Last Updated: 6/23/01 1:24 PM
TEC-II Tuning Forum
Frequently Asked Questions 2.0
Click
on the item in which you’re interested
General TEC Facts & Impressions
Particular Tuning for Conditions
Particular Tuning for Problems
V. Dealing with Hardware Problems
WRX007: 5 hours for the turbo TEC kit.
Joe Hogan: Full day
jhuang76: It took pretty much one day only, including a few mishaps on my end.
R Diamond: A full day at Rallispec.
See NickCat’s photos at:
http://albums.photopoint.com/j/AlbumIndex?u=1125321&a=8287653
Fuel Pump install photos:
http://albums.photopoint.com/j/AlbumIndex?u=1125321&a=10140456
Jammy fuel rail photos:
http://www.hools.com/fuelrail/
http://members.home.net/auto/info/tec.htm
Source: Joe Hogan
This is a project for persons that can follow directions well. If you "wing" it chances are good you'll get in trouble. You need to know the basic engine components and there function. Also, have a big breaker bar for the center nut of the crank pulley. Be prepared for a full day of work as well.
However, this is basic work that should be able to be completed by any person comfortable around the engine compartment. The tuning part is only slightly more complex but easily understood once you work through a few adjustments with Shiv.
Source: Sean McElderry
Everything is bolt on. Shiv's kit is complete with all parts needed. That means all the sensors, the wires, the injectors, and the crank trigger. Everything. It is a very well thought out kit. Installation took a full day. It would have taken less time, but Shiv's ultra-awesome, portable compression tester decided it would be cute to break off inside my engine. After Shiv fished that outta there, the rest of the day went smoothly.
Source: Shiv
With our TECs mounted on the firewall, it's virtually impossible for them to be damaged in a rally situation—unless, of course, they are mounted right next to the red-hot turbine housing. Our mounting location couldn't be more TEC-friendly. The only thing to be concern with is water. To safeguard against this, it is a good idea to brush some rubber paint (like liquid shrink-wrap) on the wire terminals. That's about it.
Source: zzyzx
Those little highly irritating 'snaps' that hold the plate the covers the gap behind the firewall. You know, the ones you have to remove to get 'behind the firewall' access to drill the mounting holes for the TEC-II.
Source: NickCat
Little tip for working with those. I believe it's either 5mm or 6mm socket, but take the socket and slip them over the barbed end under the lip then they just slip right out. Took me about 15 minutes of swearing to find out that little trick when I kicked the toolbox and shot sockets everywhere!
Source: Joe Hogan
My turbo has been an incessant leaker of oil since installed. Actually the installer claimed this would clear up after a while. Untrue—it just got worse. Well finally I had enough oil stink. After some discussion with many persons I decided to reroute the oil drain line to see how that might help.
It appears after running a XRP hose from the turbo directly to the oil pan between the exhaust header and motor mount the copious amounts of leaking oil has subsided. There is now only an occasional faint smell of cooking residue. This fix sounds easy but it took about 3 hours of twisting and thinking under the car to get the hose to fit in this area. There ain't much room in there. But if your turbo is puking oil you might try to reroute the drain. If you’re interested in the particulars please email me. Gravity is your friend. Postscript: Three days after the "fix" there’s no oil smell.
Source: Shiv
To do the above, I had to cut a good bit of length off the braided steel oil line that Minnam supplied. Properly trimmed and routed between the header and the motor mount, it's a straight shot from the turbo to the oil pan. With the standard oil line length, it had to make a big, gravity-defying horizontal death spiral around the crossmemeber/suspension on the way to the oil pan. No way oil was draining properly. Especially when it was cold and thick. This is a far better solution than putting a restrictor in the oil feel. Please do it! Your turbo (and nose) with thank you
Source: Shiv
Like a goofball, I took it upon myself to install the coolant temp sensors in the new aluminum adapters. As if you guys couldn't do it yourselves—duh. However, I used Teflon tape to seal them. Little did I realize that the Teflon tape likes to rip when the sensor gets screwed in. I just noticed this while looking at my car this afternoon. Without a proper seal, the coolant assembly will leak slightly... drop.. drop.
To fix my error, you guys will need to remove the sensor and use pipe thread compound paste instead of Teflon tape. This will ensure a tight seal. Also, be sure to tighten the sensor down really good. You may need to lodge a long screwdriver in the fitting while holding the sensor which a wrench in order to get the necessary leverage.
Please contact me if you have any problems to experience any leaks. Keep in mind, that a leak here would be very slight, so look carefully.
Source: Shiv
Look under your hood for the vacuum hose diagram. You will see something labeled the "Purge Solenoid Valve." It has one hose going to it from the intake manifold. It also has a hose coming from it and going into the "Purge Canister." The valve goes in this second line, configured so that boost pressures from the manifold cant go into the canister. Air should only be able to travel from the canister to the solenoid, and not vice versa. (Air can only blow from the white end to the black end of the check valve.)
Source: Joe Hogan
When I was N/A the air box on the TB had a nice casting blank. The blank was left of the TB facing forward as you look at the engine from the front of the car. It looks like a small circle of flat plastic. The MAT sensor, after drilling an undersized hole, screwed right in to the plastic. A little thread sealer was all that was needed to provide a snug fit.
Source: Shiv
You can install the MAT sensor anywhere you want in the intake tract. My only suggestion is to make sure that you mount it in a place that gets constant airflow. If it's mounted in some obscure nook and cranny, it's possible that it will sit in heated, stagnant air and not respond to temp swings. I've seen this before. Best to keep it mounted in a smooth tube (i.e. where the MAF would otherwise be).
Source: NickCat
That's it... now my car
pressurizes to the stock 38, but once on I get a pressure of about 41. Not too
bad for a complete drop in replacement. Install was easy and took just a little
over an hour cause I was going slowly. Good luck, hope these instructions help.
Source: Joe Hogan
Now that I'm investigating spark plugs, I thought I find the answer for a "simple" question about how to screw in a spark plug.
The NGK torque spec for 14mm shaft plugs is 18-20 ft. lbs. However, I've confirmed with the NGK tech, John (877-473-6767), that this spec is for a "dry" fitment. A lubricated thread takes less torque than a "dry" thread to achieve the same "tightness." Most of the people on this Board use anti-seize when installing plugs, so instead of guessing at the torque value, John offered this guidance: "Turn the plug 1/2 - 2/3 of a turn after the gasket touches the plug gasket seat." This event can be identified when installing the plug by hand. Once the plug is threaded as far as you can by hand (and insure that it is not cross-threaded) the gasket should be on the seat. Turn the plug with a torque wrench set at 20 ft lbs preferably. This is done to insure that you do not exceed the recommended max torque value. Once the plug has turned 1/2-2/3 of a turn you are done. This should be tight enough to hold the plug in place and more importantly allow the heat to transfer to the cylinder head correctly.
Just a bit of additional detail, the NGK plug used in the Subie is of the flat crush gasket type. Other gasket types use different values/methods. Therefore, this information may not transfer to other part number plugs. Be sure to read the instructions.
Source: efoo
I intend to tie up the bundle of wires close to the TEC-II and nail the bundle someplace solid so it takes the stress from vibration instead of the connections themselves, along with finding some kind of silicone to coat/weatherproof all the slightly exposed stripped wire ends where they enter the TEC-II unit itself. And of course, use lots of electrical conduit everywhere else to keep water away from the wires. I had the conduit, but didn't have the wires protected well enough where they actually ran into the TEC unit.
Source: wac
The terminal blocks used in the TEC-II kit produce gas-tight electrical connections if installed properly. I know, because I've worked on electrical equipment used in harsh environments. It goes without saying that this is not like your typical stereo install.
For the terminal block connections, soldering isn't at all necessary, and may result in a poorer connection than bare wire conductors. Simply stated, the terminal blocks are designed for bare wires.
A soldered wire end may feel solid when you screw down into it, but high temps and/or vibration may cause it to deform and lose much of the initial clamping force. Solid copper wire resists deformation much better, and will retain most of the initial clamping force. Another consideration is that soldered wire terminations cause a lot of stress at the interface where the solder ends. Without strain reliefs, it is usually the first place that soldered wires fail under vibration and stress. It makes sense that stranded core wires flex much easier than solid core wires, so why would you want to remove that benefit of using stranded wires?
"Boot-lace ferrules" may also work well, but they really need to be crimped before insertion into the terminal blocks. Since they make life a little more complicated than it needs to be, I'd skip it.
In summary, strip as much as necessary to make sure that all the wire strands make full contact with the clamps in the connector. Give an extra 1/16" or 1mm gap if necessary. If you break a strand or two while stripping the insulation, snip the end off and try again. A $15 pair of quality "T-Strippers" is well worth the investment for a job like this.
Source: Shiv
The fuel pump upgrade is not necessary for NA applications. In fact, it's only above 7-8psi that a fuel pump upgrade becomes necessary.
Source: Shiv
I would suggest that the EGT and boost gauge are the most important. EGT and 02 sensor voltage, used together, make a powerful tuning team.
Fuel pressure is a good idea, especially when pushing the limits of your fuel system. Oil temp is a good idea if you plan to drive on the track, or under severe conditions. Same goes for water temp. For most people, these gauges aren't critical. The a/f ratio gauge is cute and somewhat helpful in noticing strange changes that would be cause by a failing injector, a dying fuel pump or a clogging fuel filter. The TEC-II monitors this so you don't have to (as long as you are within the limits of the injector/fuel pump).
So, if I could get 3 gauges, I would opt for boost, fuel pressure and EGT. For track, I would definitely get an oil and water temp gauge as well. An oil cooler, like the nice air/air unit that Rallispec sells, would also be a good idea for competition. Don't forget to run a good synthetic like Redline.
Source: Shiv
I suspect the HKS plugs are the same as the ones from Denso. Either way, they seem to improve combustion for spark-limited ignition systems. For the TEC-II users out there, I think they are unnecessary since they have more than enough spark to bridge the plug gap. Can't imagine ever becoming spark limited. In this case, I would suggest getting colder range traditional plugs instead and put the money saved into gas money.
Source: R Diamond
The part number for colder NGK platinum plugs is: BKR7EVX-11 (stock #4685). Cost: $47.50 from Teague’s Auto.
Source: Z1500
Higher static compression creates more power throughout the rpm band, but it'll lower your maximum allowed boost before the onset of detonation. Boost is worth way more power than compression, because boost raises your compression and your total airflow at the same time. With the down side of, when you're not on the boost, you have slightly less power.
Effective Compression Ratio = static compression ratio x (1 + boost/14.7)^1/2
For a car running 8.5:1 pistons and 18psi (~max on pump gas):
8.5
x (1 + 18/14.7)^1/2 = 12.67 ECR
If you run 9.0:1 pistons and want to maintain the same 12.67 ECR (~max on pump gas), you'll have to lower your boost to: 14.4psi
[(12.67/
9.0)^2 - 1] x 14.7 = 14.4 psi
So you have to run 3.6 psi less boost to maybe pick up a tiny bit of bottom end. Or to take it even further for 9.5:1 you can only run 11.4psi. I'll tell you right now that the difference between 11.4psi and 18psi is huge. And at some point (12.67:1 in this case) you can run no boost and be maxed out on ECR for pump gas. And how fast is a N/A car with 12.67:1 compression, ask the Honda boys running 15's.
Of course you get to a point where you don't want to go the other way too much ether. A 6.0:1 car isn't too much fun on the street, but it can run 50.8 psi boost on pump gas. But the kind of turbo that could support that much boost would never spool up. You have to find what's right for you, but 8.5:1 is what most of the DSM guys like. My friend just built a motor for his DSM with 9.0:1 JE's and he is quite upset to find that he's getting lots of knock running 16psi. While lately in the cold weather, I've been running 20psi (8.5:1) on pump gas without any knock. And I drove his car and the difference in bottom end grunt isn't even noticeable. And his top end performance is sad compared to mine.
Source: SFSube
Key investment: a power inverter for your laptop. I have one that plugs in via the cigarette lighter and allows you to run anything (within reason) that has a regular plug, like a laptop power supply.
Source: Jed
Hi there. Thought you guys might be interested in my thoughts on WinTEC 2.0. Hope I don't go on too long -- I'm rather excited.
General Comments
WinTEC 2.0 at first glance does not seem like a major upgrade: the interface is the same, the feature list feels similar, and we’re still stuck with the same kludgy help system. Documentation remains cryptic at best. We even still have that embarrassing start up dialog box with the pictures of the drag race cars and the truck spinning out that has often lead my commute partner to emit engine revving noises when I make datalogs on our ride home.
But in the same spirit as the earlier Electromotive products, the best stuff is on the inside. WinTEC 2.0 extends the resolution of the TEC II platform and strives to engage the tuner to an even greater degree than earlier releases. The additional feature set is targeted aggressively towards this goal and barely gives a nod to the more appearance-oriented improvements to competitors such as Link.
New ONFLY Mode
The ONFLY mode is the most crucial feature addition to WinTEC 2.0. This new mode allows real time changes to almost all engine parameters while the car is running. You can then save changes and create a new baseline while on the fly. As Shiv alluded to, this feature allowed him, for example, to set much more aggressive minimum injector on times in his Miata and let him try switching to a divide by of one rather than two in setting his injector firing pattern.
Very few settings are barred from changes during ONFLY mode and they are all labeled carefully in the dialog boxes to remind you of this at all times. The notable exceptions to ONFLY editing are changing breakpoints for MAP and RPM (now you really wouldn’t want to do something that dramatic while the car was on, would you?) and calibration of your EGO settings (more on that later). Once you see these exceptions, if you think about them, they all make good sense.
It’s difficult to express how big a deal ONFLY is – it’s very probably going to change the way everybody tunes for the better. You’ll find you can make your changes much more quickly because you’re not constantly toggling between edit mode and datalog mode and then trying to reload parameters into your TEC. I imagine you’ll find that you don’t datalog nearly as much as you did under WinTEC 1.1.6, since you can try so many of your settings out while the engine is running. Think about it – you can define all the parameters of your fuel curve on the fly. And of course, the ability to quickly change timing on a dyno may quickly save you the price of the upgrade in dynamometer time fees. Shiv has noted that there is currently a small bug here in the beta version of the software.
New Auto Calibrate VE Tables
Feature
Here is the nod I mentioned to other program’s new feature sets. It makes great sense to add this feature. Now you don’t really have to think about adjusting the VE inputs to your table – just have WinTEC do it for you. You can choose to have WinTEC monitor your current ride or simply read in a datalog as an input for suggested changes. It’s pretty cool when you feed a datalog into the feature – it chugs along for about 10–15 seconds and the suggested changes appear in yellow on the VE table. You can accept them all, or accept/reject on a cell-by-cell basis.
This feature, in conjunction with ONFLY mode (which gives you the ability to modify AFRs on the fly), greatly quickens the tuning process. Admittedly, VE tuning was among the easier parts of learning to tune, but time savings are always appreciated.
New Enhanced GPO Control
The new GPO control still fiddles with the TEC-II’s one GPO output. We’ll expect more when the TEC3 is introduced. For now, Electromotive has simplified the operation of other controls using the GPO by creating preprogrammed forms for AC, Auxiliary fuel pump, fans, intake runners, Nitrous Oxide, shift light and torque converters. The WinTEC 1.1.6 GPO could have controlled pretty much all of these devices, but the new forms make it easier and more intuitive.
Source: Boost
I used the WT 2.0.1 for the first time this weekend, and I am very impressed!! The Autotuning was awesome, and the on-the-fly adjustment saved me a lot of time. I tuned the car in a fraction of the time I used earlier, and the result was at least as good as the old time-consuming way to do it.
Source: Jed
I'm certainly willing to pay for this upgrade. I think it's well worthwhile. The ONFLY mode will undoubtedly improve the state of tune on our cars.
Source: Jed
If you’re upgrading from WinTEC 1.1.6 to WinTEC 2.0 you will definitely need to change your EPROM. It could be the beta nature of my install, but there are no instructions on how to do this. I did it this way:
1. Don’t be having just rubbed your feet on a shag carpet. Avoid static.
2. Disconnect the battery.
3. Disconnect the wiring clips on either side of the TEC.
4. Dismount the TEC from wherever you have it mounted.
5. I left my power lines connected and had enough room to manipulate the TEC.
6. Unscrew the 4 screws on the top of the gold colored portion of the TEC.
7. Some people may need to remove a coil to get access to the 4th screw – I didn’t.
8. Gently separate the 2 halves of the TEC from each other. They are held together by chip pins, so I can’t emphasize enough the importance of being gentle and not breaking the pins.
9. Pull the chip. A chip puller would be great here, but I used a small screwdriver.
10. Replace the chip with the new one. Don’t be bending any of the pins when you insert it.
11. Reverse steps 8 through 1
Source: AustinP
I just finished doing the WinTEC v2 upgrade, which went fine, but a couple watchouts:
Source: Jed
Carrying over EGO calibrations when a 1.1.6 file is imported into 2.0 would be difficult, as there are now twelve setting points for AFR vs. volts, rather than WinTEC 1’s 8. This allows for finer resolution of the EGO. But the new settings create more of a spread between volts and AFR. The good news here is that when we discuss our AFRs with our Link and Haltech buddies, we’ll be speaking the same language. The rather alarming bad news is that WinTEC 1 files loaded into WinTEC 2 will drive AFR targets more tightly clustered around stoichometric than they did previously under WinTEC 1. While the difference is not dramatic, it still may represent a danger for users bringing over their old maps for high boost forced induction applications. The net net is that tuners must be careful during their initial runs under the new software if they are importing files from earlier versions. Per Shiv, Electromotive may try to fix this issue, but I think it would be hard.
WRX007: “the car feels great, no pings, no flat spots or backfire wow!!! it is perfect”
NickCat: “but right now I am running almost completely from the map originally provided by Shiv except for the tuning I did in the VE map. My car at wide-open throttle runs near 0 EGO correction. I tuned 65kPa a little but mostly just in the mid range, it's hard to keep the car at 65kPa while driving around data logging.”
jhuang76: “As far as test driving it goes, WOW!!! Using the most basic setup that Shiv gave me, the car pulled very strong, especially in the midrange. It feels even stronger than right after an ECU reset. For example, my car can now easily accelerate on the highway without having to downshift. I'm going to advance timing a little tomorrow and that should get me even more power. By the way, I've had the stock air intake and exhaust setup and I think the air flow restrictions are really preventing high-range performance increases.”
Imprezer: So here are my thoughts on Vishnu TEC II NA Kit after installing it on my RS. The car's power is definitely increased throughout the RPM range. The power doesn't fall after 5000RPM as suggested by some US Impreza tuners. The car feels much smoother even with super noisy Borla headers. The car's throttle response is much better than with stock ECU. The car is much more fun to drive with lots of torque and power across the RPM range.
Sean McElderry: The result? Awesome. The car runs very well. No detonation. 8psi of boost. Smooth as silk in all modes of operation (high or low load/high or low RPM). And the best part, completely tunable from now on.
SFSube: My car is now running perfectly! Smoother than a factory turbo with no drawbacks. I've never driven an aftermarket turbo even close to this refined. I wish I could stop grinning like a nutcase, but the feeling of the turbo push, push, pushing on my chest as I whip down 280.
Kartboy: My turbo isnt installed yet, but I have been driving my car with the TEC-2 installed for a few months and have had nothing but great luck with it. I haven’t had the laptop in the car for a long time. It starts and idles better than it ever did stock.
Joe Hogan: Actually it is taking me some time to readjust my throttle application. Last night on the ride home I was attempting to fit into traffic after a toll booth on the New Jersey Turnpike and spun all fours going into second gear!
Source: Shiv
At stock fuel pressures and with a sufficient fuel pump, the 550cc injectors should be able to safely support 330-350hp on 92 octane. With higher octane, a little bit more power should be possible. As for maximum boost, that's a good question. I don't know since none of us have been running turbos for that long. I can say that I, after 30k miles of heavy turbo use, have yet to see any signs of engine wear (no oil consumption and good compression). I don't see why 100k isn't possible. It really all comes down to tuning and safety margin. That's why I suggest installing and familiarizing oneself the TEC-II before going turbo. And once turbo, start off with low boost and gradually work your way up. Still, I'm not entirely comfortable with running much more than 10psi on 92 octane (be best we get here in CA). But then again, I'm reeeaaaaly conservative.
Source: Shiv
Just got back from the dyno where I ran my own MY99 turbo, which was recently converted to FWD by ISR. Boost level was 7-9psi. That is, 7psi up to 4500rpm, ramping up to 9psi by 6000rpm and then back down to 7psi by 6300rpm. I found this boost curve to give the nicest and least stressful driving characteristics. The results:
The car actually hits 252ft-lbs at 3800rpm, and then falls down 20ft-lbs for a bit, then back up to 256ft-lbs at 4600 where it remains essentially steady until 5500rpm or so. By 6000rpm, it falls off to 225ft-lbs. Not much top end roll off at all.
The actual, as measured, output was 268 wheel hp and 263ft-lbs. But since the temps were quite chilly today (60F or so), the Dynojet corrected down with a .97 correction factor. Either way, the car gets up and goes really well with peak EGTs (top of 4th gear) of only 1450F.
A stock 2.5RS dyno's at around 120-wheel hp. Also with 1450F EGTs
Matt Cavanugh-- 5 speed
MY99 Minnam turbo/TEC-II, 7psi, 3-inch exhaust, 93 octane pump gas. Tested with
FWD configuration: 224 wheel hp @ 5800rpm, 230ft-lbs @ 4600rpm.
Edwin Foo-- 5 speed MY99 Custom turbo/TEC-II, 7psi, 2.5-inch exhaust, 93 octane pump gas. Tested in RWD configuration: 228 wheel hp @ 5800rpm, 245ft-lbs @ 3750rpm.
Dyno charts:
http://www.vishnuperformance.com/ubb/Forum1/HTML/000615.html
Source: NickCat
I was g-teching the other day and the best time I got was a 5.5s 0-60 time. I brought the car to about 4K and slipped the ACT clutch out pretty fast, the hardest thing is getting a nice smooth shift into second because the car lurches so much.
From the datalog I am running about 9 psi, but I suspect that I am running horribly rich during high boost. The car has very little difference in feel between 7 and 9. I will be tuning the fuel delivery soon much better very soon.
Source: Sean McElderry
5.5s is not bad. Remember, our
sweet US trannies require a 2-3 shift to hit 60. This will make your time
higher than other cars with similar power levels. And 5.5s to 60 is really not
bad if you're not dropping the clutch. You'll see a large improvement if you
drop the clutch and shift quickly. But don't do it unless you have money for a
new tranny (just in case).
On my second run ever, I slipped the clutch from 5k and had a much better start - 1.9 60' time. And this time I shifted faster. Not as fast as I would if it was a rental car, but still faster than I normally do. Ran a 13.8. I imagine I could have shaved quite a bit more time off had I dropped the clutch and shifted as fast as I could, but my tranny was more important to me than a timeslip.
Source: Shiv
Just thought I'd drop some performance numbers that I just derived from a datalog and some gearing info. Current set up is a MY99 2.5RS turbo with a MY00 EJ25, upgraded rails, Full dog box, ACT clutch and Toyo RA-1 road race tires. Peak boost is set to 10psi up to 5900rpm, above which it drops to 7psi.
Acceleration
Testing
0-30mph: 1.8 seconds
0-60mph: 4.1 seconds
2nd gear (4400-6000rpm): 1.4 seconds
3rd gear (4400-6000rpm): 2.4 seconds
Boost Rise
(WOT in 5th gear)
2000rpm 125kPa
2500rpm 147kPa
2600rpm 153kPa
2700rpm 160kPa
2800rpm 167kPa
2900rpm 170kPa
Just for giggles, I compared the above run to the acceleration data (from radar testing) of that crazy 500+hp Bozz EVO that Sport Compact Car tested a few months back. Despite a terrible bog during the launch (0-30mph took a whopping 2.8 seconds), the EVO still managed to reach 100mph in just 8.9 seconds! With quicker shifts and a decent launch, it may have hit the century mark is less than 8 seconds. By comparison, I'm hitting 100mph in around 10 seconds. That's a pretty big difference!
Source: Sean McElderry
Today, I'm going to be fine-tuning my VE table a bit more. After everything's good, the boost is getting raised to 10psi. The problem is my injector duty cycle is already at 75% at 8psi/6500rpm. I'm hoping it doesn't go past 80% at 10psi.
Source: Shiv
You can take the injectors to 85% duty cycle. You can also back down your a/f ratio targets from 13.5 to 13.6-13.65 above 5200rpm. As EGTs go up, the o2 sensor tends to read leaner than it really is. In other words, at higher EGTs (1550+ deg C.), a 13.6:1 AFR may actually be closer to 13.5:1 AFR. Just take it easy, listen for knock and keep EGTs below 1650 deg. C! 10psi may be pushing it for a higher compression MY00 running 92 octane.
There is another difference with the MY00. At 170kPa, I'm running 16 degrees of ignition advance above 5250rpm. Right off the bat, I will tell you that the MY00 will not be able to run that kind of ignition advance due to its higher compression ratio. In other words, MY00's ultimate safe boost level, all things equal, will be lower than a MY99's.
Source: Sean McElderry
I should talk a little bit about the differences on a MY00. The injectors are different, which had us wondering at first whether or not the new 550cc/min injectors would fit. In the end, they did. The MY00 wants fuel—and a lot of it. It has to run richer than the MY99 to keep the EGTs down. The compression ratio is higher, which might mean I have to run a bit less boost than the MY99 folks to be safe. In all, the maps for the MY00 and MY99 are completely different.
Source: Shiv
That's a real tough question to answer since there are so many variables involved. But generally speaking, I would suggest that an MY99 should be limited to 10psi on 92 pump. Maybe 11psi on 93 pump. Higher compression MY00 cars should stay 2psi lower. Keep in mind, when properly managed, 8psi is still honkin' strong. Stronger (and definitely safer) than running 16psi with stock injectors and stock ECU. But, as always, these limits should be approached slowly with a good deal of diligence.
Source: Shiv
I wouldn't recommend running anything more than 10psi on pump gas. I learned that while trying to run 12psi. I had a hard time keeping EGTs down and power up. I think a well-tuned 8-10psi system will be more entertaining.
As for unleaded race gas, I suppose it is possible to run more boost. However, this will require some leaning out of the fuel calibrations. Specifically, the TOG will have to be reduced a little bit with appropriate changes to the Injector Offset Time. The VE table can be left alone. You will probably be able to advance timing (in the 10+psi rows) a few degrees as well. This is how my car went from 270 wheel hp (at 7-9psi) to 295 wheel hp (at 10-13psi). I would strongly suggest running less boost in the midrange (up to 5000rpm), making peak boost by 6000rpm. With full boost in the midrange, your engine will make a lot of torque making life difficult for itself as well as the transmission.
Source: Shiv
I get 30-31 mpg on highway. During highway cruise, a/f targets are 15:1 and ignition advance is around 35 degrees. In town fuel economy can vary from 18 to 25 mpg depending on how heavy my right foot is.
[responding to someone with 23 degrees of advance at cruise—3k RPM and 84.3 MAP.] With more advance during cruise conditions, you'll probably find yourself being able to sustain a 75 mph cruise at just 65-70 kPA. Barely any throttle. Keep adding ignition timing until you feel no additional benefits. While it's unlikely that you'll get knock during cruise, going too far with advance will result in less torque. Let us know what you find!
Source: Joe Hogan
Today the fuel economy approached 24 MPG. This included some highway driving, eight 40-second AutoX runs and some local driving. To achieve this fuel economy the spark advance was advanced to 24 degrees at 750 rpm and 38 degrees at 2500 RPM up to 80 kPa then tailed off to 21 and 22 degrees respectively at 104 kPA. The mixture is set to 15:1 in the mid range moving to 14.35 at 104 kPA. By the way, the engine is N/A. The engine pulls strong and does not ping. Actually the spark advance may move 1 or 2 more degrees in the 750 RPM column.
Source: Shiv
I think all OEM maps leave a bit
to be desired -- Subaru's included. I guess they account for those grannies
that lug their engines in 5th gear at 20mph, running Rotten Robbies 87 octane
with oil that hasn't been changed in 2 years. They also seem to sacrifice a lot
of throttle response and partial throttle torque-- probably to keep emissions
down during EPA testing. While awd Subies are hard to dyno, I suspect their
big, low output 2.5L engines are even more detuned than the Miata engine. Where
else can you get 15% more power at the wheels through a free flow intake?
Yeesh...
Source: Efoo
Essentially, you can log everything the sensors can see. For Shiv's kit, that means RPM, MAP, MAT, TPS, O2 sensor, AFR, EGO corrections, knock sensor, etc.
Source: TEC Manual
You can datalog for up to 27 hours straight with the current software.
Source: Shiv
Nope. It's not going to be CARB approved. “For off-road use only.” This is because phase sequential operation creates more pollutants during cold (when the valves are cold, atomization is compromised), which is critical to EPA testing. Just another situation where performance takes a back seat to emissions (even if it is just for a few seconds during cold start). If anything, the car should sniff test cleaner with the TEC-II than with the stock ECU.
Source: Gotcha
I pulled the codes in RS'ted's 00MY RS with now a Minnam turbo kit and TEC-II. There were only 2 codes. “High MAP reading” (no big surprise there) and “no A/F sensor reading” (imagine that, guess the OBD couldn't connect to the A/F with laying over on the workbench). But those were the only codes; I think that speaks pretty highly of the quality of the TEC-II. If it can make the Subaru ECU happy, it can do anything.
Source: Shiv
All o2 sensors, used by themselves, are inappropriate for tuning. This even applies to the wideband units that are so popular among aftermarket tuners. These sensors simply reading the wrong thing (o2 content) and are influenced by a number of other things (ignition advance, EGT, combustion characteristics, etc.) This makes them less than reliable. An EGT gauge offers far more pertinent information. Used together, they make a powerful tuning tool.
Source: Shiv
It shouldn't be a problem. I've had to jump other cars in the past with
no problems. Just make sure your voltage isn't low in the first place or the
TEC won't get enough voltage. The TEC needs at least 9v to run. Chances are, if
your car is running, it will see more than that-- even if it is jumping another
car.
Source: Shiv
I suspect a lot of what you are experiencing may be
psychological. At 20% throttle, you're probably seeing close to 95% power.
That's just the nature of mechanical throttle bodies. Also, at full boost,
wide-open throttle, you may be activating enrichments which temporarily reduce
torque by over-fueling. Check your AFRs. Same thing goes for ignition timing.
Wide-open throttle could be inducing slightly higher MAP, which would be
reducing you timing a few degrees, which is quite noticeable under boost.
Again, do datalogs and compare these values in wide-open throttle and partial
throttle situations. I suspect a lot of what you are experiencing may be
psychological.
Source: Shiv
It's actually a coil and fuel cut. (There is an option for ignition retard but I don't recommend it for manual transmissions. It's really only used for launch control—staging boosted drag cars with torque converters.)
You can also set a 2nd rev limit which is triggered by shorting the manifold air temp signal to ground. This can be accomplished by rigging up a simple toggle switch. With it set to, say 3000rpm, you have a valet mode.
For drag racers, this auxiliary limit can be triggered by the clutch switch. This would allow you to keep wide-open throttle between shifts as the revs would instantly drop down the aux. rev limit once the clutch goes in.
You can also control the hysteresis of the rev limit. This dictates how "smooth" or "rough" the limit is. With a small hysteresis, you won't even notice that you are up against the rev limit. The revs will simply stop. With a bigger hysteresis, it will bounce back and forth at the rev limit (like stock)... whap whap whap...
For what it’s worth, I like smaller hysteresis. Feels easier on the tranny. I guess the stock rev limit was designed to be obvious to even the most unobservant drivers
Source: Shiv
I'm sure I'm more conservative than others but my car is
set to 6600rpm. The setting I provided in the baseline program should be a
stock-like 6250. For most modified N/A cars, there seems to be sharp torque
roll off just after 6000rpm. So a higher-than-stock redline may not be
justified. Turbo cars seem to hold up a bit better so a few extra revs may not
be such a bad idea.
Source: Shiv
Just a suggestion guys: when you order stuff like idle control motors,
software, magnetic pick-ups, etc., go through me instead of Electromotive. Not
only are my prices lower but you will also get the right part since I'm more
familiar with the car and your set up than Colleen is.
Source: Shiv
One of the nicest things about good engine management is that you can give up compressor and intercooler efficiency and make up for it in good EFI tuning.
I find it funny when turbo kit designers address efficiencies and then slap on a rising rate fuel pressure regulator and a high flow fuel pump. I've always believed that the engine management portion of the kit is the most important component--by far. A good example is my Miata. I'm using the least efficient form of forced induction (Eaton Roots-style supercharger) without any form of intercooling. Measured compressor efficiency is in the mid 50s with 10-15hp of belt-drive parasitic losses. At 4-5psi of boost, intake temps are 170F. Yet, the car is running 25-30 degrees of timing at wide-open-throttle and making nearly 180hp to the rear wheels without ever pinging.
Other people are running mega-buck ball bearing turbos, front-mount ICs, lots of boost, and still making less power and suffer from catastrophic engine failure. They are using piggyback additional injector fuelers and ignition controls.
Source: Shiv
A good portion of the throttle response and low/midrange torque gain is a result of the TEC-II's phase sequential injector firing order. One out of two fuel squirts is fired at the back of the intake valve. Since the valve is hot, the fuel atomizes immediately. Next injector firing squirt happens during the intake stroke, when the inrushing air pushes both the new squirt and the previous pre-atomized fuel vapor into the combustion chamber. The entire fuel/air mixture swirls and mixes together. Then the piston starts to come upwards, compressing the intake charge. When the spark is fired, it is fired for an unusually long 120 deg of duration. At lower engine speeds, when air/fuel swirling/mixing is not complete, the long spark helps facilitate the entire burn. All these events improve low-end to midrange power and throttle response. In other words, I don't think fuel or ignition alone can take the credit.
It's more than just two squirts that matters—it's when the two squirts occur. I don't know of any stock engine management system that runs phase sequential fuel injection. They almost always run full sequential instead. This is because phase sequential operation creates more pollutants during cold start (when the valves are cold, atomization is compromised), which is critical to EPA testing. Just another situation where performance takes a back seat to emissions (even if it is just for a few seconds during cold start).
Source: Shiv
Appropriately sized injectors make a world of difference. Both in terms of performance and drivability. Stock MY99 injectors flow approximately 280cc/min according to the flow bench at RC Engineering. Perfectly sized for a 165-185hp car. Still too small for any turbo system. With 8psi of boost, my injectors went wide-open at 3500rpm.
Before I put the 550cc injectors on my car, it seems as if my 7psi turbo was operating at a higher duty cycle than I once thought. Initially, I assumed the Impreza RS had 320cc injectors. Turns out, after measuring them, they were actually 'lil 280cc squirters. With the TEC installed, I recall them reaching a nominal 100% duty cycle at 4000rpm or so. With higher fuel pressure (80psi), they reached 100% duty cycle at 5000 rpm. Simply unacceptable. They only reason the injectors can provide fuel (sort of) for a 6-7psi turbo system is by going static. This means that instead of squirting during the intake stroke, they keep squirting fuel all the time (even during the exhaust stroke when the intake valve is closed.) The fuel puddles behind the valve and eventually gets in the engine during the next intake valve opening. With big 550cc injectors, running phase sequentially, things are a lot cleaner and better controlled. I'm reaching 80% duty cycle by 6600 rpm. This is while running 12psi of boost.
Source: Shiv
There are two basic types of injectors, high and low resistance. The former is rated at around 13ohms while the latter is around 2.5ohms. Most OEM cars use injector drivers designed to drive high impedance injectors. If they attempt to drive low resistance injectors, they will likely be damaged. Most aftermarket EFI computers have beefy injector drivers that are capable of meeting the current demands of driving low resistance injectors. They can also drive high resistance injectors as well (as the TEC-II kit does).
Which is "better"—low or high-resistance injectors? In big injector/small motor situations, low resistance injectors will always be better, offering faster response (lower ramp up time) and better overall control. For a Lucas type injector (RC), a low resistance injector can operate with on-times as low as 0.9-1.2ms without binding up. High resistance types, on the other hand, can go down to 1.6-2.0ms. This is important when trying to achieve a nice low, stoich idle with big injectors.
For what it’s worth, with a 750rpm idle, my low resistance 550cc injectors operate with a 1.5ms on-time—still well above the point in which I start running into trouble. With the ability to go lower, I could probably run 650cc injectors with no idle/drivability problems. However, if the TEC-II could only drive high resistance injectors, I would already be at (or beyond) the point of trouble. By "trouble" I mean having to live with a rich, non-stoich idle. One solution would be to raise the idle a bit to, say, 1000rpm. With at-idle on-times behind closer to the injector’s minimum injector on-time, the problem would be "fixed". Another alternative would be to live with a rich idle which may eventually lead to clogged cats.
Source: Shiv
Boost-dependant rising rate fuel pressure regulator are semi-effective ways of supplying fuel in aftermarket turbo/supercharged cars. As manifold pressure goes up, so does the fuel pressure—usually about 7 psi of fuel for every 1 psi of manifold pressure. This tactic is used in stock-ecu'd, stock injector'd applications where the stock control system can't compensate for boost by itself. In fact, the stock computer still operates the injectors as if the car was still stock (for the most part, anyway). Increasing fuel pressure is a "back door" way to increasing injector flow without actually changing injectors.
The proper way to increase fuel flow is by running a constant fuel pressure and larger injectors. To run the larger injectors, you need to re-map the fuel curve, which requires a fully programmable ECU. If you can do this, there is no need for a rising rate fuel pressure regulator. In fact, if you used one, it would make your tuning more difficult.
Source: Efoo
The VE table is definitely modifying gamma. It's a pretty neat concept actually. Think of it like you have a rubber sheet stretching over all the various axes of engine parameters like rpm, MAT, MAP, etc., and the TOG is a first cut at making the sheet stretch to fit over the optimal performance curve. Of course, it's not a perfect fit, so you have the VE table and IOT to shift the whole sheet around (IOT) and VE lets you poke and pull at various points in the sheet to make it fit certain curves better.
In mathematical terms, that TEC2 equation tries to approximate a fuel curve that depends on a lot of factors like RPM, load, etc. The real fuel equation is some super long monstrosity that includes all the above variables, but the TEC II equation simplifies it down to a simple linear equation that only involves the biggest variables, MAP, and TOG.
All the other variables are handled via those lookup tables for offsets that modify Gamma. So it doesn't look nearly as confusing. Of course, that may not make things look any easier for you, but at least think of just how much more complex it would be if you had to solve an equation with 10-12 variables in it instead of being able to tune using a few at a time!
The biggest implication of this approach, IMHO, is that it totally matters which order you tune the variables in. For example, so far, Shiv has been making us all tune the VE tables at WOT first before doing other things like ignition advance, MAT compensation, etc. If I were to tune using ignition advance first, my final mapping could look significantly different, but produce the same results. The point here is that for this system of equations, there are at least n different solutions, where n is the number of variables in the system. By starting from the same baseline map and tuning the same variables first in the same order, we will have tuning experiences which are similar and therefore numbers that work for one of us should work for others. Therefore, it's important that (a) we follow the same steps that Shiv took and (b) Shiv took the right steps to begin with using his knowledge of which variables have the most impact on engine performance and reliability (like EGO CR, knock sensing, etc.).
Source: Shiv
For the TEC, you only need to plug in two numbers for the raw fuel curve—like a slope and a Y-intercept. The Y-intercept (injector offset) effectively tunes for idle and the slope (time-on gamma) tunes for load. Takes about 10 minutes and is the foundation for fuel delivery. Unlike other systems, the TEC (like the stock ECU) does not operate on a conventional map, constantly linearizing between breakpoints. Instead, it uses this simple algebraic running fuel algorithm, which is very quick and hard-wired. This gets you 80% there. The rest of the fuel tuning is done in the VE table, which tunes the raw fuel curve for the VE characteristics of the particular motor. This is a lot easier (and more effective) than tuning by punching in duty cycles or gammas in 80 or so individual cells. This is why I've always been such a big fan of the TEC.
Source: R Diamond
GAMA is not a word, either in English or Greek. Instead, I believe Electromotive were thinking of the Greek letter G, which is properly spelled gamma.
http://www.dictionary.com/cgi-bin/dict.pl?term=gama
http://www.perseus.tufts.edu/cgi-bin/morphindex?lang=greek
Source:RS'ted
http://members.aol.com/dvandrews/ems.htm
Source: Shiv
I'm afraid that I don't know of any good, modern engine tuning books. And with the variety of EFI systems available, it's unlikely that one size will fit all. The best way to learn about engine tuning is through trial and error on one's own car. That way, one can actually learn what misfire, lean run, over-retard, etc., feels like. Unfortunately, tuning is still two parts science and one part seat of the pants. This means that "feeling things out" is still just as important as studying EGTs, a/f ratios, and ignition curves.
Source: Shiv
Having the factory ECU still think that it is controlling the engine is
quite a nice thing. Because of this, the stock ECU will keep the fuel pump
running, the fans functional, all emissions devices operational, idle control
system working, etc. Much more "hack" is involved when running an
aftermarket EFI system in stand-alone mode. The fact that it is completely
isolated from the operational fuel/spark system means that it cannot "do
anything" to mess with primary engine function. No matter how confused it
gets (primarily from not getting an o2 sensor signal).
Source: efoo
Can someone explain to me why most of the maps I've seen have the amount of fuel correction drop off under high rpms? Does the engine not work as efficiently or something, necessitating less correction?
Source: Shiv
As with any pump, there comes a time where pumping/volumetric efficiency
goes to hell. Above a certain engine speeds, the air simply doesn't have enough
time to fill up the cylinder on its intake stroke. I suspect the pumping
efficiency losses are caused primarily by the air drag through the head (valves
are pretty small openings for all air to squeeze through). Cams can improve VE
by opening the valves more and/or for a longer duration. But that's a science
in itself.
While the presence of boost helps to even out
most VE humps and dips, high rpm roll off is inevitable with any motor. But it
works out okay in the end as most motors reach their peak horsepower 1000rpm or
so shy of the point at which their rods would want to fly out of the block.
It’s nice to keep a little margin of safety.
Source: Shiv
So what do we do? Simple. We err on the side of conservatism. We run a bit richer under boost (you don't need a wide band to know when you're too lean). We also run with a 3-4 degree safety zone against knock. If you are running rich, with reasonable EGTs with a 3-4 degree safety zone, you're in safety land. Sure, by running extra rich with a conservative ignition advance map, we won't be able to eek out the very last bit of horsepower. But we sure as heck will be in safety land. And chance are, we'll also have more than enough horsepower to keep ourselves happy. There are countless data points to support this argument.
The goal is to see AFRs of 13.5-13.6 under boost (in 4th gear) with small (0-5%) negative EGO corrections. Big corrections are bad. Also remember that too little ignition advance is nearly as bad as too much. Advance right up to the point of mild detonation and then retard 3-4 degrees. Once both of these areas are dialed-in, the knock sensor can be activated and tuned. Please let me know when you do this. Improper knock sensor tuning can result in over-retard, which is bad. Also, start off at 5psi and gradually work your way up.
But also keep in mind, the goal is to make safe, reliable and drivable horsepower, NOT to run X amount of boost. Boost pressures are not a good representation of horsepower. 10psi of properly managed boost will provide more HP and far better reliability than 20psi of ragged-edge, overly retarded, lean running boost.
Source: Shiv
Over-run is just a fancy word for engine braking (i.e. decel caused by no throttle opening). This yields some rather low MAP values. You will need to make sure that your decel MAP trigger is well above the MAP values you see during over-run in order for the injectors to shut off. You'll probably end up with 45kPa or so.
Source: Shiv
There is no longer any over-run backfire like there was when I was running the stock ECU/afc/itc/fpr/etc. All it took was some adjustment of the MAP/TPS-activated fuel cut-off parameter which cut off the injectors during over-run/high vacuum conditions.
I
also noticed that there is no longer any fueling anomalies associated with
changes in battery voltage. With the stock ECU, I used to run slightly leaner
under boost when the headlights were on than I would when they were off.
Changes in EGTs confirm that this wasn't a monitoring error. With the TEC, I
adjusted the battery voltage compensation accordingly and evened everything
out.
I've
also compensated for ignition timing in relation to intake temperature. Before,
during periods of heat soak, the car would want to ping. Now, ignition timing
gets retarded as a function of intake temp (as well as MAP and rpm). Same with
fuel delivery. Before, I also used to
run into transient lean spots during sudden openings of the throttle. This
caused both knock sensors (stock and J&S) to jump in aggressively and sap
power. Now, I have activated an acceleration enrichment which is triggered by changes
in boost as well as throttle position. The result: no lean spots under
throttle/boost stabs. I've also tuned
the o2 sensor feedback system to always stay closed loop. This can be done
without excess, slow and annoying o2 correction as the base fuel map is now
dialed-in for the particular turbo motor. This means that the o2 sensor is only
used to make small (+/- 3 to 4%) corrections.
It's
also interesting to see some quite dramatic "steps" in the engine’s
VE—not only as a function of RPM, but also as a function of load. As a result,
my VE table looks strange but seems to work rather well. Also gone is the gross
over-retard I used to feel (from the stock knock sensor) in the 4-5k rpm range.
With that now gone, torque is up substantially in that range. Highway pulling
power is exceptional.
Source: Shiv
You can, but every car seems to be a little different—even same model year cars with same mods. It's also a good idea to invest in some learning time so you'll be comfortable with making future changes when/if you go forced inductions—hence the idea of starting off with a baseline file and moving up from there. When boosted, the margins for good, safe operation are too small to rely completely on someone else's map. While it's possible to have a one-size-fits-all map, it would be detuned, slow responding and a bit soggy in an effort to account for the lowest common denominator (highest actual CR, lowest octane gas, laziest injector, etc.) What's the fun in that?
I'm of the opinion that you will need to learn the system to extract maximum safe performance. It just takes time. But from trial and error (and support from the forum), you'll get there. No one said programmable EFI is easy. But it certainly ain't rocket science.
Source: Shiv
Just to set the record straight, I'm a big proponent of map sharing. It's one of the advantages of using a laptop programmable ECU. It's also one of the advantages of having a tuning community like the one we have here. It's a great way to exchange info, no question about it.
However, in this case, bad info can potentially blow up your motor. What makes matters worse, is that good info in one car may bad info in another car. So, great care must be taken in order to protect yourself. And this can only come through knowledge and experience gained from a good deal of first-hand tuning, as frustrating as it can be. Sometimes this will take days. Sometimes weeks. Sometimes months. But without this experience, trying to use and evaluate someone else’s map is nearly impossible (and quite unsafe)—especially when introducing boost into the equation. That's why I'm encouraging everyone to get along on the learning curve by starting off on the baseline file provided, as nasty and rough as it may be. Chances are, that only one or two parameters need to be tweaked to get the car to drive reasonably well. This, by the way, is mostly the byproduct of running big 550cc/min injectors and having to deal with not-so-insignificant engine-to-engine variance. Once that is taken care of, the fun (or painful, depends on how you look at it) part starts. And don't kid yourself, it will take some time to get to the point where you not only understand the concepts (that's the easy part) but can become proficient in seeing cause/effect relationships. The TEC's sheer flexibility will also mean that there is more than one way to get from A to B.
But once that point is reached, map sharing is a vital tool in getting to the next step in which you see how other people did things differently, yet still managed to reach the goal. Along the way, you'll probably catch and correct things that, gone unchecked, could have done nasty things to your particular motor. But a few keystrokes and it's gone.
Source: R Diamond
Source: Shiv
Adjust Time On Gamma (TOG) and Injector Offset (IOT) until the corrections are close to zero at both idle (influenced by Injector Offset Time) and 5500rpm (influenced by Time On Gamma). Both TOG and IOT will influence each other as well, so a little poking around is necessary. Just remember to adjust (either positive or negative) IOT by small amounts (.125ms at a time). What you are doing with your TOG and IOT adjustments is dialing-in a basic raw fuel curve. The VE table adjustments will "massage" the raw fuel curve to properly match your engine's breathing characteristics at places other than wide-open-throttle 5500rpm and idle (at these two places, your VE table should be zero, obviously). My baseline map has a TOG and IOT value which should be pretty close to what you will end up with. But, as always, engines will differ.
Source: Jkav
Verify that the target AFRs at full load are acceptable, or err on the
rich side. Get the car running with the baseline TOG and IOT values you have,
let it reach operating temperature -- don't worry yet if the idle is wonky --
and go out do a WOT 3rd gear run to redline. Have someone watch EGO CR during
the runs. If EGO CR goes positive and starts marching up toward the max of the
authority range you set (make sure its ~+/-.4 for your first run), back out and
adjust (increase) TOG. Repeat until there are no positive EGO CR values -- you
want them to be zero or slightly negative.
Then,
after TOG is dialed in, adjust IOT in .125ms bites until you get -.02 to -.05
CR at fully warmed idle. The small negative correction at idle will make tuning
hot start much easier since EGO feedback is turned off for 20 seconds after a
hot start. The "extra" enrichment will help get a smoother hot start.
Source: NickCat
The one main thing I can say that will make values vary wildly is whether you are on the gas or off and whether your fuel cut has triggered. If your fuel cut has triggered, you are going to be running super lean, and the numbers will reflect that. Conversely, if you are cold starting, and gathering those numbers that will be reflected as probably running rich.
Some simple rules for data gathering:
1. Try to always make sure you wait till the engine is warmed up before gathering data.
2. Keep track of outside temp. This will proportionally affect the numbers you see.
3. Just a note: Cruise is one of the hardest things to tune. You will always get variations in corrections; it’s just best to get as many as you can (while on throttle, never off).
To make sure you are getting values that are on throttle make sure you are filtering when the TPS is over .50 that way you pretty much get rid of all the useless off-throttle data. When I was playing with access databases to group the data. I was using the following filters:
CLT > 80
TPS > .50
This will get rid of most of the bad data. You have to tweak the rest to get what you want.
Source: Jed
Source: Shiv
I'm sure I don't have to say this, but I would advise not running your turbo car on the dyno until you know everything is all sorted and safely tuned. Dyno's can be pretty nasty to cars that aren't running well. Drivability should be the first goal. After that, dyno tuning should only involve VE table and ignition table manipulation. Do it the other way around and things get confusing and messy.
Source: Shiv
The adjustable parameters in the engine monitor screen are best exploited on either an engine dyno or a steady state chassis dyno. In the real tuning world, the ability to hold an engine under a fixed load and a fixed rpm is highly valued. I believe Dynojets became popular because most people didn't have the ability to do rpm/load breakpoint tuning.
But even on an internal dyno, one can still tune sector by sector, from say 2-4krpm, then 3-5krpm, then 4-6rkrpm, and finally, from 5-7krpm. And during each of these testing zones, finding the sweet spot of ignition advance is done by adding and subtracting timing until you reach max torque. Same applies with fuel. This can be done in two ways. One, by just bumping up or down the desired AFR target (offset) in closed loop mode. Or two, by adding/subtracting GAMMA in open loop mode. On high-powered cars, the latter may be more helpful since o2 feedback can take too long to zero in on the desired AFR.
The truth is, the inherent linearities of any given motor coupled with the operational linearity of the TEC system makes dyno tuning quite easy. The best place to start is by finding the engine's critical rpm break points. This is basically the points at which the torque curve changes in slope. Where the torque curve is flat, the ignition and fuel requirements will be essentially flat. Where torque is high, ignition will be relatively retarded and fuel delivery will be relatively generous. The opposite applies to areas of low torque (at high rpm).
Source: Shiv
There are many ways to approach dyno tuning. Perhaps the simplest way is to first take a stab at fueling, then timing, then fueling again.
I would suggest running open loop during testing. Through the engine monitor screen add/subtract 2% fuel at a time and see where it helps by looking at the torque curve. In those places were torque improves, add/subtract that into the associated VE table cell. After a few runs, you should have a VE table tuned for max torque at all rpms given the timing map.
Now, do the same with timing. Small increments. See where the changes help and where they hurt and incorporate them into the ignition advance table. At lower engine speeds, more advance will have the most effect. At higher speeds, advance gets less critical. Tune right up to the threshold of knock and dial it back an extra 1-2 degrees. At the end, you should have a nicely shaped timing curve that looks somewhat like the inverse of your torque curve. That is, where torque is high, timing should be lower. Where torque is low, timing can be higher.
Now, go back and add 2% fuel across the entire range. See what happens. In those areas that don't lose power, keep the extra fuel in there as added safety margin. By now, you should have good fuel/ignition curves. Just make sure that EGTs stay under 1600F (as measured close to the exhaust valve in the header runner). Your EGO voltage will probably range from 4.4 to 4.7v at WOT.
Source: Sean McElderry
I'm very happy with my off-boost drivability. While tuning my MY00 Minnam Stage II, I found that I needed rpm break points at the following: 900, 3500, 3850, 4250, 4650, 5200, 5750, and 6250.
Those are the key areas where the volumetric efficiency of my engine changed. By using those points in my map, I now have dead even 8psi of boost from 2900rpm to redline in the cold and 3100rpm to redline in the heat.
My MAP values are 30, 60, 100, 120, 150, 170, 187, and 200. They seem to work fine for me.
Source: Shiv
In dyno tuning my 2.5RS turbo, I settled upon what I believe to be good RPM breakpoints. That is, points in which engine VE makes some swings. Of course, every car is going to be different. Especially those running different turbo configurations, different boost levels, and exhaust/intake piping. NA cars are going to be very different as well. So, consider this just a starting point.
Here are some critical RPM points:
3650 Seems to be a point at which the intake manifold is tuned for. Lots of torque to be had here. Needs a lot of fuel and a bit less advance here.
4100 Seems to be a zone of poor VE. Needs less fuel and more advance.
4650 Good VE again. Lots torque here. Peak torque, in fact. Lots of fuel, a bit more ignition retard.
The torque curve is pretty linear from there to redline. Not flat, but linear. Almost a straight line to redline. But then again, I'm tuned to run 7psi up to 5000rpm, ramping up to 9psi by 6000rpm. So, as always, your mileage may vary.
Source: Sean McElderry
4th and 5th gear are pretty essential for tuning, in my opinion. I find that in the lower gears, things happen too quickly. The EGTs are optimistic in lower gears as well. Stay up really late, use the interstate to tune, and pray the whole time that there are no cops around.
Datalog Settings—Speed: 3, Length: 4. I do one 4th gear run from 2000 to 6500rpm, view the results, make changes, and repeat. Sometimes I will do more than one 4th gear run to see the affects of a changing MAT or something of that nature. It's also important to datalog normal driving conditions as well—but when tuning for WOT, I do the above procedure.
Source: jhuang76
On the highway, I can usually not speed too much and get from 3000-6000 in 4th. Tuning in 5th seems to be unnecessary. I'll have my VE maps set so I'm at around -4% EGO correction WOT at almost all RPM in 4th. Then, I'll run WOT in 5th at obviously limited RPM (don't want to get nabbed for 40+ over limit), but for those RPM values, I'm still around -4% EGO correction. So, I assume through extrapolation that my higher RPM in 5th should still be fine.
In short, I would do aggressive and active tuning in 3rd on a side road. I'll verify the accuracy less often when I'm driving on the highway in 4th and 5th. It's worked so far.
Source: Shiv
For a normally aspirated car, go ahead and do your tuning in 3rd gear. N/A cars don't accelerate that hard.
Source: Shiv
My suggestion is to tune for the coldest conditions were
the engine would place most demands on the fuel system. Keep an eye on injector
duty cycle, as you cold weather folks will suck down more fuel, per psi of
boost, than us fair weather fiends.
Source: Shiv
I'm afraid that the numbers do not re-align themselves with changes in breakpoints. You will have to recreate the curve.
Source: Shiv
Here's the fun part. Richer mixtures will reduce EGT. How much depends on a number of other variables -- the most significant one being ignition advance. However, after a certain point, additional fuel will actually raise EGTs. Lots of people don't recognize this relationship and end up over fueling their system.
This is because excessive fueling will slow down the burn to the point where it is still burning on its way out of the exhaust ports. As you can imagine, this will make your EGT probe very hot. I don't like to think in "AFR" number because they are flaky and dependant on other variables. And rarely are they really indicative of the actual air/fuel ratio entering the engine. But it's not unusual for a 5% change in fueling (+- 5 in the VE table) to alter EGTs by as much as 100F.
Source: Shiv
Well, it's really going to boil down to a couple of things. First, the catalytic converter on the Minnam turbo kit isn't the best. It's been known to harden and crumble with time. If that happens, it's best to hollow it out and install a new cat somewhere downstream in the intermediate pipe. The reason I say not to put it where Minnam put their cat is because there isn't enough room to put a properly sized cat of sufficient flow capabilities in there. That's why the Minnam cat fails with use. Size it big and wide and you'll be okay in that respect.
As for tuning, I think as long as you have the car operating in good closed loop fuel control (back and forth lean/rich at a good pace), you'll be fine as far as fueling goes.
As for timing, more ignition advance will raise your NOx level. So, you'll probably find that more retard is necessary for passing. You might have to play around with this on the sniffer itself. This is why it's best to find a "friendly" Smog station during your first sniff test. He should allow you to tweak until your car runs lean. Then save that program on your laptop for the next time.
For what it’s worth, my Smog tester in NJ wasn't very friendly. He wouldn't let me make changes to the TEC-II at the station. Instead, he told me what was failing and let me drive around the block, tweak it, and drive back up and sniff again. Okay, so he was a little friendly
Source: TEC Manual
SE0 Temperature based, 1-second starting enrichment -- This extra Gamma is added to the current Gamma for one full second after the start of engine cranking. Its full value is added in if the coolant temperature is -30C and then drops linearly to zero at 80C. Try 1.5 GAMMA.
SE1 Constant, 1-second starting enrichment -- This enrichment adds to GAMMA for a duration of one second after cranking starts, regardless of coolant temperature. Try .2 GAMMA.
PW0 Fixed, 1-second starting pulsewidth -- This enrichment adds a fixed amount of additional pulse width to the injector firing regardless of MAP or kPa values. It is used primarily for turbo charged engines, which do not show much relative pressure at cranking. It is added in for a duration of one second only if engine RPM is below 400 and the coolant temperature is less than CLT0. Try 3ms for turbo engines, 0 for normal engines.
CLT0 Coolant temperature below which PW0 activates -- This temperature is used in conjunction with PW0 to determine if the engine is in a cold cranking condition. Below this temperature the extra starting pulse will be injected. If the value is set to 80C, the starting pulse will always be on.
ASE0 Temperature based, 20-second starting enrichment -- This enrichment ramps down over 20 seconds from the start of cranking. The amount of fuel added depends on engine temperature. At -30C, the full, programmed GAMMA is added. There is a linear drop-off of GAMMA to a value of zero at 80C. Start with a value of 1.0 GAMMA.
ASE1 Constant, 20-second starting enrichment -- This enrichment ramps down from the programmed GAMMA at the start of cranking to zero GAMMA 20 seconds later. It is always used when the engine is cranking. Try .1 GAMMA.
Source: Sean McElderry
The "correct" amount of ignition advance is based on more than CR, A/F ratio, and Octane Rating. Volumetric efficiency and intra-cylinder pressures also affect how the mixture burns, and thus have a direct correlation to the "correct" advance value for a particular operating point.
The reason advance vs. RPM is not linear is because the engine has different volumetric efficiencies across the revband.
The reason advance drops as load increases is because you have a lot more air in the combustion chamber. This leads to higher cylinder pressures, which makes the mixture easier to ignite; consequently, the burn both starts and completes faster, which requires less advance.
Source: Shiv
Most (including myself) would argue that tuning for knock by ear is still the best way to go. While some people wear detonation cans, there is a good deal of extraneous noise than can confuse the issue. Still, if so desired, one can also connect the TEC-II's knock sensor to a set of headphones and listen. But the ears are great listening tools all by themselves. Just keep the windows up and the radio off during the tuning process. Once tuned and knock-free (with a nice 3-4 degree safety margin), you can set-up and activate the knock sensor for back up.
The key to ignition tuning is to take small steps at a time. If you advance by increments of 2 degrees, you'll pick up mild knock before it gets serious. Developing good knock detection skills is mandatory for any engine tuner—especially when turbos and superchargers come into the picture!
When doing ignition tuning, remember to disable the knock sensor.
Otherwise, it will be impossible to dial-in appropriate ignition curves.
What is MBT?
Source: Shiv
MBT = Minimum ignition advance for Best Torque
It's a good indication of how efficiently your system is running. If you're running at MBT, you're not going to make more torque through additional ignition advance. If you're running below MBT, you want to run more advance. Sometimes, octane requirements (in higher boost applications) makes it impossible run at MBT. This is why 105+ octane race gas makes it possible to extract loads of hp in high boost applications since MBT can be reached without detonation.
Source: Shiv
Rumor has it that you want maximum expansion approximately 14 degrees after top dead center for maximum torque. This means that the burn needs to begin even earlier, which means that the spark plugs need to be given the signal to spark well before top dead center (this is why it is referred to as "advance"). How early depends on the speed of the burn which, in turn, depends on engine load, fuel octane (which is usually the limiting factor), air/fuel ratio, engine speed and probably a few dozen other things.
For a 7-8psi turbo EJ25 running on pump gas, you will never be able to run your car at "optimal" advance since octane will be the limiting factor. Instead, you will probably want to operate 3-4 degrees away from the onset of audible knock. This will probably be in the 15 to 20 degree BTDC area. More boost will require even less advance. There will come a time when too much boost will require too much retard which will cause EGTs to rise to dangerously high levels. This is why it's important to have an EGT gauge. 'Cuz knock isn't the only thing that can hurt the engine.
Source: JKav
I think Shiv meant to say "More boost will allow even less advance." The peak-expansion at-14-degrees-ATDC thing still holds regardless of boost level. And the upward-spiraling EGTs can also be detrimental to the precious, precious sacred turbo.
Source: Shiv
What is relationship between torque and ignition advance? Think of a curve. At first, the slope is very positive, then less positive, then zero, then a little negative, and then very negative. If you catch the curve early enough, then yes, 2 degrees can make a big difference. But once you’re at maximum brake torque (ideal state of tune for a ping-resistant NA car), a couple of degrees here and there may not be noticeable.
Source: Efoo
The knock sensor detects knock it will retard as necessary. You can tell it how many degrees to retard by per knock incident, and how fast to bring the timing back if knock is not further detected.
Source: Shiv
The knock sensor, fortunately, does not make permanent changes in your ignition map. Once the timing is pulled back, it is added back in user-defined increments until it reaches the proper, map-specified ignition value. The only downside of this is that if you will have to revise your ignition map manually if knock occurs. But knock sensors aren't perfect. They will pick up noise that isn't necessarily knock. They are also incapable of catching every knock. This is why I suggest dialing the car in with the knock sensor disabled or with a very high sensitivity. Your ears are the best tool for knock detection.
Source: Shiv
I set it up so that it's very aggressive once it hears detonation. However, the activation threshold is high enough so that engine noise won’t trigger retard. I set max retard for 10 degrees with a rate of retard (when sound level is above threshold) of 5 deg per engine event and a rate of advance (when below threshold) of .25 deg per engine event. What this means is, instead of being invisible to the user, you will actually feel a sudden reduction in power once the knock sensor jumps in. When/if this happens, you need to back off ignition advance at that particular rpm/load point. This way, you are using the knock sensor as a fail-safe and tuning aid—not as a form of active, full-time ignition tailoring. The latter resulting in less overall advance/less power/higher EGTs due to the characteristics of knock hysteresis—once knock starts, it requires a lot of retard to stop.
I should point out that different cars will
end up with different knock sensor parameter numbers. Best to trust your ears
and makes sure that you and the knock sensor sees eye to eye. You can test its
functionality by advancing timing a few degrees (in the engine monitor screen)
to induce knock and then see whether the knock sensor jumps in to the rescue.
One or two pings may sneak by but, after that, the ignition timing should pull
back quickly. Needless to say, it's a good idea to tune the knock sensor before
the car is turbo'd and running high boost. Hope that helps. For those who are
still in the fine-tuning stage, I would suggest that activating and tuning the
knock sensor is the last thing you need to do. Trying to construct an
ignition curve with it jumping in and out all the time is nearly impossible.
Source: Shiv
The knock sensor microphone is tuned to pick up certain frequencies that
are typically characteristic of knock. However, many engines (especially built
ones) are noisy at certain engine speeds. In most every application, the knock
sensor should be disabled above 5000rpm due to excessive valvetrain noise.
While I've found the EJ25 to be unusually quiet in this respect, making life a
bit easier on the knock sensor, a built motor may be way too klanky. I would
suggest advancing your timing until you hear the onset of detonation with your
ears. As soon as you hear it, of course, get off the throttle. As long as you
don't stay there and groove to its beat, the engine will be fine. This is important since running too little
advance is nearly as bad as
running too much of it. Too little advance will also make your car feel very
sluggish with a tendency to run rich. You'd be surprised how much a difference
a few degrees make (as long as you don't ping!)
Source: Shiv
The 4500 rpm recommendation really only applies to klanky race engines with solid lifters. I've found that keeping the knock sensor activated at all times (set the inhibition rpm to, say 7000rpm) works just fine for the relatively quiet and smooth EJ25. The key is to set your threshold high enough so that only knock (which produces huge sound spikes) will trip the retard function (I use 45 as a threshold). Also, I would recommend setting the rate of retard very aggressive (I use 5 deg/engine event) and the rate of advance very slow (I use .25 deg/engine event). This way, you'll actually feel the TEC pull back timing when knock occurs. This will encourage you to do something about it before continuing on your flogging ways
It's also a good idea to only allow 7-10 degrees if total retard authority. Be careful that the knock sensor isn’t knock retarding when knock is not present. In other words, 99 percent of the time, the knock sensor should be completely inactive. You can verify this by data logging and looking at the total advance numbers. Hope that helps.
Just remember that too little advance can be almost as bad as too much of it—especially in a turbo application where you're right at the edge of safe EGT readings.
Source: Sean McElderry
Spikes to 35 are normal, background noise, in my experience. I set the sensitivity of my knock sensor to filter out anything below 45. I think I'm going to change that to 40 though. When my engine knocks at RPMs below 4500, I will get a huge spike in my datalog. Above 4500, there is so much engine noise that it isn't picked up all the time. So it's very important to listen for knock above 4500, since the datalogs will essentially be lying to you at that point.
Source: Joe Hogan
According to Shiv, when his engine pings the knock sensor spikes to 100.
Source: Shiv
I've also set up some pretty conservative baseline fuel maps. It's set up so that O2 sensor feedback is always one and always corrects to the negative side under boost (rich base fuel map). What this means is that, with the help of acceleration enrichments, the car initially runs a little rich under boost, leaning out slightly as the car stays on boost, until a/f targets are reached. This approach seems to allow us to get away with a rather aggressive ignition curve without running into detonation problems. It should also give the user some safety margin when fine-tuning the system. Once again, the high-speed data logging features (directly downloaded to the laptop's hard drive) made analysis very simple.
Source: Shiv
The best way to go about the leaning out process is through an EGT gauge. Also, keep an eye on the EGO voltage more so than the nominal AFR numbers (since different people are running different o2 sensors with different calibrations.) Don't go any lower than 4.5 volts under boost. You will probably end up somewhere between 4.6 and 4.7v.
Source: efoo
I'd just take it easy and lower it one small step at a time, listening for ping along the way. I did that with my car and am running AFRs of 14.35/14.4 under boost now -- my exhaust definitely has a gas smell to it, but it doesn't make me sick any more.
Source: Shiv
Don't forget that when it comes to fuel tuning, changes in the AFR table can be just as effective as changes in the VE table. This is especially turn when running big injectors (like almost all you guys are) and minimum on-times that make it impossible to run lean AFRs under heavy vacuum/engine braking.
For example, with a minimum on-time of 1ms, you may find that the leanest the car will run at 30kPa, between 2000 to 4500rpm is 14.3:1. Note: This condition will only occur if your decel fuel cut is not triggered which may happend when slowly getting off the gas or not lifting off the throttle fully). If the AFR targets in those zones are 14.5:1, your EGO correction will stupidly pin itself to the negative side, trying in vain to get AFRs lean. And when its pinned at a big negative correction, the engine will lean stall once it comes to an idle. And if you don't come to a stop, you will feel a big lean bog if you get back on the gas. Yucky.
Upon first glance, one would be tempted to keep taking fuel out of the VE table in those aforementioned zones. However, this approch will not work because the minimum on-time of 1ms is the controlling, supreme force Instead, set your AFR to, say, 14.25. This will ensure positive EGO correction which has no negative side effects under these conditions.
A good way to set your low load AFRs is to do a datalog of a 5000rpm-to-idle coastdown and set your low load AFR targets to be as rich (or slightly richer) than the AFRs you see in your log. When doing this datalog, you will need to:
Source: Shiv
During light loads (such as cruise), running AFRs of 15-16:1 is fine. In fact, it's desirable as it improves fuel economy. As load goes up, however, more fuel is needed for max power and cooling effects.
Just to avoid any potential confusion, the nominal value of 14.1:1 from a TEC-II is roughly equivalent to an o2 sensor voltage of .93. Due to the nonlinearity between air/fuel number and o2 voltage, it may be a better idea to simply look at o2 voltage. Divide it by 5 to see the raw output. 4.3/5=0.86 volts. Rule of thumb would be to aim for .86-.88v for an NA car and .95-.96v for a turbo car.
Also keep in mind that o2 sensor voltage will be affected by RPM, EGT, ignition advance and a number of other variables. Also, don't forget that not all o2 sensors are calibrated equally. In other words, there is no "ideal" o2 sensor voltage for power. Even off much-hyped wide band o2 sensor. They'll get you 80% of the way there. Use EGT and open ears to get you the rest of the way.
Source: Shiv
At idle, almost every engine will need to operate richer than stoich (which is approximately 14.7:1). This is mainly due to unavoidable engine efficiencies. At those low airflow rates, intake swirl and tumble is pretty weak. This means that the air and fuel mixture isn't properly mixed/homogenized. Because of this, it's easy for the spark plug to misfire as it may not come in contact with a good enough localized air/fuel cocktail during the spark event. Adding more fuel to the mixture ensures that the cylinder is full of the right cocktail, even during low airflow situations.
Whereas under cruise, at higher engine speeds, the intake swirl and tumble is more than enough to ensure burnable in-cylinder air/fuel homogenization. Hence, the ability to run relatively lean.
One should keep in mind, however, that running excessively rich at idle isn't good for the catalytic converter, which may clog and crumble over time. Nor is it good for in-cylinder lubrication as the excess fuel can wash away the oil the cylinder walls, eventually leading to metal on metal contact. What is too rich? It really depends. But if you're idling with o2 sensor voltages of a steady 4.7v and up, you should try to lean things out a bit. With big injectors and 45-50psi rail pressures, I've been able to bring on times down to 1.1-1.2ms. This usually results with a nice closed loop rich-lean-rich-lean scanning at idle (even when the target is a relatively rich 14.3-14.4:1).
Source: Shiv
If your AFR targets too rich for N/A, you are literally throwing away torque. I suggest backing off the a/f targets in the ~100kPa (WOT NA) range. You can keep the higher MAP cells with the same a/f ratio to ensure proper a/f under boost. I would even do this when you are turbocharged. No use running extra rich when you're not on boost.
(Source: phone conversation, Shiv’s method for determining the best N/A AFR target): Using the engine monitor, apply a global AFR offset until you see about 4.4v on the EGO sensor at 100kPa. Note the “desired AFR” in the monitor screen that gets you 4.4v, and put that number into your AFR table. You may wish to retard ignition timing a few degrees globally while doing this tuning, just to be safe.
Source: Shiv
That's normal. Under light load, the air/fuel mixture will oscillate between rich and lean. This is typical of closed loop feedback. Just try to make it center around 0 correction to ensure that the targets can be reached under all conceivable atmospheric conditions. Under load, however, the A/F ratio should stabilize and go rich.
Source: Joe Hogan
The question: The EGO
corrections in second gear were between -5 and sometimes as high as -10 at the
higher rpms. However, in third gear the
corrections were 0 to -2 from 4k RPM, (second to third shift), to 6k RPM.
Literally a flat line. While this made me feel good, why the pronounced
difference between 2nd and 3rd gears?
Source: Sean McElderry
I see the same thing, except I've only done a comparison between 3rd, 4th, and 5th gears. In 3rd gear, my EGO corrections will go to -10% to -15% at WOT/6500rpm. In 4th and 5th gear, they stay at 0%.
My only thought on this was that things happen much quicker in 3rd gear (redline comes really quickly) and the TEC-II doesn't have time to catch up because the EGO corrections are setup pretty slow (to get a stable idle). Does that make sense?
Source: Shiv
You will always see bigger negative EGO corrections in lower gears. This is due to a decrease in engine VE due the engine speed's quicker rate of rise. Your engine will be most volumetrically efficient (i.e. hungry for fuel) in steady state conditions where RPM is constant. This is the reason I suggest tuning in 4th gear, as it's better to have too much fuel in lower gears than to have too little fuel in higher gears. This is also the reason EGO feedback remains on at all times.
Yes, it does take some time for the TEC-II to zero-in on the desired a/f ratio. This means that it may take a second or so where it is playing "catch up". This is most evident in the lowest gears. This is also why it's best to tune for negative corrections. Extra fuel has very little effect on hp. Too little, on the other hand, can be dangerous.
Source: Shiv
During an overrun, the injectors turn off and only come back on when: a) TPS goes above the "closed TPS voltage" value, or b) engine speeds drop to 1600rpm (or whatever the injector turn-on engine speed value is in the decel parameter window). When the injectors turn off, GAMMA will go to 50 and EGO correction will zero out. EGO correction will also go to zero when enrichment is triggered.
Source: JKav
You should have it set up to give about 2-3% "downward" CR everywhere (taking out fuel) except idle. If you're seeing huge CRs, it could be a couple of things: too-rich a/f targets causing rich misfires, the O2 sensor falsely reads lean and adds more fuel, etc. Try leaning your on-boost a/f targets slightly, say to 13.8 or so, and do a data log. Lay into the throttle gently and back out if you hear knock. Input the CRs you're seeing into the VE table at the respective breakpoints. Theoretically you can disconnect the O2 sensor and run the car off the VE table and have no effect whatsoever on performance. But it’s wise to use O2 feedback for minor adjustments to the fuel delivery. Again, you really would like to see CRs of -0.02 to -0.03 under boost.
Source: Shiv
Not only are they normal, they are also intentional. They are there because I forced the TEC to try to achieve a rich a/f target during engine braking (less than 30kPa). I would rather the positive corrections (as you have seen) than negative ones. The latter can cause a stall or lean run bog when the engine falls to idle.
If you look at your a/f ratio table, you'll see that the lowest MAP row has very rich a/f ratio targets. This is where the positive corrections come from-- from the TEC trying to enrichen in order to achieve those rich targets. Usually, the EGO correction will just pin itself to the positive end of its authority. A little trick I learned when tuning small engines running big injectors with high fuel pressures
Source: Shiv
With well tuned TOG and IOT values, the TEC will ensure that you're always pretty close to where you want to be. The only thing that can mess things up is EGO feedback, which is why I encourage all to disable it at wide-open throttle (once one gets comfortable with tuning). For novices, the worst thing leaving EGO feedback enabled will do is result in a richer-than-ideal a/f mixture as it always fails on the lean side -- causing the TEC to add in fuel.
Another good example of why O2 sensors make poor tuning tools when asked to operate away from stoich. If you look at the datalogs (or at the EGO calibration settings in your calibration file), you'll see that the nominal AFR value of 14.36:1 corresponds to a wide range of EGO voltages. So it becomes possible for the TEC to change fueling quite dramatically yet still achieve the desired nominal AFR. One can fix this by recalibrating the relationship between AFR and EGO voltage (in the EGO Calibration Screen) but you're still left with a flaky sensor. Best to ignore it all together when tuning at WOT. Just disable EGO feedback above 85kPa and you'll be much happier.
Source: Shiv
We do use O2 sensors to tune our TOG and IOT. However, it's important to know their limitations. In other words, it's unrealistic to expect repeatable EGO CR numbers since the sensor's performance isn't very repeatable. Fortunately for us, when the sensor misreads, it misreads on the lean side. This means that the TEC will just add more fuel and more fuel never hurt anyone. This method of tuning is sufficient (and inherently safe for the engine) when it comes to dialing in the raw fuel curve. Fine-tuning for maximum power (especially for boosted applications) is best done with an EGT gauge, which provides a much clearer window into how the engine is operating.
Source: Shiv
The reason why o2 sensors are used for tuning is because they are inexpensive and standard on any modern automobile. While we are using them just like everyone else, all of us should be aware of their limitations when it comes to tuning.
O2 sensors are sufficient for gross tuning as they will always lead you to a "safe" (rich) outcome. As a result, they are just fine for filling out a VE table. In fact, the next generation software will use the O2 sensor to self-calibrate the VE table. But when it comes to fine tuning for maximum power, you need a dyno (or a really sensitive butt). Boosted guys will also need an EGT gauge, as max power is not good if it is accompanied with 1700F exhaust temps.
Source: Shiv
In the EGO Parameters, which is in the A/F Ratio Table screen. You'll see both a MAP and RPM switch. Both need to be satisfied in order to disable EGO feedback/closed-loop. In the baseline, the MAP and RPM switch are set at 208kPa and 7000rpm, respectively. Basically ensuring that EGO feedback will never be disabled.
Source: Shiv
I typically run with a 10 authority range, with EGO disabled above 110kPa and 3000rpm. Both conditions need to be met before EGO feedback shuts off. When I was NA, I ran closed loop at all times with good results. However, I wish I had experimented more with running open loop at wide-open throttle. I think it would have resulted in more predictable fueling (at least in my case).
Source: TimZ
I use a 5% authority range, but this requires the VE map to be dialed in pretty well. It's good to remember that you can't rely on the EGO feedback to correct for anything but steady state cruise conditions. It does not respond quickly enough to correct for poor mapping under wide-open throttle conditions.
Source: Shiv
Changes in engine VE imply that it takes different amounts of fuel to achieve the same air-fuel ratio. For example, if you set all the a/f targets to 14.6 (not advisable, of course!), in order to minimize O2 feedback correction, your VE table will have to be filled out. You'll see that you will require more fuel in the midrange, when torque is the highest. And you’ll need less fuel at both ends of the RPM band when engine breathing suffers and torque is low.
In the real world, your motor will want to run at different a/f ratios in different conditions. Under boost, you'll set your a/f ratios lower (richer). So, in effect, your VE table not only represents your engine's breathing characteristic, but is also influenced by the a/f ratio targets your engine wants to run at for best torque. Clear as mud, no? Let me know if this doesn't make sense.
As I've mentioned a few times before, for higher boost applications, where the engine wants to run rich, the narrow band o2 sensor has limited functionality. In these cases, I would advise more experienced tuners to run open loop above, say 130kPa. This will require either a dyno and/or a good EGT gauge for proper, safe tuning. For less aggressive tuning, nothing is wrong with staying closed loop at all times. At worst, it will just result in a bit extra fueling. Running a little too rich never hurt anyone.
Keep in mind that this suggestion is only for boosted applications. In NA applications, the engine doesn't need to run very rich. This means that the standard o2 sensor is capable of providing accurate information.
Source: Shiv
VE mapping doesn’t change to any significant degree with ignition advance manipulation. Under some circumstances, you will notice that more advance does make the engine run slightly leaner due to a more complete burn (since it starts earlier in the combustion cycle). But I've only really seen this relationship at full load conditions. But the difference is subtle (but worth looking it, of course).
Under cruise, you should be able to target afrs leaner than 14.64:1. Try going leaner and leaner until you feel oscillations caused by going slight misfire and then richen it up a bit—just enough to smooth it out. I ended up at 15:1, for what it’s worth.
Source: Shiv
From what I've seen, engines will always have poor VE under cruise conditions. So it's not unusual to see big negative VE corrections below 80kPa. Looking at Imprezer's NA car (with intake, exhaust and headers), his 105kPa VE cells are only 5-10% greater than that of cruise.
Source: Shiv
As long as you aren't getting bogs of any type off idle,
don't worry about big corrections at idle. It's natural and is bound to happen
when you are using big fat injectors on a relatively small motor at just low
loads. It's only under load that corrections need to be managed for maximum
performance.
Source: Shiv
The big injectors make is impossible to achieve a lean 15:1 AFR at the super low MAP values seen during engine braking at higher rpms. The injector would stall if it weren’t for the minimum injector on-time keeping them open. Instead of having huge negative VE corrections, you may want to try making your AFR targets in those cells richer. In my 500cc/min injector'd 1.8L Miata (!), I have the AFR set to 14.3:1 across the entire board at 35kPa. This eliminates the big brainless negative corrections I would get otherwise. My injector cut off is triggered to happen at MAP values less than 30kPa. Don't set this too high or your car will get the herky jerkies. Just make sure that your closed TPS voltage is set accurately. If it is too low, the TEC will never think that you are fully off the throttle. As a result, it will never cut fuel.
Source: TEC Manual
Manifold Absolute Pressure Activation
DCCL3 MAP reading below which DECEL activates -- This is the manifold pressure below which the engine is in the over-run (excess vacuum) condition. It should be set by observing the manifold pressure at idle, neutral and subtracting 8kPa x BAR. Recheck the number by causing the engine to go into an over-run condition, such as in coast down and observe if the manifold vacuum drops below this DCCL3 number.
RPM Limits of Activation
DCCL0 Cold RPM above which DECEL activates -- This is the lower RPM limit at which the deceleration fuel cut-off will keep the injectors off if the coolant is -30C. Above -30C, a straight-line linearization to DCCL1 is used to determine this RPM. Once the RPM drops below this point, the fuel injectors turn on and DCCL4 enrichment is added for one second.
DCCL1 Hot RPM above which DECEL activates -- This is the lower RPM limit at which the deceleration fuel cut-off will keep the injectors off if the coolant is at or above 80C. Below 80C, a straight-line linearization to DCCL0 is used to determine this RPM. Once the RPM drops below this point, the fuel injectors turn on and DCCL4 enrichment is added for one second.
Throttle Position Rate of Change Sensitivity
DCCL2 TPS rate above which DECEL activates -- This is the rate the throttle position must be closing to activate the fuel cut-off. A higher number is less sensitive; a lower number is more sensitive. Try 7 volts/sec.
Post-deceleration Enrichment
DCCL4 One second enrichment after DECEL ends -- This enrichment is added to the current GAMMA for 1 second when the fuel injectors come back on. It is intended to help refill the manifold and wet the dry cylinder walls. Try .1 GAMMA.
Source: Sean McElderry
In my car, there is a huge difference when fuel cut doesn't occur. Bucking occurs because the injectors are trying to fire just a little bit. The car "wants" to decel, but the injectors are preventing it from happening. When fuel cut occurs, it's a nice, smooth decel all the way down to below 2000rpm.
Source: Efoo
My settings are rather aggressive, but they work for a very gentle cut that occurs often.
DCCL0 = 1500
DCLL1 = 1900 (yeah, I'll explain in a sec)
DCCL2 = 0.75
DCCL3 = 35
DCCL4 = 5
I actually spoke with Electromotive on the phone regarding these settings (these are from them). Essentially, my car has a problem in that revs drop like a rock once I lift off the throttle. I have a lightened flywheel and I think that is why. In everyday driving the revs drop so fast that the decel doesn't turn the injectors back on until it's too late and the engine quits (like when I'm coasting to a stoplight and am out of gear). So I turn the injectors on really early in order to make the drop not so drastic. Otherwise, the DECEL kicks in almost immediately on the highway though. Very gentle. I might change DCCL4 down to 2% after looking at your map though; I think that might account for part of my part-throttle bucking problem.
Source: Shiv
Make sure that your "TPS at closed throttle" value is dialed in correctly (see TPS Parameters). If it is set too low, the TEC will never know that the throttle is closed and will not shut off the injectors. Also, make sure that your decel parameters are set sensitive enough (both MAP and TPS trigger). For starters, set the MAP value to 55 kPa the TPS trigger to 2 volts/sec. This will make the injectors very quick to turn off. Then, try making them gradually less sensitive until it feels right. Anyone else have any advice or some numbers that they are happy with?
Source: efoo
I've found that my current settings (3 V/s and 50 kPa) make my car feel like I've triggered a VTEC "off" switch every time I let off the gas. The car just suddenly starts slowly so fast. The kPa value subjectively feels like it triggers what throttle "position" the decel will trigger – the higher it is, the less you have to back off to trigger it.
The Volts/sec value modifies how fast you're laying off the throttle. If it's high, then you've got to basically just quickly take your foot off the gas in order to trigger decel. If it's low, the slightest upward movement of the pedal will trigger it.
For that stock
feel where you can smoothly get on and off the gas, I’m planning on raising the
volts/sec value pretty high, like 4 Volts/sec, and keeping the kPa value the
same. I want the decel to trigger quickly so my car doesn't buck when I do
quickly lift my foot (higher kPa), but I don't want it to trigger every time I
lift the throttle (higher volts/sec).
Source: Shiv
In an effort to make my dog box'd 2.5RS turbo feel a little more "normal", I took the time to carefully set up my deceleration parameters for maximum smoothness. For a while, it was a compromise between the injectors not shutting off easily enough during decel or shutting off too readily. The former would keep the car going when the throttle was closed. While the latter would keep the car herking and jerking incessantly. Neither was ideal.
So, what I did to smooth things out was to tune in a "semi-fuel cut" during decel. First, I would set the actual decel triggered fuel cut to be quite insensitive, with MAP trigger set to 40kPa (or right around the idle MAP) and with a TPS trigger of around 5 volts/second (instead of the super sensitive ~1 volt/second that many folks are using.
This tweak, alone, would cause the annoying problem of unintended... ahem... cruise control when the throttle was closed. Not anymore. I set the lowest MAP row (35kPa) of the VE and AFR tables to achieve a lean 17:1 a/f ratio. As always, the VE table and AFR table need to see eye-to-eye in order to avoid big EGO corrections that would result in bogs or soggy throttle response once the throttle was applied again. But once set up correctly, a slow lift off the throttle would induce a nice engine braking state without the big herky jerkies associated with a true fuel injector cut. But once the throttle was lifted off quickly, the injectors would completely shut off for maximum engine braking. The end result is really nice drivability. No more jerks around parking lots. And more importantly, much less tranny lash while driving around casually. I like it a lot. Give it a shot!
Source: TEC Manual
ACE0 Throttle position rate of change sensitivity -- Throttle Position Rate of Change sets the sensitivity at which the TEC will go into the acceleration mode. If the throttle is moving faster than this rate, as in a rapid depression of the pedal, the acceleration enrichment will start. A low number is very sensitive, 52.5 volts/sec. is off. Typical value = 7.
One Second Acceleration Enrichments
ACE2 Temperature based, one-second ACCEL enrichment -- When the throttle or MAP/MAF moves faster than the sensitivity, 1 second of extra GAMMA (fuel) will be added to the current GAMMA. This amount is fully added when the coolant is -30C and ramps down to nothing at or above 80C.
ACE3 Constant, one-second ACCEL enrichment -- When the throttle or MAP/MAF moves faster than the sensitivity, 1 second of extra GAMMA (fuel) will be added to the current GAMMA.
Variable Time Acceleration Enrichments
ACE4 Time for variable time ACCEL enrichments -- Several acceleration enrichments are provided which can be adjusted in length of time and amount. This value sets the time over which these acceleration enrichments will be activated. Typical value = .5 sec. Only choose .25, .5, .75, 1, 1.25, 1.5, 1.75 or 2.
ACE5 Decaying, variable time ACCEL enrichments -- This enrichment is activated for ACE4 number of seconds and its amount is a function of how fast the throttle was changed. If the throttle moved at 13 volts/sec the full amount of this enrichment would be added in. Typical value = 1 GAMMA.
ACE7 Constant, variable time ACCEL enrichments -- This enrichment adds to the current GAMMA when the flow of manifold pressure is changing rapidly. The amount of this enrichment is fully added for the ACE4 denoted period regardless of the rate of change. Typical value = 0.
Variable Time, Injector Pulsewidth Enrichments
ACE8 Fixed pulsewidth enrichment -- This enrichment adds a fixed amount of time to the injector pulsewidth when in the acceleration mode. Unlike GAMMA based enrichments, its amount is not varied by engine load or MAP/MAF. Typically used only on high boost turbo engines.
ACE9 Time to add the fixed pulsewidth enrichment -- This is the duration in time to add the fixed injector pulsewidth once the acceleration mode is detected. Typical value = .3.
Source: Shiv
The TPS voltage is dependant on battery voltage which, to a small degree, is dependant on how many accessories you have on. At night, you have lights on so you may lose a tiny bit of voltage. You should set the closed TPS parameter to the highest number.
Source: Bud Man
My decel setup probably is arguably all wrong, but after endless trial and error, it seems to work quite well. Fully closed TPS voltage is .43, the lowest I see. If I raise it, the injector cut-off is abrupt, and I hate "the jerk". I have very lean target A/F ratios in the 30 kPa row which has eliminated the surging typically noticed when the injectors don't turn off and are shooting for a richer target. I haven't noticed any problem with this approach related to flirting with the minimum injector on-time. The net result is that my car transitions to decel cut-off and back on more smoothly than the stock ECU did. And when the injectors don't shut off it's still very smooth with some really cool sounding background burbling.
Source: Shiv
I run a MY00 EJ25 at 10psi. With a 10:1CR, running that kind of boost is tricky. Running it with sufficient ignition advance (for good power and low EGTs) is even harder. During a full boost, wide-open throttle run from 4000rpm to 6000rpm, my ignition advance numbers range from 10 to 21 degrees. This keeps peak EGTs below 1500F. The high flow fuel rails help as well.
But one thing I was noticing is that during a dyno-style 2000 to redline wide-open throttle run, the car would not ping. However, if I were to lay into the throttle above 4000rpm, I would hear a few transient pings. I attempted to retard the timing. But the amount of retard necessary to make the ping go away was a bit excessive. Completely killing the power the car used to make during a wide-open throttle low rpm to high rpm pull. Another try was to enrich the partial boost cells in the VE table. But doing so made the car run to rich during the kind of sustained partial throttle/partial boost runs that one regularly does when tooling around the street.
However, since the ping only occurred during high rpm, vacuum-to-boost transients, it dawned on me that the car must be running too lean during fast boost rise. By too lean, I actually mean "not rich enough." Instead of increasing my TPS-triggered enrichment amount and duration, I focused on dialing in my MAP-triggered enrichment. In the end, I set ACE1 (trigger sensitivity) to a reasonably sensitive 0.5kPa/ms. ACE4 (enrichment duration) was set for a reasonably quick 0.25 sec since the lean-run seemed to be very short term. But ACE6 (enrichment amount) was set to a pretty aggressive 15%.
TPS triggered enrichment was left pretty skimpy because, with turbo lag, the change in engine load lags behind swings in throttle position. Trigger was set at 14v/sec. Duration was set to 0.25sec. Enrichment was set to 5%.
And what do you know... it worked. I could lay into the throttle super aggressively at high rpm without inducing even the slightest crackle. In fact, I found I could advance my high rpm timing a couple of degrees. While I can feel a very slight reduction in tip-in boost power response, the power delivery is smooth and still very urgent. The engine ends up being stronger and happier with this change.
Note: For supercharged cars, I would reduce the amount of MAP-triggered enrichment and opt for more TPS-triggered enrichment. This is because load closely follows Throttle position.
Source: Joe Hogan
Source: jhuang76
Get your car running perfectly while it's warm, and then tune your cold start settings. What I did was to get the car to operating temperature. Then, have the VE tables set so the first MAP and RPM points (very bottom left) are near idle. Then, have it set to zero. I would adjust my IOT until my car would run at 14.2-14.3 AFR in open loop. Don't even mess with the cold start settings until you've got the car running right warm.
I set my VE tables at WOT and idle at zero, then did some test runs to see my EGC. I adjusted my IOT and TOG until I could idle at nearly zero EGC, and likewise for WOT. I ended up having to drop my TOG from like 9.2 to 8.4 (which explained why I had like -15 at WOT for most of my RPM range). Then, I started (and am in the process) of tweaking the values in between.
Source: Shiv
You should set the GAMMA to offer zero correction at WOT at your peak hp point (around 5500rpm). Above that, you will probably see negative correction. Below it (and above 3000rpm) you'll probably see positive correction. But it sounds like everything you are doing is fine.
Source: Shiv
There is a MAT-based enrichment that you can adjust. But before you do that, try out Angus' calibration file. If anything, a colder day should result in more positive corrections due to increased air density.
Source: Shiv
Source: Shiv
Source: Shiv
Source: Shiv
Source: Sean McElderry
Source: Shiv
It really depends. A little extra fuel never hurt anyone. However, if you experience moderate or severe detonation, you will need to add a lot of fuel to quell it. Not recommended. Better to retard timing. However, the caveat about timing retard is that you can over-retard which is not good for the engine as it raises EGTs. This is why everyone should have an EGT gauge.
Source: Shiv
Source: Shiv
Source: Shiv
Source: Shiv
Source: Shiv
The TPS triggered enrichment isn't very effective at tuning out tip-in lean spots. It's more effective for tuning out lean run during boost rise/sudden high load. For good idle tip-in, you should make sure that the lower left portion of the VE table is dialed-in properly. This means that not only must your 800rpm at 40kPa (idle conditions) must be tuned. But also your 800rpm at ~80kPa. Needless to say, that's a hard zone to tune. You may have to lug your car in 5th gear at 20mph to reproduce it. You may find that you want to set your VE table to be a little rich in that zone.
Source: Joe Hogan
My solution to this was to allow the deccel function to engage at a kPa much higher than it was intended to engage. Not perfect but it avoids the tip-in problems associated with negative EGO CR.
Source: Shiv
Hi Joe (and everyone else) -- While your technique will work, it does so (as you know) by sacrificing a bit of drivability. Sometimes, especially around parking lot driving, you don't want your injectors to shut off. If they do, you'll get an abrupt shake. It’s easy to get used to. But not quite stock life. I recommend setting the deccel function to engage 5-6kPa less than the 35-40kPa you see at idle. As for TPS trigger, I usually put the setting on 4-5 volts/sec. This usually mimics the stock ECU's deccel function. However, as you know, you will see big negative corrections under deccel if (and when) the injectors remain on.
To eliminate this problem, you'll need to enrichen the a/f ratio target in your lowest MAP row (which should be set about 5kPa less than the typical load you see at idle). While the idle MAP row will be full of normal, near stoich a/f ratios, the row just below it will be full of very rich values. Absurdly rich, in fact. So rich that it will max out your EGO correction range (in the positive direction). In Angus' car, I programmed the a/f ratios in the lowest row to be anywhere from 12:1 to 13:1. To determine what a/f ratios to use in the lowest row, do the following:
1. Set up your deccel parameters so that the injectors always remain on (no deccel cut-off). Do this by setting your MAP trigger super low, like 5kPa.
2. Turn off EGO Feedback. You can either do this through the software or by pressing the letter "C'" in the engine monitor screen.
3. Next, datalog a 3rd gear coast down from 4000rpm to idle.
With the datalog, look at the a/f ratios recorded. Note at which engine speeds they occurred at. They will be rather rich as dictated by the low load and 1.0-1.1ms min. injector on-time.
Next, set up the lowest row in the a/f ratio table to have slightly richer values than the values you recorded at the engine speeds they were recorded at. Next, go back and set your deccel and EGO parameters back to normal.
What this whole process effectively did was to determine the leanest a/f ratios that are achievable with the given injector min-on time. Setting the a/f ratio targets slightly richer than that will ensure positive corrections whenever the injectors stay active during engine braking.
Source: Shiv
Source: efoo
Source: Shiv
Source: Shiv
The torque dips and peaks are, for the most part, a true indication of any given engine's VE profile. Of course, it assumes a close to optimal state of tune but we all have TEC-IIs, right?
Source: Shiv
Source: Shiv
Usually, lean run causes idle roughness while excessive rich run may yield a lower pitched, lazy drone-like sound usually accompanied by super stinky exhaust fumes.
Source: Shiv
There a couple of things to look at. First, when the idle is oscillating, take a look at EGO CR. If it is moving back and forth with the oscillation, then the two are related. That is, the EGO correction system is causing the oscillation. If this is the case, go to EGO Parameters and "slow down" the feedback system. One step at a time, try increasing your Instantaneous AFR divisor Error and Number of Samplings for Average Error. Take small steps. If you slow it down too much, your o2 feedback will be lazy and hard to tune off of.
Also, make sure that (at idle) the EGO CR isn't too far away from 0. In fact, it should be in the positive side a few percent. If it is in the negative side, try reducing your injector offset by 0.125ms. This should bring the corrections into the positive zone. If it doesn't, try it again.
Also, make sure that your actual injector on-times aren't "flirting" with your Minimum Injector On-Time. This will cause excessive correction as the EGO feedback system will try in vain to achieve an a/f ratio that it can never reach.
Also, make sure that your a/f ratio target at idle isn't too lean. Typically, it will want to idle smoothly at 14.3:1. If you go much leaner it may oscillate.
Here's another cute little trick. Oscillation can also be caused by a VE table that’s slightly off. For example, during oscillation, your MAP may bounce from 35kpa to 40kpa. It is very possible that the car wants to run rich at 35kpa and lean at 40kpa, or vice versa. This will make it impossible for EGO correction to stay on top of things. As a result, the EGO correction will chase its tail. The goal is to make sure the A/F ratio is constant at all conceivable idle MAPs (lights on/off, fans on/off, AC on/off). You may have to make adjustments in the VE able to ensure this. Also, at different loads, some engines want to idle at different A/F ratios. This will require some changes to the A/F ratio table. Perhaps new breakpoints as well. Maybe one at 50kPa and one at 40kPa. I had to do such juggling in my Miata. Running 550cc/min injectors on a 1.8L motor at a constant 45psi of fuel pressure is no joke. My Impreza was much more forgiving.
Source: NickCat
One small thing that I also did today, after talking with Shiv last night, I changed my AFR at idle to 14.3 instead of .4. This made a big difference! My engine is now finally running a much, much stronger idle that almost doesn't move. I think the TEC and the ECU idle control were just fighting with each other over the proper AFR.
Also a little tip, which just occurred to me today... tune you car at idle when it's fully warmed up. You have to have the lower left cell in the VE maps filled out correctly for idle when warm before you start modifying your warm-up enrichments. You can't modify warm-ups at 80 because the car is already warm, so you need to have the base number in the table first, then work on the warm-ups that are based off that number.
Source: efoo
Nick, you say the lower left cell of the VE table needs to be tuned in properly first—so in my case, that's @ 1300rpm and 30 kPa. Currently that is 0—should I just rev the car up to 1300rpm and note what MAP reading I’m getting, then tune accordingly?
Source: NickCat
Yes that's the right cell: 1300 RPM and 30 kPa. Best way to check for the correct value it is to first have the car warmed and idling (outside if possible so you're not pulling warm air or inhaling exhaust for a few minutes). Then note the EGO correction at idle. Take this number and add the appropriate number to gamma it to make the EGO Correction = 0.
For example:
EGO Correction = -3%
Put a -3, in the GAMMA cell on the monitor screen (then hit enter), to add this value to the correction base and EGO should go to 0.
Wait and note if the EGO changed and if the correction is 0... or close to it.
Source: Shiv
The bottom left cell in the tables will heavily influence idle. If you put a 14.2 in
1300rpm/35kPa, you will see something awfully close to 14.2 in the Desired AFR (monitor screen).
As for EGO correction, however, it's better to see a slightly positive number at idle. A negative number may cause a little lean spot at throttle tip-in.
Under full load conditions, it’s better to see the EGO CR subtract fuel than to add it (i.e. a slightly negative EGO after correction). The latter often implies too lean of a fuel curve, requiring EGO feedback to jump in and save the day.
Source: efoo
I now have a nicer idle! My A/F ratio in the lower left corner was already 14.3, so I left it alone. Here's what I've done:
· reduced injector offset to -0.625 (was -0.500)
· VE table entry to 8 @ (1300rpm, 30kPa), and 5 @ (1300rpm, 40kPa)—I'll explain the second one later
· EGO params—# engine events/sample = 6 (was 8)
· EGO params—divisor = 32 (was 16)
· EGO params—#samples = 8 (was 4)
This produces a very nice idle right between 650 and 700rpm.
The reason for the second VE entry change—whenever the radiator fan turned on, my idle would start hunting again. I could tell that my EGO CR was hunting as well—it was running a little lean. When the radiator fan turned on, revs went up to about 800 and MAP = 40kPa instead of 675 rpm and 35kPa at idle with no radiator fan. So I figured I'd add a little fuel at the next MAP point (60kPa) to make the interpolation slope a little lower.
Now it only hunts for a few seconds, then settles down, when the fan turns on. And when the fan turns off, it takes only 2-3 seconds to settle down. Same goes for when A/C is turned on and off.
Source: NickCat
I have to agree with Shiv after seeing my AFRs this morning. I have an almost perfectly smooth idle now. It idles between about 700-730, but usually stays about 713 and sticks. I did, however, tune for a zero to negative EGO correction and I have that lean tip in Shiv spoke of. So I will be adding a little more to the lower left cell. I only have 2 in there right now.
Source: Joe Hogan
First off my I found that if the ignition was advanced much more than 21 degrees in the lower left box of the ignition table the idle would suffer. On my MY99RS excessive advance at idle introduces roughness without any real benefit.
Next I worked ... and worked ... and worked on getting the VE table to where the EGO adjustments at idle when warm was no more than 5% positive. This is repeatable and very steady. The lower RPM VE values produce very little EGO adjustment around town. Most under 3% positive or negative. When on the throttle hard the adjustment is more pronounced. But then only a transient adjustment until things settle down. I think all of this together allows the engine to smoothly decelerate and capture a steady idle.
Even when the car is unloaded, in N or P, at idle and a sudden load comes on ... a cooling fan for instance ... the idle only oscillates a couple of cycles and then settles down. In D or R the idle never varies. Remember the car is 4EAT. The EGO correction parameter is still set to 16, which is average.
Source: Shiv
You can add some more coolant advance during start up. When the engine is cold, fuel atomization and combustibility decreases; hence, the need for more advance. You'll find that adding a few more degrees of advance during warm up may reduce some of the richness you are experiencing. Just be careful about adding too much. And make sure it all goes away by 65F or so. Give it a shot and let us know what you find.
Source: Shiv
Don't worry so much about absolute engine speed. The stock idle control logic is quite complicated and will vary target engine speed depending on a number of variables. Just care about idle stability. A smooth idle is possible on virtually every setup. It just takes a little bit of tuning, as every car seems to be a little different. And when you are using 550cc injectors, these little differences need to be addressed if a smooth non-oscillating idle is to be achieved. A lot of it has to do with slowing down or speeding up the EGO parameters. Read each of those F1 help screens and try to figure out what you are seeing on the engine monitor screen (corrections, afr, injector on-time, etc.).
Source: Joe Hogan
The car seems not to mind 15 AFR on idle. I see that during the end of the cold start routine that I've leaned out considerably. However at 60C there is some major adjustments as the AFR reverts to an EGO adjusted 14.3.
Source: RS'ted
A 14.1 entry in the lower left corner field doesn't necessarily equate to a 14.1 target A/F ratio at idle, as the TEC linearizes between data points. I have found it necessary to enter "14.15" in order to actually have a target A/F ratio of 14.3 at idle.
Source: Shiv
It's not so much the target that is important. Instead, it's important to get good closed loop rich-lean-rich afr fluctuations during idle. Going either too rich or too lean during those fluctuations will cause engine oscillation. Tiptoeing just around stoich (~14.7:1) will provide best of all worlds. Typically, you will want see actual afr jump around between 16 and 14.2. By adjusting desired afr and manipulating EGO parameters, you can achieve exactly this.
Source: Shiv
Achieving those A/F ratios will have a lot to do with fuel pressure (during decel), injector size, injector impedance (which affects min. on-time) and MAP pressure (during decel).
First, make sure your injector minimum on-time is set low as possible. While setting it too high can make it easier to achieve a stable idle, it can also make it impossible for you to run lean during heavy vacuum (like during decel). Running the big low impedance 550cc/min injectors, my min on-time is set at 1.0ms. During decel, my actual on-times don't come close to that. Instead, with 17:1 a/f ratios, the hover around 1.05-1.07ms. Then again, with the rails and high flow fuel pressure regulator, I am not getting any fuel pressure problems under vacuum (i.e. higher-than-stock fuel pressure during low load conditions).
If you are currently running the stock injectors, which are high impedance-type, you may have problems operating them down to 1ms. Instead, you may find that your minimum allowable on-time to be 1.2-1.6. You will have to find it through trial and error. To find this value with version I, do the following:
1. Set your minimum on-time to zero.
2. Turn off EGO correction.
3. Download and start your car.
4. Lodge a folded up piece of paper behind your throttle actuator. You will need to prop it up so the idle settles in around 2000rpm. This is to make the engine harder to stall as well as to induce some extra heavy vacuum.
5. With the engine running, look at the injector on-time. If the engine is running smoothly without any misses, the on time can go lower. The idle is to see how low you can go before your engine starts to miss (that is, until an injector stalls). So, in small 2% increments, start subtracting GAMMA from your engine monitors screen. The car will start running leaner and leaner. Don't be surprised if your a/f ratios go as lean as 17:1 before your car starts to miss. In fact, in one customer MY00 turbo, I've seen it go as low as 18.5:1 before the engine started to miss. The goal is to see how small the on-times can get without running into injector stall. Once you find that lowest allowable on-time, put it into your minimum on-time parameter. That's it.
6. Also, check your voltage compensation in your MAT compensatation parameter screen. Set it to 110. The default from Electromotive is set too low.
Source: Shiv
Don't know if I've made an announcement about it, but for the last couple of months, all of the new Subaru and Miata baseline files run between 7 and 14 degrees of advance (+/- a few deg for idle control). Too much advance raises combustion temps, which raise NOX emissions. Too little kills throttle response. Important to find a happy medium. If I recall correctly, the MY99 Subaru baseline runs around 14 degrees at idle. The MY00 runs around 7 degrees at idle.
The initial ignition advance, in most instances, should be set the same as the first column in the ignition advance table.
What is the function of the minimum injector on time?
Source: Joe Hogan
The Minimum Injector on Time’s function is to keep the injector from stalling. Or is this used to put a "floor" under the idle to eliminate very lean conditions with idle oscillations? Using a Minimum Injector on Time that is just below the minimum fuel required for a stable idle works well to flatten the idle.
When you get big negative adjustments the idle will be dead flat but you end up with a tip-in lean spot. The Minimum time can be adjusted in very small increments until all eventualities are accommodated. For example, try 1.04950 then 1.04925 until the correct Minimum time is found.
It has been my experience that the Engine Monitor Screen indicates the calculated Injector on Time. However, the program will limit the lowest time by referring to the Minimum Injector on Time setting. You will note that if you set the Minimum Injector on Time too high the Monitor screen will attempt to adjust to a very low value and the EGO will go way negative.
Source: Shiv
The on-times on the screen (and on the datalog) are calculated—not actual. And when they fall below the minimum on time you set in your calibration file, you know that you are asking the TEC to achieve an unachievable a/f ratio target. Took me a while to figure that one out myself.
Source: Shiv
If you're idling with big negative CRs at idle, you will need to reduce Injector offset a bit. Similarly, your TOG may be a little too high. I would keep it around 10.5-11.0 and set the IOT to whatever it takes to bring CRs down to close to zero at a fully warm idle. You'll probably end up with an IOT between 0.75 and 0.5ms.
However, during cold start, you will need extra fuel to keep from rough running or stalling. During warm up (and the first 20 seconds of hot start), the EGO correction system runs open loop (no EGO CR). The tricky part is tuning for all the various types of enrichment (1 second, 20 second, coolant-based, etc.) Unfortunately, living in the warm land of CA, my cold starts may not work quite well—it never gets cold here! But you may want to copy Joe Hogan's, Efoo's, or anyone else who's nailed it and lives in your winter wonderland.
Source: Efoo
For those of you living in colder places, you may want to turn up the injector on time for temperatures below 30C (PW0). Shiv's default map puts it at 2ms in order to avoid having to crank the car twice (once to get gas into the engine, once to actually start it) I've had to up it to 3.5-4ms. Note that this has nothing to do with warm-up situations, just cold start!
I wonder what this will do once it gets warmer, but for now, it certainly sounds a lot nicer than the "whrr! whrr! whrr!" sound followed by dead silence.
The help text states you should try 3ms for turbo engines. I've found that 3.5ms lets me start up on the first try within 1 second every time in the morning. 3ms did start, but it took a couple seconds of cranking and the engine sounded like an anemic, wheezing horse on its last legs when it did start, before finally revving up to idle speed. It just didn't sound right.
Source: Efoo
Since I have the (mis?)fortune to live up north, I actually use the left half of the starting enrichments graph, unlike you California weenies. Here are some numbers that I've come up with that seem to start the car up reliably from temperatures ranging from 0C upwards, without staining the garage floor with tons of black carbon deposits (which is what was happening with the default numbers I first received from Shiv).
Starting Enrichments:
SE0: 40%
SE1: 0%
PW0: 1ms
CLT0: 30C
ASE0: 30%
ASE1: 0%
Warmup Enrichments:
-30C: 60
-20C: 40
-10C: 30
0C: 20
10C: 14
20C: 8
30C: 6
40C: 3
50C: 0
60C: 0
70C: 0
80C: 0
Generally, I found that the default warm-up enrichments are way too high for lukewarm-start (30-40C) conditions -- often the engine would run super rich and sputter and die because the EGO corrections don't start kicking in until 40C. I start my car when MAT=35C all the time these days, so I am reasonably sure that this should work for you.
My car: MY99 2.5RS turbo, TOG=11.0ms (scale the numbers up by 3.8% if your TOG is still 10.6).
Source: Joe Hogan
Recently I've changed TOG to 11 and the IOT to -0.5. This has helped during cold weather. The MAT temp enrichment/enleanment is at 9%. This will be changed as I test more during various temperature ranges.
Source: Shiv
To those that are still experiencing idle oscillations, check your idle speed. Chances are that it is too low. Anything under 750rpm can be considered too low. For some reason, some cars have this while others don't. There's an easy way to bump up the idle to a nice 800 to 850rpm. On the back side of the throttle body hinge (facing the firewall, just under the main spring), there is a small, brass Phillips head screw. It is secured into the throttle body and held tight by a small lock nut. You will need to loosen the lock nut and rotate the screw clockwise. This will "prop" the throttle body mechanism open on the other side. Rotate it until idle speed rises to 800-850rpm. Then tighten the lock nut again. Next, you will need to reset the throttle position sensor (the black plastic box on passenger side of the throttle body). You will need to loosen the Phillips head screws that hold it in place. Then, rotate the sensor until you see (at idle) a TPS voltage of 0.47v. Voila. The whole process should only take a few minutes. The slightly higher idle will greatly improve idle stability. Just don't over do it and end up with a 950rpm idle. 850rpm should be more than sufficient. Try it out, but only if you are experiencing idle lumpiness.
You should really have to only do an ECU reset once unless you do some serious retuning in the idle zone. Keep in mind that all these tricks (raising idle and ECU reset) aren't necessary for a smooth idle. But they seem to make it easier for those who don't have lots of prior engine management tuning experience
Source: Sean McElderry
I did this a few weeks ago to raise my idle. And it worked...my idle is stable and is between 800rpm and 900rpm (depending on conditions). In the process however, I completely stripped my idle control screw. Do not try to tighten the sucker down very hard. Just tighten the bolt a little bit and things will work fine.
Why is the idle more stable with the A/C on than with it off?
Source: Shiv
The reason the idle is slightly more stable with the AC on is because the engine is sucking down more fuel under the load. A minimum load, those big honking injectors are teetering on their mechanical efficiency.
If you think this is a bit tricky, see what the Brits do. Using Link ECUs with stock injector drivers and 480cc injectors, they have to idle stink rich to get anything which resembles an idle. With 550cc injectors, at least one person is seeing idle a/f ratios of 11:1 (pinned on the rich side of his a/f ratio meter). I think they are less sensitive to noxious gas fumes than we are
Source: Shiv
Try this:
As for the idle oscillation problem, if all else fails, try adjusting the throttle stop on the throttle body. There is a 8mm lock not that needs to be loosened. Then, just rotate the small screw (tighten it) a turn or two until you get a nice 850-900rpm idle speed. Then tighten the lock nut.
For some reason, the 2.5RS is set with a very, very low idle speed of 600-700rpm. This causes the engine to lug unnecessarily. Couple that with big injectors and an intentionally slow closed loop feedback system, and idle stability is sacrificed. Bumping the idle up a couple hundred rpm will make your life a lot easier. Of course, if you do this, you will need to reset the "TPS voltage at closed throttle" setting in your calibration file as it will have raised 0.03-.05 volts. That's about it though. Don't bother adjusting your stock TPS. As long as the closed throttle voltage is under 0.5v or so, the stock ECU won't fault or contribute to idle funkiness.
PS: In raising your idle speed, you may need to readjust your Injector Offset. For example, when I adjusted my throttle stop (TPS voltage went from .41 to .47), I had to change my injector offset from -.625 to -.500. You will find that a lot of your VE table nonlinearities around idle is caused by too low of an idle speed. With an 800-900rpm idle, the engine will behave more predictably.
Source: Shiv
A stumble typically indicates lean run, which is entirely possible on a cold engine. The colder an engine is, the more fuel you need to feed it. Try increasing your coolant based enrichment value at that coolant temp. See if that helps.
Source: Sean McElderry
Make your car richer during warm up.
My Calibration settings are:
Calibration Screen->Fuel Enrichment->Starting Enrichments
SE0: 60%
SE1: 10%
PW0: 1ms
CLT0: 30 degrees C
ASE0: 20%
ASE1: 16%
Calibration Screen->Fuel Enrichment->Warm-up Enrichments settings
%FPW = 90, 70, 50, 30, 20, 5, 0, 0, 0, 0, 0.
Source: TimZ
I don't see a big reason not to do the EGO feedback at 45degC, but I do see a very good reason to fix the underlying problem.
If you are running so rich at startup that you (or someone with a sense of smell) can smell gas inside the car, then the possibility of flooding the car while attempting to cold start in cold weather gets pretty high.
While this doesn't sound so bad at first, you should realize that once the plugs get wet with fuel and misfire, it's very likely that they will become carbon fouled, at which point they will pretty much never fire, until they are either replaced or the carbon is burned off the insulator with a propane torch. Believe me - this will leave you stranded.
I have seen this happen, both with the TEC, and even on a pretty exotic, high dollar European sports car (which shall remain nameless) with a marginal OEM cold start calibration. It was marginal, but it sounds like it was still better than what you currently have.
It's not that hard to dial in the coolant temp-based enrichment. All you have to do is monitor the engine on cold start, and use the gamma adjust to tweak the mixture as the engine is warming up. Try to get it as close to stioch (or what ever you normally run at idle) as possible, without the engine stumbling. You'll have to take some notes as to what the coolant temp was for various amounts of gamma adjust, and then adjust the coolant temp based cold start enrichment table accordingly.
Source: Joe Hogan
The trouble I'm having is a slight hesitation at throttle tip-in off idle when the coolant temp is below 60C. This is something that also occurred with the stock ECU. MY99RS has a 4EAT so this problem is not likely to trouble most of the other RS' on the board.
Source: Shiv
I suspect this tip-in dead spot may not be fuel or ignition related. Automatic transmissions tend to add another level of complication when it comes to getting torque to the rear wheels. I know that my Infinity (4 speed auto) tends to resist kicking down a gear and/or engaging firmly when cold.
The only thing I can think of that can otherwise cause a tip-in dead spot is a lean run condition, which is easily datalog-able.
Source: Joe Hogan
Apparently this riddle has been solved. The coolant based gamma enrichment (ACE3) was the key. I added 10% in the Gamma enrichment and 100% in the coolant based TPS movement based enrichment. It is located in either the MAP or TPS enrichment windows. Like magic the hesitation/stumble just disappeared. Over the next few days I am going to try to remove some of the enrichment—to save the environment, and my gas budget
This worked for the MY99RS w/4EAT. This vehicle is still N/A but probably will work for most vehicles that suffer this problem. This field increases the gamma when the engine is in the warm-up phase. Goes to zero adjustment above 80C.
Source: Shiv
If you think the EGO system is over/under correcting or is too slow/fast, then you can adjust the EGO parameters accordingly. The F1 help will tell you how to tweak the numbers. Just one small step at a time.
Another possibility is that your EGO feedback is forcing your injector to operate with too small of an on-time, causing it to stall. You can check this by looking at your AFR swings. If you suddenly swing to big lean numbers like 18-20 with injector on-times of under 1ms, you know that the injectors are stalling. Then, raise your minimum injector on-time according. Just be careful, if you raise it too high, you will not be able to achieve a reasonably stoich idle. It will be too rich and your EGO system will try to fight it in vain, causing big negative corrections.
As for EGO parameter settings, those who are having idle problems, try setting the AFR divisor to something big like 32. This is to compensate for those big honkin' injectors since a little correction yields a lot more fuel. You can probably set the Average divisor error to something relatively small like 8.
Source: Shiv
It sounds like the EGO correction is going very negative due to the inability to achieve the desired a/f ratio target. This would occur when: a) the minimum injector turn-on time is set too high, or b) the a/f ratio target at idle is too lean.
It's a
"false" correction as the minimum injector on-time is "road
blocking" the actual a/f ratio. EGO correction becomes stupid and keeps on
correcting in vain. Confirm this by looking at real-time injector on-times (in
the monitor screen) and see if they go below the minimum injector on-time
(whatever it is set to in the basic injector parameter screen). If so, reduce
the minimum on-time. If you go too low, the injector will stall and your engine
will miss. You'll see the a/f ratios go super lean for a second. You may find
that the lowest you can go is just above 1.00ms before stalling occurs. If you
still are seeing these wild corrections, try making your a/f ratio target at
idle richer.
Source: TimZ
The programmed minimum injector on time and the injector's actual minimum on time are not necessarily the same thing.
As you probably realize, the injector's actual minimum on-time is the point at which the pulse to the injector is too narrow, and the injector doesn't open at all.
The number that gets programmed into the TEC is a guess at what the actual number really is. Its purpose is to keep the TEC from sending the injector a pulsewidth that won't turn the injector on.
The problem is that if your TEC's min on time is programmed too low, it will still be possible for the injector to get a signal that won't turn it on, and you will see a lot of the idle traits that you are mentioning.
As you observed, richening up the VE table also makes this go away -- essentially you have increased the commanded pulsewidth a bit, and gotten over the hump. Problem is, (assuming that the min on time is in fact set too low) you stand a good chance of having the problem come back later, if, say, atmospheric conditions allow the motor to pull a little more vacuum.
One thing that you could try would be to monitor your injector pulsewidths at a decent idle, and set the min on-time to just under that value. Set the VE table back to where it was, and see if you still have a problem. You can try inching the min on time up, until the problem is cleared.
Also, you should be able to tell if the min on time is set too high, as the idle AFRs will suddenly be too rich.
Source: Shiv
Yes. My suggestion, as always, is to do the engine management first and the turbo later. I think this is the safest way to go. And finally, there are no disadvantages of running the big injectors and high flow fuel pump when NA. Other than you will only use 45-50% injector duty cycle. J
This will ease the tuning process because you will only be introducing one variable at a time. The maps I include with the kit are for a turbo EJ25. Of course, they work quite well on a non-turbo engine. The only difference is that you only use a small portion of the big map until the turbo gets installed. You shouldn't require a major tuning overhaul once the turbo goes in. In fact, tuning on an NA car first will optimize the NA portion of the maps, which will improve overall performance once the turbo goes in.
Of course, given
the choice, it's also better to learn the ins and outs of the TEC-II while the
car is NA.
Source: Sean McElderry
I'm pretty sure you can't lower boost with the TEC-II. But you can call Turbonetics and order a 3psi spring for $10.
Source: Shiv
It’s not a good idea to go to with a three bar MAP sensor if you only plan to run less than 15psi of boost. The reason behind this is reduced tuning resolution. Some 2 Bar NA guys (soon to be turbo) know what I mean as the whole NA portion of their fuel and ignition maps is only a small portion of the whole tuning landscape. If you eventually have some big boost aspirations down the line, you want to upgrade to a 3 Bar MAP sensor later. It will cost $80 (for sensor and connector). Changing MAP sensors, however, will require the re-tuning of your VE table. A Bar of properly managed boost is a lot of horsepower. Don't be to concerned with the big boost numbers run by others. It's apples and oranges once you start tuning correctly
Source: JKav
The TEC only sees 0-5 volts regardless of which MAP sensor (1, 2, or 3 bar), which is why it needs you tell it which one you're using. So yes, you will lose a bit of resolution going to a higher bar sensor.
Keep in mind, the TEC will linearly extrapolate beyond all your breakpoints. If running 14 psi boost, it is best to set the highest breakpoint to something lower than the max boost, say, 12 psi. Don't get too carried away with this approach, though.
Similarly, set your last MAP and RPM breakpoints slightly below the highest expected MAP and
RPM. Likewise, set your lowest MAP and RPM breakpoints slightly higher than your idle values. This will help improve the "granularity" of your tables by tightening the range.
Source: efoo
· installed MAT sensor in intercooler
· installed boost control solenoid
· relocated O2 sensor to new turbo downpipe
Now, software-wise, I've just done the following for now:
· made sure that the GPO output for the boost control solenoid is completely zeroed out
· adjusted EGO CR allowance back to 40% to allow the TEC to correct for lean/rich situations under boost
· turned knock sensing back OFF
· ignition timing for all values above 105 kPa is set to values in Shiv's baseline map
I expect to have to do the following once I start the car up:
· make sure the O2 sensor is working in its new spot in my turbo downpipe!!
· make sure the EGT and boost gauges really work
· recheck the idle MAP readings, with and without accessories running, calibrate idle settings if necessary (shouldn't have to do much, but who knows)
· run the car in neutral up to get some kPa readings for the various revs, watch boost control, EGO CR factors along the way, listen for knock
Once I'm satisfied that the car will at least not kill me as soon as I back it out of the garage, I'm going to start redoing the VE table maps. Of course, this only applies when the car is completely warmed up (80C).
From what Sean posted earlier, I gather that my MAP reading under WOT will be different now. So one thing I think I might try to play it safe is to keep revs low, under 3000 rpm, for the first WOT run, and tune that first, then use those numbers to extrapolate a VE curve for the higher revs that is closer to the real one. Then I'll make a WOT run up to redline in 4th.
Source: Shiv
Another thing which may help is a less restrictive exhaust system. The
greater the exhaust restriction, the greater the tendency for detonation.
Source: Shiv
No turbo timer, I'm afraid. I tried to keep costs low as possible in include only the bare essentials. Although Jkav (see below) is more qualified than I am in this topic, I would say that cool down period really depends on how fast the turbo is spinning, which depends on how hard the turbo was driven right up to that point. Typical cool down time would probably range from 20 seconds to 1 minute depending on this factor. My suggestion would be to slow down your driving (stay out of boost) for the last couple of minutes of your drive.
Another good thing you could do is run a good full synthetic oil like Redline. Cheap (well, not that cheap) insurance against oil breakdown and turbo/engine damage.
Source: Jkav
If you're using a water-cooled turbo and synthetic oil then you don't need one. Otherwise, you can still get away without one by, like Shiv says, staying out of boost for the last mile or so. Not a big deal when you consider that what's good for the turbo is good for the engine too.
Source: Jude DeMeis
Think twice before installing a turbo timer. We've had two episodes around here of cars rolling and causing damage when the handbrake slipped and/or failed. One turbo-timered Mitsubishi rolled into some meathead's car in a parking lot and the DSM owner almost got into a fight. Another VW that was parked in neutral took out three cars in our office lot causing a big insurance claim. Take Shiv's advice and just cool it down for a minute before you park. If you use synthetic oil, coking will not usually happen either.
Source: MattC (MY99 Minnam Turbo, not yet TEC2)
My third gear EGTs are right around 1500F—that is the lowest I can get them. I haven't done 4th gear pulls in a while, but I am guessing they are around 1650F. This is at 10psi of boost, and 80psi of fuel pressure.
Source: Sean McElderry (MY00 TEC2 Turbo)
I think my EGTs agree with MattC’s for the most part. At 6250rpm/WOT in 4th gear, I'm seeing around 1650F. At 6500rpm/WOT in 4th gear, they're around 1700F.
Source: R Diamond (bone stock MY98 baseline readings)
The magic EGT is 1400 degrees Fahrenheit. That's where the car seems to enjoy staying during 60-80mph cruise. Even during WOT blasts up to 100mph, it doesn't go over 1450. At 55mph, the car will range between 1300 and 1400 degrees. Warm idle is 800.
Source: Z1500
A thermocouple works by measuring the difference between the probe tip and the cold junction. The cold junction is where the wires coming off the back of your probe connect to the extension wires coming off the back of the gauge. If this cold junction is seeing colder air than it used to, it will throw your gauge readings up by the same amount. I would assume this is what is happening because winter is now in full force, and your cold junction is probably exposed to outside air. If you want to see consistent readings year round, move your cold junction inside the cabin so your climate control can keep it at a consistent temp. By the way, the industry standard is to calibrate EGT gauge's for a 70F cold junction. Just a thought, I could be wrong.
Source: Scottie
Might I suggest Secondary Air as a possible reason for the EGT increase? In some modes of operation, secondary air is used to help oxidize HC and CO emissions. This process produces heat, hence the more secondary air the more heat. With colder temps comes denser air. If the secondary air control system doesn't have the means to compensate for such density changes, then you may get more secondary air (from a mass standpoint) under the same circumstances during colder weather than warmer. Remember, I said it was just an idea. An idea based emission controls in general, not necessarily Subaru in specific.
Source: Shiv
Fuel pressure should always go up with boost in a 1:1 ratio. Starting at 0psi of manifold pressure (108kPa), you should see 43psi rail pressure, if you are running 5psi of boost, fuel pressure should go up to 48psi. 10psi of boost should see 53psi of fuel.
The opposite holds true for manifold vacuum (less than 108kPa). The more vacuum, the less fuel pressure. At 20" of vacuum, you should see around 35psi of fuel.
This relationship ensures a constant pressure differential between the fuel rail and intake manifold. This is why there is a vacuum reference line running to your fuel pressure regulator.
However, for those running high flow fuel pumps with factory regulators, the fuel pressure may never go below 43psi -- even under heavy manifold vacuum. This is because the tiny factory regulator cannot return enough of the massive amounts of fuel flow back to the fuel tank. In other words, the high flow pump is overpowering the tiny regulator. This is also why you may see fuel pressure oscillations at idle. In terms of performance, this isn't a big deal since closed loop o2 feedback can compensate for the fluctuations in rail pressure.
My kit will include a high flow fuel pressure regulator which will replace the factory regulator. This will take care of the pressure anomalies by being able to return enough fuel back into the tank. In other words, the rail, pump and regulator will all complement each other.
Source: Shiv
It is a two port skinner valve that acts as the boost controller. Whether a 3 port solenoid works better is unknown to me. I've used the Profec and EVC IV, both of which (if I recall correctly) use 3 port solenoids. I've tuned the TEC to objectively match or even out-perform both. I suppose it's an instance where extra tuneability makes up for any inherent advantage(s) in solenoid design.
Source: Shiv
0-100 refers to solenoid duty cycle. 0% means the solenoid is closed and therefore the wastegate will stay shut, and 100% means the solenoid is wide open and the wastegate will snap open.
More duty cycle, more bleed, more boost. In first setting up/tuning a turbo car, I would recommend zeroing out the whole table, and adjusting the wastegate down to 5-6psi. Then tune ignition and fuel accordingly. Gradually creep up by adding "GPO (general purpose output) Offset" in the engine monitor screen. You'll find that each addition 10-20% will raise boost by a pound or so. But take small steps because things can get hairy towards the top end. After you have your max boost to where you want it, you can massage the GPO table to prevent top end roll off, improve low-end boost rise, etc. Improper tuning can result in boost spikes and oscillations.
Source: Shiv
Here's the skinny on boost control. The GPO solenoid is doing nothing other than actively venting the wastegate signal. This, in turn, fools the wastegate into staying closed longer-- raising the boost in the process.
Before tuning your boost maps, one has to first realize that the boost controller can only raise boost, not lower it. That means that, with the boost controller installed but lifeless (with zeros in your GPO table), you will have to set your mechanical wastegate screw to provide a boost level lower than you really want to run. For example, if you eventually want to run 10psi, you will have to set your wastegate adjustment for less than that. On the other hand, setting the wastegate too low will make it impossible for the boost control solenoid, when active and on the job, to vent enough wastegate signal to achieve the desired boost. A good rule of thumb is to set the wastegate to 3-4psi less than the desired boost level. If you want to run 10psi, set the wastegate (with the boost control solenoid installed but closed/deactivated) to 6-7psi. With some of the Minnam turbo kits, this will require some serious tightening of the wastegate adjustable screw. But, as always, be careful and adjust and test in small increments.
Source: Shiv
So now you're driving with the boost control solenoid installed nothing but zeros in your GPO table. You see, at wide-open throttle, 6-7psi of boost. In reality, it's likely that your boost curve won't be totally flat and consistent. In lower gears you'll see less boost than in higher gears. For this reason, all boost testing should be done in higher gears (like 4th gear). You will also see some RPM-dependent boost level changes. While you get 7psi in the midrange, boost may taper off to 5psi in the top end. Or vice versa. This is normal. Just make sure that the peak boost you see is 6-7psi. All the other boost nonlinearities can be tuned when you dial-in your boost maps.
Now comes the fun part. While fun, caution must be taken as an overboost can cause some serious damage to your engine, morale and wallet. As said earlier, up to this point, your GPO table is empty and full of zeros. Instead of sticking numbers in there, reloading them into the TEC, testing, and doing this over and over again, we're going to use the engine monitor screen to make temporary and global changes to the boost control solenoid.
In the engine monitor screen, start adding GPO offset, in 5% increments, testing after each tweak (!!!), until you reach your desired peak boost. As the offset number goes up, you will hear the solenoid ticking quite audibly. It sounds very different than knock, by the way. Once that peak boost is recorded, pull up your GPO table and input that GPO offset value you ended up with in the engine monitor screen. You will only want to fill in the cells above ~130 or so kPa. If you fill in the lower MAP cells, you will hear the solenoid ticking all the time (even during cruise and idle). That's annoying.
Download the new program (with the adjusted GPO map) into the TEC. Now, your basic boost map is set. However, you will notice that you still have uneven boost levels throughout the rpm (i.e. top end fall off). Now comes the tricky part that requires some good judgment. First, note at which engine speed(s) that fall off is happening. Go to your GPO table and start increasing the values in the appropriate RPM cells (up and down all on-boost MAP levels). Play it safe and adjust up in 5% increments. Similarly, if you notice a boost spike at a certain RPM, you can lower the boost by lowering GPO values in the appropriate cells. Through this GPO table tweaking, you can create the kind of boost profile you desire.
Note: The engine monitor screen can only make +/- 30% GPO offset adjustments. If you find yourself reaching the end of the desired adjustment range, you will either have to adjust your mechanical wastegate screw or download the offset into the MAP and add in a new gob of offset on top of that.
As always, please let me know if any of this is unclear. Please be careful when tuning the boost maps as a bad cell entry can cause a terrible boost spike (and possible detonation).
Also, all of you who are running 2 bar MAP sensors should have an overboost protection that I programmed into your file. To confirm this, go to Sensor Parameters. Click on MAP Sensor. The high-voltage sensor failure should be set, if I recall correctly, at 4.60v. With 5v being a full 14.7psi, 4.60v is roughly 12-13psi. Please don't test it out. Once that is triggered, the CE light will go off and you will experience a very obvious full cut. For a safe fuel cut, the default MAP value should be very low. Low enough to prevent combustion instead of a dangerous lean run. Set the default MAP value to the lowest possible value.
Source: Shiv
You can tune the TEC-II's boost control for more gradual boost rise characteristics. Perhaps something like 7psi at 3000rpm, 8psi at 4000rpm and 11psi at 5500rpm. While this shouldn't affect full out acceleration times or peak hp numbers, it will reduce that big fat wad of torque you would otherwise get between 3500-4500rpm and during the next gear grab. Doing so makes both the tranny and the clutch a lot happier.
Source: Shiv
Go to MAP sensor parameters and check out your high MAP voltage failure value. It should be set to 4.65v, which corresponds to approx 11psi of boost on a 0-5v 2-bar map sensor.
Joe, (and just Joe), you may want to raise it to 4.8v since you may already be brushing close to that boost level. If that value is exceeded the TEC will default to the default MAP value which is set to something really low like 30kPa. This is basically a boost-activated fuel cut. Very obvious as some have found. Hey, it's better than a blown motor, eh?
This method of overboost protection will not result in lean conditions. Lean run conditions are usually result of having, say 5-10 percent less fuel than desired. Cutting fuel by a whopping 90% will result in absolutely no combustion. In fact, cutting fuel by 50% will do the same.
For what it’s worth, A/F charges loose volatility with mixtures much leaner than 18:1. During the 11psi fuel cut, the a/f mixture works out to be somewhere around 100:1. While there is still some fuel being squirted out by the injector, it's just not anywhere close to being enough to do anything interesting.
Source: Shiv
Check your oil drain line. It's a
gravity drain, so if it doesn't slope downwards all the way from the turbo to
the oil pan, it will get pushed through your turbo seals and get into your
exhaust. Also, those who don't run cats will notice a lot more fumes
(either oil or gas) and puffs of smoke than those who do.
Burnt oil can range in color from black to blue to white. I've seen it come in all sorts of hues. Mostly, it depends on how dense it is. All cars squirt fumes out the back but the cat tends to obscure it completely. Without a cat, lots of groovy crap tends to come out. But still, you should look into it. First thing to do is jack your car up and make sure your oil drain line is going downhill all the way from the turbo to the oil pan. There isn't much vertical room to work with so you will have to work at it. I ended up cutting some excess drain line off. But then again, I have one of the earliest kits.
Source: Shiv
It's a good idea to keep an extra magnetic pickup in your glove box. It's the only single point of failure in the whole TEC system. If, for some reason, the pickup mount comes loose (because you didn't use Loctite on the bolts!), the pickup can smack against the spinning toothed wheel. If that happens, the pickup gets destroyed and the TEC can no longer read engine speed. If this happens, the engine will shut off and refuse to restart. Needless to say, the car is sidelined until you toggle back to the stock ECU or get another pickup. To make things more annoying, it's the only sensor in the system that can't be purchased at your local auto parts store. So, contact me if you want a backup. The price is $30+shipping through me. $47+shipping through Electromotive.
It appears that jhuang76’s crank trigger pickup came loose and hit the toothed wheel while he was driving on the track. I'm not sure how this could happen if the bolts are tight and Loctited in place. I have a feeling that the screws holding the magnetic pickup in the holder weren't tight enough, letting the pick up slide towards the spinning toothed wheel and... whack! Hence, my suggestion to keep an extra pick up handy.
Source: kartboy
I don’t think the toothed wheel can get damaged like the trigger assembly. Just watch Battlebots and see what a flying toothed wheel can do to anything in its way. I have my extra pickup in the car with the little alan wrench to tighten it just in case!
Source: Shiv
O2 sensors will, at times, be able to clean themselves as long as you keep from operating in rich conditions. So, basically, drive on the highway in closed loop operation, and stay out of boost for 20 minutes or so. Just stay off boost and drive around.
EGO correction off = open loop
EGO correction on = closed loop
How and why should I keep my
TEC dry?
Source: RyanC
It's those stanky exposed TEC connections, I'm tellin' ya! I've had the same problem in the past, after hosing off the engine bay the car would run like crap for a few minutes until it dried off. I have electrical taped over the whole connector and down the wires about 3-4 inches, but it's not a permanent solution. Since then though I haven't had any similar problems (knocking on wood very loudly right now). Didn't someone say a while back that they were going to investigate a weather-resistant 'box' for the TEC?
Source: Shiv
It's probably water sneaking through the ignition wire boots—either that or the wires themselves being knocked loose by the high-pressure water spray. That’s pretty easy to do when dealing with a flat four as the valve cover is horizontal and close to the ground. You may want to put a dab of silicon around the seal to prevent water from seeping in. This will keep them from coming loose. I can assure you that unless you put the TEC on the outside of the car, water will not affect it.
Also, where have you guys mounted the TEC unit? If it's in the area right in front of the ABS system, then yep, water can get into the TEC unit as a puddle tends to form right there. Especially if the passenger side hood vent is removed. In that case, not only will you want to mounted the TEC on spacers (to lift if off the surface 1/2" or so but you will also want to put some sort of cover/shield on it. High output coils and water don't mix!
This is why I prefer to mount it vertically on the driver's side firewall. Mounted there, there is no way water will ever touch the TEC unit unless blasted by a high pressure unibody spray from a drive through car wash.
Source: Efoo
My TEC is mounted on the firewall according to your recommendation, Shiv. But it does get wet. If it's raining or snowing outside, the rain just flies in the hood scoop, misses the intercooler, and hits the TEC unit. I can attest to this personally as I've checked my TEC unit after heavy rains and documented a rather soggy TEC input connector. That's why I'm building a cover for mine.
Source: Shiv
In that case, yes, a little plastic shroud would help keep those coils and terminals dry. I guess the idea solution would be to properly duct the intercooler. Get two birds (better intercooling and a dry TEC unit) with one stone!
Source: Shiv
The gauge and the MAP sensor don't read in the same scale so don't bother comparing them -- even when the both read in inches of Hg. One is absolute, the other is vacuum. Confusing? Yes. J
|
TEC-2 |
PSI (absolute) |
in-Hg |
Boost Gauge |
|
|
|
|
|
|
10
kPa |
1.5
psi |
27
in-Hg |
|
|
15
kPa |
2.2
psi |
25.6
in-Hg |
25 |
|
25
kPa |
3.6
psi |
22.6
in-Hg |
|
|
35
kPa |
5.1
psi |
19.7
in-Hg |
20 |
|
50
kPa |
7.3
psi |
15.2
in-Hg |
15 |
|
65
kPa |
9.4
psi |
10.8
in-Hg |
10 |
|
85
kPa |
12.3
psi |
5.0
in-Hg |
5 |
|
100
kPa |
14.7
psi |
0
in-Hg |
0 |
|
110
kPa |
16
psi |
Boost! |
+1 |
|
135
kPa |
19.7
psi |
Boost! |
+5 |
|
150
kPa |
21.8
psi |
Boost! |
+7 |
|
155
kPa |
22.5
psi |
Boost |
+8 |
|
180
kPa |
26.1
psi |
Boost! |
+11 |
|
187
kPa |
27.1
psi |
Boost! |
+12 |
|
200
kPa |
29 psi |
Boost! |
+14 |
Source: Shiv
The tachometer is off. This is very, very common as most tachs are optimistic. The TEC will always give you a perfect rpm reading as it is accurate to the 1/4 degree all the way up to 14k rpm. If you want a more accurate gauge reading, you could wire the tach to the TEC.
When my RS was stock, the 6250rpm rev limiter kicked in when the gauge was reading 6700rpm.
Source: Shiv
If you go to the sensor parameter screen and look at the O2 sensor parameters, you will see that it is set to fail if it detects lean run for more than 255 seconds. This will only happen if you are trying to zero-in on an AFR that it can never reach. For example, if your AFR target is 14.3:1 at idle, yet your actual AFRs scan between 14.35:1 and 15:1, and it stays this way for more than 255 seconds (from a prolonged idle), it will trigger a sensor failure. It rarely has anything to do with the sensor actually failing. It's almost always caused by setting unobtainable AFRs (due to EGO authority range restrictions) and staying leaner than the target for an extending period of time. The solution is to either reset your AFR value to match the VE table or vice versa or set a wider EGO authority range. In other words, don't buy a new sensor.
As for the o2 sensor, if you are bouncing around on both sides of your AFR target (which is good as long as the idle is stable), your intermittent sensor failure is probably caused by steady state cruise where your AFR target of say, 15:1 is out of reach for some of the reasons I mentioned earlier. In 5 years of playing with TEC-IIs, I've only seen one real sensor failure (MAP sensor). GM sensors are very reliable.
Source: Shiv
Typically, when o2 sensors fail, they lose output voltage, which means that they will read lean. This will cause the ECU to add in extra fuel. The end result is that your car will run unusually rich. This will eventually clog your cats, reduce your fuel economy and foul the o2 sensors and spark plugs.
Source: Shiv
Do this by unclipping the injector connectors, one at a time.
Every time you unclip one, the engine should lose a cylinder. You will know if
the car is only running on 3 injectors. It would sound very unbalanced. To test
for fuel leaks, just wrap each injector with tissue paper, go for a boost-full
drive, pack it, and sniff the tissues. Don't sniff too long though.
Source: Shiv
It's easy to check if the MAP sensor is functional. Just turn
on your engine monitor screen with ignition power on, but engine off. Your MAP
pressure should read atmospheric. Depending on your altitude, that will be
somewhere between 95 and 110kpa. If it reads something like 80kPa, congrats,
you have a bum MAP sensor. If this is the case, shoot me a personal message and
I'll work on getting you a replacement.
Source: Shiv
There seems to be a little confusion on how to confirm your ignition timing with a timing gun. Well, here's how you do it:
Start your car. The timing gun has three leads coming out of it. Two are for battery power and ground (it powers the gun). The third is an ignition pick-up sensor. This is attached to the ignition wire going to cylinder number 1 (front left cylinder when standing in front of the car, looking towards the rear of the car). Attach all the leads accordingly. If the gun has an adjustable dial, make sure it is set to "zero". Right now, if you press the trigger it will start strobing away like mad.
Now, with the car idling and the engine monitor screen up, click on the "Set Advance To Zero" button. Your ignition timing will immediately drop from a normal at-idle 18-20 degrees to a rock-steady 0. Then with the timing light, check your timing. Use the timing mark on the stock pulley and see what timing mark it lines up against on the stock plastic timing gauge (on the engine cover). The timing mark should be slammed at the "0" mark plus or minus no more than 1 degree. If you are off by plus or minus 6 degrees, the magnetic pick-up is referenced on the wrong tooth. If this is the case, the ignition numbers that you think (according to the TEC) you are running, are off by plus or minus 6 degrees. Fortunately, the design of the pick-up makes it very difficult to be off this much but it's worth checking anyway.
Well, that's it. Just remember to click "Clear Zero Advance" when you’re done. Driving around with zero advance will not only feel disgusting, it will also raise your EGTs to uncomfortable levels. Be wary of this button. Use it only for checking timing. Do not drive around with zero degrees of advance. Please contact me if you have any questions.
Source: Shiv
A customer called and told me his crank trigger assembly (toothed wheel) was out-of-round by 0.012". Maximum out-of-round tolerance is 0.008". I hope it is an isolated incident. But just to make sure, I would like all the new TEC users to check something out. After you airgap your magnetic pick-up to 0.040-0.050", rotate the crank pulley (while it is bolted to the engine, of course). Look at the airgap as the pulley rotates. It should remain relatively constant at all times. If it doesn't, we have a machining issue to deal with. Please contact me if this is the case. Sorry for the inconvenience.
For those who are already running the TEC-II and have experience no erratic RPM pick-up signals (viewable in datalogs), you're in the clear.
Electromotive recently redesigned their crank pick up wheel to accommodate much larger air gaps and out-of-round tolerances. The recommended air gap for a 6" wheel used to be .024" to .028" with out-of-round tolerance of just .002". With the new wheel design, gaps and tolerances have increased by 200-300%, making things a lot less critical and much easy to install. Every one of my customers, new and old, has this new trigger design.
No one really knows how much tolerance is possible due to the recent design change. However, I have yet to see a gap that is too big or a wheel that is too out-of-round. But still, please let me know if you suspect that you have excess tolerance in your pick-up or are experiencing any erratic RPM data. Sorry for the inconvenience.
You would know if you had it. If you took a datalog and looked at your RPM graph, instead of seeing a smooth curve, you would probably see major fall outs or spikes. The engine probably wouldn't even run. But if it did, it would run very erratically and no amount of tuning would make it better. The funny thing is, I've never experienced it myself. Nor has anyone I know of. I suspect it is one of those things—either the engine runs or it doesn't.
Source: Shiv
TOG should not affect fuel pressure one bit. Fuel pressure is controlled totally by the FPR. Determine whether the drop in fuel pressure global or is it just under wide-open throttle? If it's global, I would tighten down the adjustment screw until you see 43psi with the FPR referenced to atmospheric (vacuum line disconnected.) The internal spring may have just loosened up slightly. Unusual, but possible. The only other explanations for a reduction in fuel pressure are a clogged fuel filter or a dying fuel pump. But in those cases, the pressure drop only occurs under load/periods of heavy fuel consumption.
Source: Shiv
I've only seen the rubbing problem on one other car. I suspect there is some variance in respect to the shape of the plastic engine cover. Perhaps it is bowed out on some cars, resulting in rubbing when the .125" thick wheel is mounted behind the pulley. Unfortunately, short of removing the plastic cover and shaving some of the plastic support brackets, there isn't any quick fix. But before you do anything, just make sure the pulley is on nice and square and not slanted (unlikely). As for long term, there's really nothing to worry about other than some scuffmarks on the plastic. It will self clear within a few hours of operation.
Source: Shiv
Check whether you have any leaks in the exhaust system (before the o2 sensor). This could potentially allow air (with o2) to enter the exhaust stream, making the o2 sensor read lean-- adding fuel in the process. Also, check whether your injectors are working properly.
Source: jhuang76
All I had to do to get my car going again was to swap out the injector controller (simply disconnect TEC-II connector and plug back stock) and disconnect the Magnecor wires from the TEC-II and plug it back into the original distributor. I reconnected the stock distributor connector and disconnected the power feed between the TEC-II and the battery. It took about 10-15 minutes (8 minutes struggling with the driver side nearest to cabin fuel injector connector because the stupid bracket right there—and two minutes healing a slightly burned finger—the engine gets real hot).
If I had a turbo and I set for zero boost (open up the wastegate and any bypass), it'd be like the NA car right? So, even w/ a turbo, if the TEC-II has problems, I could easily convert it back to a near-stock, OK running car.
Source: Shiv
It's extremely unlike that anything is wrong with the TEC unit. They are paragons of reliability. I suspect something else is going on, especially if the engine monitor screen is working with the ignition power on. In this case, it's only when you crank the engine that communication breaks. That suggests that it's a voltage issue (i.e., the battery isn't providing enough voltage during cranking.) You should check the voltage during cranking. The TEC, like most aftermarket ECUs, is very sensitive to voltage fluctuations (high output coils and injector drivers). A voltage shortage will shut it down immediately.
Source: Sean McElderry
1) Turn the key to On and load a .bin file.
2) Turn the key to Off and remove the key from the ignition.
3) Put the key back in and start the car.
This same thing happened to me
several times and whenever I followed the above steps, it went away. I had to
make sure I turned the key to Off before I started the car after loading
a .bin file.
Source: Joel Gat
My Bro's Mustang has had every possible electrical problem that a car can have, including several fires due to Ford ingenuity and wiring. That being said, he once had a problem that made absolutely no sense -- sounded much like this one, but he was running off the stock ECU (modded chip). I could not figure out what the heck was going on until I had this super strange grounding thought. What if the block wasn't grounded properly?
Well, in his case, it turned out that the engine tore itself free of the big fatty wire that attaches the block to the chassis. I know it sounds silly, but my suggestion is clean off and recheck the connections for the ground for the TEC, first. Then check the block ground too. If the wire shows any signs of unhappiness, replace it with a big fatty grounding wire and clean off the connection spots at the block and chassis as if you were going to put a contact lens on it and then in your eye. I mean clean!
On my bro's car we figured out that the engine was grounding through the starter motor (which had its own ground) and so under heavy use, the starter motor started developing internal shorts and all sorts of wacky things happened. Replacing that wire and the starter made the car run perfectly.
Source: Shiv
You will always need to turn the engine off in order to load a calibration file. Only engine monitoring (and engine monitor adjustments) can be done with the car running.
Source: Efoo
Part One: I finally figured out why my car would buck and jerk every time I went over freeway expansion joints. My BOV wasn't letting enough air by. Essentially, I would hit a bump, my foot would lift a little bit, therefore closing the throttle plate a bit. But turbo is still spinning, so all this air piles up in the intake manifold (since throttle is still partly open), MAP sensor reads "holy cow! boost!" and references the VE table to find out that I want +20 GAMMA there, therefore dumping buttloads of fuel into the engine. Engine bogs big time. I panic and stab the gas pedal, adding even more enrichment as TPS acceleration factors come into play, along with the end-of-decel enrichment. I'm dumping 150% GAMMA into the engine. Engine continues dropping in RPM, gas pedal has no noticeable effect. Finally, around 2000rpm when that 18-wheeler behind me is about to rear-end me, the engine levels off, burns off the excess fuel, big pop from the exhaust, and it starts running again. Repeat for every expansion joint.
Man, that was one tough one to figure out for me, since I knew nothing about what was going on. I figure you guys know better, but I'm posting this anyways so the next idiot doesn't pull an Edwin and thinks "hmm, gotta make sure that BOV doesn't open under boost. Let's tighten it down so tight it doesn't move!"
Part Two: My car will buck at part-throttle at random times. It used to be almost all the time, but I "cured" that by loosening up my BOV so it barely holds in the air at idle. And that helped a lot. I was able to go for 15-20 minute drives with no problems. However, I just went for an hour drive, and near the end of the trip, the bucking was bad. The jerking was so bad that the MAP sensor went off and triggered an intermittent fault in the TEC-II monitoring software.
The funny thing is, I took a good look at my boost gauge while this was happening. The needle was not moving at all. You would think that such bad jerking would result in some change in manifold pressure at least. And it took an hour for it to surface this time. It used to happen when I went over bumpy roads and the like.
Yes, my car does seem to go crazy when the jerks occur—the O2 readings go all over the place. I thought that it was due to hyper activeness on the part of the EGO corrections, but I've slowed those down considerably, along with turned off EGO correction altogether while I'm driving, and I still get these oscillations.
Source: Jkav
Bucking can also be attributed to aggressive accel/decel enrichment sensitivity/GAMMA settings. The BOV "trick" may just mask these by introducing additional air to damp out the fueling.
Source: Boost
I once had the same problem on a
Mercedes with a turbocharged 2.3. It turned out to be improper ground to the
TECII unit. It grounded through the serial cable. I made an extra ground cable
and hooked it to one of the screws that holds the coil, and the symptoms were
gone.
Source: Shiv
The only critical sensor necessary for spark is the magnetic pick-up. If (and only if) you wired your stock CE light up to the TEC-II, you can diagnose a pick-up failure by seeing if the CE light turns off during cranking. If it doesn't, the magnetic pick-up is dead and needs to be replaced. If you did not wire the CE light up to the TEC, you can check the RPM value in the engine monitor screen during cranking. If it stays at zero during cranking, the pick-up is gone.
The only other reason I can think of that would cause your car to not spark is if the TEC-II has lost its switched 12v battery input (SW BAT). If it did, it will not turn on despite full battery voltage. To test this, you can hook SW BAT directly to constant alternator/battery 12v. If the car starts, then your regular SW BAT source is flaky. If the problem still persists, the magnetic pick-up is likely to be the culprit.
Source: Shiv
Check to make sure the wires in your connectors are making proper metal on metal contact. In the past, some folks haven't stripped enough insulation of the ends. Either that, or they have bend the bare ware back on the insulation. Either will lead to inconsistent contact and would explain why plugging and unplugging the connectors got the car started again. Also check your SW BAT splice. If it was poorly spliced, the TEC will not get the power to switch ON. Car won't start. I recommending using a butt connector and shrink wrap instead of the splice. This will make for a more reliable connection.
Source: RyanC
I don't think the map gets lost; when I installed my used TEC it hadn't been in a car in 3 weeks and the map was still there. Like Tom said, check the plugs. If that fails, you may have to selectively check each wire end for corrosion/looseness. I couldn't start my car on two occasions. One time the wiring was all gunked up from the winter, the other time my SWBAT hookup (at the OEM coil pack wiring harness end, not the TEC end) had jarred loose. I flipped open the lid of the connector, wiggled it (just a little bit), and snapped it back down. Car started right up (after I spent 2 hours redoing all the wire ends before trying this).
Source: Joel Gat
More details of what you see under the hood would be helpful, but knowing nothing else, I'd suspect a blown line somewhere. Check all oil lines to and from the engine, all vacuum lines, PCV line (this one, because if it's blown off, oil will get everywhere -- then figure out why it's blown off), etc.
Source: Jude DeMeis
You want to put a socket wrench on the crank pulley bolt and see if the engine turns smoothly clockwise. Also, pull the front covers and inspect the timing belt. If the engine spins freely (i.e. is not seized), the next step would be to disconnect the fuel pump fuse and do a compression test of each cylinder using the starter motor. If the engine is seized or the compression is hopelessly low on any of the cylinders, then the motor has to come out.
Remove all four spark plugs and try to turn the engine. Since this defeats any compression, the engine should then be very easy to turn by hand with a socket on the crank. Verify that it rotates smoothly.
Source: Joel Gat
When you try to turn the engine by turning the crank pulley, make sure the car is not in gear! Remove the plugs and try again. Also, if that's no good, remove the starter. I know it's kind of a pain in the butt, but that's something I learned from a junkyard mechanic. Sometimes the starter gear pops off and gets stuck between the end of the starter housing and the flywheel. It makes the flywheel impossible to turn (or close to it because to turn it requires bending the starter's shaft that normally goes into that gear) and the starter makes a whining sound because it's trying to turn but either can't, or is, but is just spinning against the now knocked off gear.
And as a last resort, dial 1-800-638-5773, Subaru Suzuki Kia Auto Recycling. With a core exchange, they'll set you up with a new 2.5 engine for under $1500, I think (depending on how the market is right now).
Source: TEC Manual
Coolant Temp (CLT): GM 25036979
Manifold Air Temp (MAT): GM 25036751
Heated Exhaust Gas (EGO): GM 8990741
Knock (KNK): GM 1997562
Source: SFSUBE
Gather
round the engine block my friends, and hear my sad tale…(happy ending included)
Genesis:
Sunday
morning broke crisply and cleanly in Burlingame. 8:00am sharp, cool and sunny,
with a mixture of scootering kids and bouncing hot-moms blazing past my eyes,
which were glazed with anticipation. Gathered around the gleaming white 99 on
the small suburban driveway lay my humble collection of tools. With Tec-II in
one hand and a freshly minted printout of the install instructions in the
other, I looked skyward, to summon Vishnu to my engine compartment! Come on
down ye beast, mark this combustion chamber with thy tell-tale criss-crossing
of glorious flex-loomed management magic!
With
colas and milk chocolate to fortify my shaking newbie hands, the first day
passed like a desert mirage, with a lot of wrenching about, and slow,
pathetically slow progress. I replaced he plugs, installed the coolant line
sensor and new injectors, did the basic Tec-II pick-up install, and the MAP
sensor. My friends, generous enough to offer their home and driveway for
Sunday, re-enlisted essentially without choice for day two,…visions of
real-time RPMs and air to fuel ratio numbers hounded my sleep.
On Monday Vishnu gave me clear skies and renewed confidence that I could complete the install in a reasonable two days, even with what seemed like dozens of missions to the auto-parts store for things I needed. Everything seemed ok so far. In fact, this install was truly a learning experience, a fast survey of an engine I’m completely unfamiliar with, and very enjoyable and educational. In went the O2 sensor, in went the Throttle Position Sensor, I was on install fire baby!
But
then the tale turns dark my fellow tuners, seekers of TEC-II knowledge, and
general reprobates, as a heretofore unknown part of my very own personality
crawled forth like some creature from Poe’s darkened byways. One inclined to
the frothings of Freud may call it my Car Ego, others may say it was a product
of the day long sun, others would be heard saying “What a stupid, bloody fool!
He should be fed to the King's dogs! Where are my coconut halves” Either way,
here is what I did…
Upon
testing the install of the pulley, when finalizing the pick up position, for
some reason I became utterly convinced the custom made toothed portion was off
by 90 degrees, certainly a disaster in the making! I called Shiv, who admitted
that it was possible the toothed plate was mismounted, but it definitely didn’t
sound right, in fact it was pretty dang unlikely. Could I please call him
before the final pulley install so he can walk me through it if I thought there
was anything wrong? Of course I said, since it was obviously a good idea.
At
that point, I decided to modify arguably the most critical component of the
system, with no experience and no finesse. I disassembled the pulley, hex bolt
by hex bolt, stripping the hex sockets. I switched the plate 90 degrees, and
threadlocked the bolts back in. Then I installed he pulley without calling
Shiv. Then I methodically wrapped everything in flex-loom. Then I called Shiv
to start the car. The car wouldn’t start. Phone consultations. A methodical 2
hour check of voltages, connections, everything, painfully re-tracing it all.
hen it really gets good. In a huge show of support, Shiv drives all the way to Burlingame
at 8:30PM to help rescue the only Tec-II install that didn’t essentially start
right up. “Why on earth won’t it start,” pondered Shiv out loud. We check
everything, a blast I can assure you after doing it already twice. Everything
checks. And then reality hits me in the head like a balloon filled with sand.
“Dude, what did you do to the pulley?! Your off Top Dead Center by…looks like
90 degrees” My world crashed down around my spotlit Subaru. Oh god. Out comes
the pulley and indeed, my on the fly modifications were the culprit, the pulley
teeth are rotated way off. (I’m an expert, yeah I’m an expert and should change
this. Why keep it the way Shiv and the rest of the Tec-II users like it when we
can do it my way? Should I mention this to Shiv? Why do that? I'm right!
lol)
With
the problem solved and with 2 broken hex bolts in one hand, and a thrice
removed pulley in he other, I watched Shiv drive off into the suburbs, after
bidding me with a farewell Unacceptable.”
Do
not take the pulley apart.
Postscript:
Once
the pulley was back together the car fired up first time. Shiv was always a
good sport about this, in the way one might be around an especially dim child.
I'm checking my water for mercury
later
all, angus