E38 LS7 Idle AFR drifting

[billybrun]

Hey all,

I have been trying to get the idle tuned up. It's an LS7 with a mild cam (11 degrees of overlap), worked and milled heads (11.8 SCR), I/H. The idle speed and manners are right where I want them, but I can't get the AFRs to stay steady thru warmup. Right now starts fine and AFR follow commanded, but gradually gets richer. I can bring the VVE higher to match the warmed up AFR in the idle area, but them it will start too lean. Any help would be nice.

Here... COS in progress_0008.tun Log_0094.efi

Thanks.

[swingtan]

Without even looking at the log....

You should set the VVE only when the engine is at normal operating temp (assuming SD tune). This is the base line for all other adjustments. I have no idea why you would tune it for cold idle then try and correct for normal operating temperature. All base settings should be corrected to suit where the engine spends most of its time. I don't know too many circumstances where an engine spends most of its time in the warm up phase.

Once the VVE is set for normal operation, use the ECT and INVLVTEMP modifiers to correct the warm-up phases.

Simon

[billybrun]

Thanks Simon, I have been following that method, and when I set the VVE where it should be for normal temperature it leans the the few minutes after start and eventually settles to correct AFR. When you say ECT and INVLVTEMP, Are you referring to B1602 and B0146? B0146 is set to 1.0 above 80c.

I have a lot to learn, thanks.

[billybrun]

swingtan, I found and old thread of yours which lead me to the right answer. I fixed my lean starts by making B0141 an equal AFR across the table, then adjust B0146 to make the actual AFR a strait line all the way from start until normal ECT, now I can control the actual AFR from start-up to normal ECT using B0141. Thank you.

[Yasei]

Is there a way (a correction factor) to get the cold start actual AFR closer to the commanded AFR or do you just need to command a more richer AFR? I have around 2 AFR difference between the commanded and actual AFRs during a cold start. VE is pretty much spot on in these areas when the engine is warmed up. I'm just commanding a lot richer open loop AFR at lower ECTs/INVLTEMPs to get a AFR I want. B1602 perhaps?

[billybrun]

B1602 didn't work for me, but B0146 did. That way "commanded" does not follow "actual" during warmup; "actual" follows exactly what you want in B0141.

[swingtan]

I meant to answer this, but got side tracked....

To get the actual measured exhaust AFR to follow the commanded, you need to have the fueling tables accurate. So far the mentioned tables won't actually fix and differences between the commanded and the actual.

B1602: This is just the correction to the commanded airflow during warm up. It has no effect on actual fueling.

B0141 through B0148: These all deal with the commanded air fuel ratio and have no impact on any differences between the commended and measured AFR.

Getting warm up fueling spot on will come down to how pedantic you want to be and more importantly, how much time you have to spend on it. Most shops will not be able to justify significant time and multiple dead cold starts to get the fueling dialed in. I've done it, on my own car where I can and do spend significant amounts of time testing different settings.

Basically, you'll need to fix the normal running temp fueling first, as LTFT's are applied to the cold running OL fueling. Then you need to get the dynamics correct for the warm up cycle. Depending on the local climate, the tuning may be easier or harder, but if you get the fueling close enough and feel it is safe to do so, you can disable the LTFT's and clear the learnt data. You want very accurate fueling though....

Here's a cold start where the ambient temp was 16'C.



And here's one where it was 0'C




Note the bottom chart where both commanded and measured Lambda are plotted together. The goal is to get the fueling correct under all conditions, then just command the required fueling. Oh, and these are both with E85, hence I'm using Lambda and not AFR.

Simon.

[Yasei]

Ok, I've already tuned the normal running temp fueling. The LTFTs stay within +-3%. I just don't seem to be able to determine which tables are the ones to adjust for warm-up fuel. I'm able to richen the warmup cycle fueling, but don't know how to zero out the difference between commanded and actual. This way I need to command very rich fueling during the warmup to avoid lean condition and idle fluctuation. Doing this on my own car and have all the time so I'd like to make it perfect I'd really appreciate guidance on this!

BTW, I'm using the "double-stoich-half-ifr"-mod in this tune. That has an effect on the intake valve temp and thus an effect on the fueling as well, doesn't it?

[Yasei]

So here's the way it is now. This is with unmodified commanded AFR.



The engine wasn't completely cooled down, 6deg IAT and 16deg ECT. Anyway, between 50-70deg intake valve temps there is a huge difference between commanded and actual AFR, up to almost 4AFR. After about 115deg intake valve temp, the difference has pretty much zeroed out (still in open loop and LTFTs zero all the way). I want to make it perfect

The car is a supercharged/cammed/etc LS7. ID1000 injectors with correct data, double stoich AFR half IFR -method.

[Yasei]

Still haven't been able to figure this one out. Seems to be related to the double-stoich-half-IFR -method. HPT got a fix as they're able to affect the INVLTEMP-calculation. Any chance we could get it for EFILive as well?