E40 Car bogs at 3000 RPM, airmass calculation spike

[ACCLR8N]

2005 LS2 6spd GTO single front mount turbo using the custom boost OS.

The issue I'm having is under moderate throttle conditions (35%) there is a spike in fuel to the engine. In the highlighted section of the log you can see the step at 3200 RPM from one frame to the next. Commanded AFR is still at 14.7, short and long fuel trims are stable through the event. The only DTC is P0103 which is expected as I have the MAF failed at 10Hz for this round of tuning. You can see the MAP, MAF, Throttle position are all stable. In my VE table this is a sharply decending area of the graph, but in reviewing the log there is a large jump up in calculated airmass.

Is there some RPM based protection scaler somewhere that I missed?

[joecar]

Do you have a way of logging VE to see what the ECM thinks VE is when DYNAIR and AIRPERSEC step up...?

Is IAT staying constant...?

What happens if you eliminate/smooth the sharp down change in the VE table...?

[ACCLR8N]

Yes I can log SCALEDCORVE_DMA, but I can't play it back in the scan tool. I will have to try and watch it.

I reviewed other logs and the IAT is constant within ±1° C

I can compensate by putting a deep negative trench in my VE table. This is what AutoVE, Calc.VET and GELF VXR's methods all do. Running up my VE table along the 3200 & 3600RPM columns. Attached is an earlier unsmoothed result that eliminated the driveability symptom, but not the problem.

[joecar]

Hmmm... there must be some other airmass contribution from some other table (AIRPERSEC and DYNAIR indicate this extra contribution).

[kangsta]

reduce the VE in the 2800 col and take another log

[ACCLR8N]

I reduced the 2800RPM column 5%. It didn't get any worse, but I'm not sure it was better. It is still very evident in the log.

[kangsta]

Can you add E40.SCALEDCORVE_DMA and GM.APCYL_E40_DMA to the log?

[ACCLR8N]

E40.SCALEDCORVE_DMA was logged. It doesn't show up in V7.5 Scan tool because of naming issue in the hexadecimal file. I'm told it can be reviewed in V8 or renamed in the log file. I will add APCYL.

[kangsta]

id probably flatten out that whole area between 2800 and 3600 rpm to make sure its not an interpolation thing and then tweak in the VE.

[joecar]

The log file has to be edited to change the name of the SCALEDCORVE pid.

[c.u]

You may not want to log the MAF at this time. I`v seen it skew reading when tuning sd.

[ACCLR8N]

Gave it a try at flattening that area out quite a bit. It didn't make any difference. Looking a frames 2580-2590 of the attached log for example still show the big step up in calculated airmass. The suggested PIDs and fuel states have been logged but I haven't figured out how to rename the SCALEDCORVE yet to review.

Any other ideas on what could do this. I combed back through my tune table by table looking for something related to 3200 RPM and nothing stood out.

[ACCLR8N]

You may not want to log the MAF at this time. I`v seen it skew reading when tuning sd.

Just don't log the MAF freq and g/s? Or should I fail it in some other manner? At the moment I have it failing the rationality test above 10 Hz.

[c.u]

Dont`t log the MAF and g/s.

[joecar]

If you get a MAF DTC then the MAF is being ignored.

[ACCLR8N]

id probably flatten out that whole area between 2800 and 3600 rpm to make sure its not an interpolation thing and then tweak in the VE.

That was the issue. The rise (hump) in my VE table preceding the bogg was the problem. I have not identified the root error that created this hump, but getting rid of it by hand solved this issue.

[joecar]

In retrospect, that hump in your VE tables looks wrong...

VE table should look like this: each MAP slice should look like a dyno TQ curve.

[kangsta]

That was the issue. The rise (hump) in my VE table preceding the bogg was the problem. I have not identified the root error that created this hump, but getting rid of it by hand solved this issue.

Strange VE shapes are common to "Auto-VE" methods due to lag in the WB reading and fast changing RPM/MAP readings so the area you are correcting isnt really the are that needs correction. This shows up more on high-resolution VE tables.

[ACCLR8N]

Thanks for your input. I have been reviewing the trends and sculpting the table by hand. I have my WB mounted 18" back from turbo exit per Innovate's recommendations and it is clearly at least a frame behind in the logs.

Temperature has been a factor. The first 20-30 minutes of drive time is to be ignored. The car is operating leaner once everything is thoroughly warmed up.

Still, all three auto VE methods seemed to be off by a factor that continues to elude me. Every time I would hit a wall at 6.3% LTFT and then it would jump the other way. I'd like to hear from somebody that is running a boosted E40 and got one of the auto methods to work. I can see why you would want a steady state dyno.