Low SELBEN's with Calc.VET?

[ferocity02]

I'm working on tuning my turbo '01 Silverado (12212156 OS w/ MAF and 1bar MAP) and decided to try Calc.VET in order to work on the MAF and VE tables when out of boost. At near-zero TP the fuel trims were maxing out negative (-12.5%), everywhere else they were -8% to 0%.

This was giving me SELBEN's down to 88% in just a couple of the MAF FREQ rows around 3375Hz. The rest of the rows were much closer to 98-99%, and the curve is still pretty smooth, no dips or discontinuities. Does this seem right?

I started with a stock MAF map for my truck and the MAF sensor has been moved to the cold piping. It's at the end of a ~14" straight section with a 90 deg bend right after it. The BOV is about 8" before the MAF. I also ditched the PCV system, so perhaps that would throw off the MAF calibration slightly too.

[joecar]

Is DFCO enabling...?

[ferocity02]

Is DFCO enabling...?

Nope, disabled it.

[joecar]

Post log file.

[joecar]

The MAF physics means it has a x^4 transfer function (meaning it is curved), so it can't have any steps in it.

[ferocity02]

Log attached and a screenshot of the MAF table after applying the SELBEN's from the log.

[ferocity02]

I flashed the adjusted MAF map back in and datalogged again, and the trims still went max negative on decel. So I adjusted the MAF map once again, and they are STILL going max negative.....

Can a stock MAF map be that far off?

With the new MAF table the driveablity was worse too, it would buck in decel.

[ferocity02]

Ok... after 3 rounds of logging for Calc.VET, my MAF curve now has a dip in it at 3250 Hz, as in the values to the right and left of this point are higher. The curve is no longer monotonically increasing. I assume this is incorrect?

The only thing I can think of is that the higher vacuum during decel is causing something to happen making for a rich mixture. The trims at idle are fine, and these have a lower MAFREQ than decel. I pulled the line to the BOV so it always stays closed and got the same results. There are no vacuum leaks (these would cause a lean mixture anyways). No leaks after the MAF.

Perhaps I entered my injector data incorrectly?

[joecar]

Are the trims close to zero now...?

Did you apply the transient filter...?

[ferocity02]

Are the trims close to zero now...?

Did you apply the transient filter...?

I haven't tried the newest MAF curve with the dip because I assumed that this behavior was totally incorrect. I can this morning.

My filters remove data with SAE.ECT below 165F or above 220F, remove data with a change in SAE.TP greater than 5% in 100 ms, and remove data with SAE.MAP above 102 kPa (to remove possible boost data).

[ferocity02]

Fuel trims still going max negative on decel... I think something else is wrong here.

One thing I've noticed, is that lets say I'm coasting down a hill in gear, so my RPM and MAFFREQ are constant. The AFR starts off at ~12 with the fuel trims at max negative. Slowly but surely, the AFR climbs back up to stoich despite the fuel trims staying max negative and there being no change in RPM or MAFFREQ.

The log shows a good example of this starting at frame 10400 going to 10600. Here I was coasting down a hill, MAFFREQ, MAP, RPM sand fuel trims are all relatively constant, but the AFR goes from ~12 to stoich over about 15 seconds... Kinda seems like the fueling is lagging behind or something... Or maybe it has something to do with the turbo spooling down?

[ferocity02]

I put the stock PCV system back on to see if that made any difference and it didn't, fuel trims still going max negative on decel, and are pretty good everywhere else. Before the turbo it would go lean during decel but the fuel trims would stay around zero...

[joecar]

Either: cylinder airmass is being added incorrectly (which then adds fuel), or fuel is being added...