4,000 RPM misfire, low load

Hello All!

I haven't been active lately, or even been diligent with my tuning studies but Im certainly glad to be back! Done some changes to the car over the last few months, so tuning has been at a standstill. One problem Im having has me and a couple knowledgeable buddies stumped. Just added some ID2000's injectors to the STI to combat my lack of fuel with ethanol. Im currently having some very odd issues with the tune. At low MAP regions on the VE table, roughly 8-10 MAP(PSi) around 4,000rpm, the afr leans completely out and misfires. Once passed this 4,000rpm zone, even at the same MAP, misfire disappears. For instance, when driving off from a stop, slowly going through the gears to approach speed limit, I experience the misfire everytime at 4,000. If I were to go WOT in second gear around 2,000rpm, I never touch the low MAP of 10 and avoid the lean misfire. I have a properly calibrated AEM 3.5bar MAP sensor, just as an fyi. So it's definitely low load dependent, but why? Fuel pressure is good. Plugs are good. Injectors are new and properly calibrated. Ive attached a datalog to show how things are pretty linear when it happens, no crazy dip in IDC, VE, Igntition, etc

Left out some very key components to my issue. No matter how much fuel added in that region of the VE map, it still misfires and leans out.

I faced a problem like that recently. Turns out I had overrun fuel cut enabled, and the trigger RPM was exactly where I got the misfire (when the fuel cut with only 2% throttle). Perhaps your issue is similar.

What car and ECU had the problem?

Nothing?! Come on guru's! I can't do this alone!

Hi Stu, this isn't an issue I've personally experienced on a Subaru. Can you give us a little more information about what you're tuning with, and when did this problem develop? ie was the car running properly prior to changing to the ID2000?

The car is being tuned SD via COBB ACCESSPORT. Problem developed after replacing some 1,400cc injectors with these 2000's

Ok I'm unfortunately not familiar with the SD tuning in COBB so I'm not sure how much help I can offer. What you're describing sounds very much like the problems you're likely to see in a MAF based tune on the Subaru if your Load Compensation table isn't tuned correctly. Typically this happens a little lower in the rpm though - Perhaps 2500-3000. I have no idea what COBB does with this table in an SD tune but it might be worth considering?

It's very odd and very frustrating. There's nothing in the recorded data to indicate any issues. I'll check the load compensation table. I really dislike the hundred tables that make up the functionality of the factory ecu. Doesn't make any sense!!!

I understand your pain. A factory ECU often makes an aftermarket ECU look like a walk in the park. If you're only making minor changes to the mapping then there is little you need to do and this makes the tuning on an OE platform really easy. When you're making significant changes though and swapping to something like an SD tune with large injectors then there is often a lot more work to be done.

Unfortunately, I cannot find the load compensation table, maybe due to the sea of tables. The 3 speed density compensation tables available are coolant temp, intake temp and barometric pressure. This is becoming a major headache.

I'm sure you checked this already but what is your SD Mode set to? 0= MAF, 1=Speed Density, 2=Hybrid. I was thinking that if you have the Hybrid Mode selected that you might be getting interference from the Closed Loop Circuit in the Mass Airflow Set-up because your MAF Curve isn't tuned. I'd check your A/F Corrections for that Load Range and see if your getting any "adders". Let us know what you figure out.

Im not familiar with your tuningplatform at all so bare with me.

Are you able to log things like injector pulsewith? if you are then it should show up wether it is lean because less fuel suddenly comes out of the injectors, or if it actually and litteraly goes down the drain and into the exhaust due to freak contitions that just blows the charge past the combustionchamber. Maybe the cams are doing something funny at that point?

Also The "Load Compensation Tables for TGV Open/Closed" are under the "Sensor Calibration" tab in the ATR SD Software.

Setting is 1, or SD mode. There is no AF correction or learning because I have forced open loop.

The datalog I posted shows pulsewidth during the affected area. I've ran these cams and engjne on 2 other sets of injectors, 93 octance and E85, without an issue. This is a gremlin of a problem

Hey Stu,

I know you said your Injectors were properly calibrated but is it possible that you don't have the correct "Fuel Injector Trim (Small IPW) Data"? If you have the correct information from ID, which fuel are you having the issue tuning right now E85 or 91? It would make sense for petrol fuel to be more susceptible to inaccuracies in the injector characterization, with regards to Small IPW Trims, especially with such large injectors and such a low fuel flow requirement (Low Load). If your trying to tune High Flow injectors with a low flow requirement and be accurate, I think proper Small IPW Data is critical. Have you tried adjusting the "Fuel Injector Trim (Small IPW)(Max RPM)" value at all? The stock value is set coincidently to 4000RPM, so at this speed the ECU will no longer apply the Small IPW Trims to the injectors because the ECU assumes that the Injector is now in it's "linear operating range" and there is a little more predictability to the injector flow rates. So raising this value to 4100rpm could allow the ECU to correct for the "Lean Misfire" by continuing to apply the Small IPW Fuel Trims to the injectors and allow you to reach your target AFR. Conversely decreasing this value to 3900rpm for instance, might allow you to readjust your main Fuel Table to achieve the desired AFR at that Load Value. I'd prefer to allow the Injector Trims to do their job, instead of trying to rob Peter to pay Paul. Just thinking out loud here.

Car is solely being tuned on E85 only. ID provides tried and true plug n play data for the Cobb AP and the ID2000's. The max rpm limit for small ipw is raised to 10,000rpm, per ID's recommendations. I honestly do not think this is an output sensor type calibration error. Forcing open loop in this situation is MORE beneficial. There is no correction, no compensation. Closed loop would probably make this worse.