Why negative fuel trims after autoVE

[Gelf VXR]

Ive been thinking about the general advice for LTFT after AutoVE of the VE table, LTFT's should be -1 to -4%, because if positive extra fuel will be added in PE mode.

LTFT's only add fuel in PE, they do not subtract.

Which makes me think if the LTFT's are negative and the adjustment isn't made in PE mode then AFR will be rich than the commanded PE AFR anyway, by not compensating for the LTFT -% value. Which kind of defeats the reasoning behind not having positive LTFT's.

How ever, if you have positive fuel trims, fueling is corrected in CL and fuel also added and thus compensating when in PE mode which should be to the commanded PE AFR?

I also not that if the VE table is not as close to BEN = 1.0, then the predetermined (stock) spark value will not be at optimum. If you add to the VE table so that LTFT's are negative, the calculated g/cyl value is higher than the true value and thus a lower spark value is looked up.

Like wise if the VE table is lower and LTFT's are positve, the g/cyl value is lower than the true value and the spark value looked up will be higher. However, I think with dynamic fueling the PCM will go with the MAF g/cyl value and not the VE table in this circumstance.


Any thoughts?

[WeathermanShawn]

LTFT's only add fuel in PE, they do not subtract.

Which makes me think if the LTFT's are negative and the adjustment isn't made in PE mode then AFR will be rich than the commanded PE AFR anyway, by not compensating for the LTFT -% value.
How ever, if you have positive fuel trims, fueling is corrected in CL and fuel also added and thus compensating when in PE mode which should be to the commanded PE AFR?

If you add to the VE table so that LTFT's are negative, the calculated g/cyl value is higher than the true value and thus a lower spark value is looked up.

Like wise if the VE table is lower and LTFT's are positve, the g/cyl value is lower than the true value and the spark value looked up will be higher. However, I think with dynamic fueling the PCM will go with the MAF g/cyl value and not the VE table in this circumstance.


Any thoughts?

I personally think you are correct in all your points. I try to get my LTFTS near zero. The problem is it makes WOT fuel computation difficult with + fuel being added. But, you are right. You have to then manipulate the airmass calculation in PE Mode to get the right WOT Commanded Fuel. Then to further complicate matters, if you hit PE in a lower RPM, chances are (if MAF-enabled) the MAF Freq (g/s) value was adjusted for a negative trim..now your too rich for WOT.

You are also correct that when any manipulation of the airmass (VE or MAF) alters the load computation for spark. In just the manner you described.

It would be a nice option to be able to just adjust the Trim for specific FTC cells. You might be able to eliminate that problem. Or if MAF tuned, a PE Threshold that operates by MAF Freq instead of MAP.

I would just try to get your LTFTS as close to zero..Then fueling and spark computation will be very close.

Good observations..:)

[mr.prick]

Always filter out positive LTFTs during PE. :bangin:

I agree with Shawn :) if LTFTs are screwed up after OL tunig what was the point?

[Gelf VXR]

Always filter out positive LTFTs during PE. :bangin:


I'm not saying your wrong, just challenging the accepted rule.


We know one bank runs leaner than the other, with negative trims on both banks, the difference is still preverlant under WOT & PE.

I'm thinkin that if you autove and aim to get your rich bank values 0% to -1% and the lean bank +2% to +3% or were ever they need to be, when you go into PE or WOT, THE PCM will add the xtra fueling to the lean bank only and balance the AFR in each bank.

[WeathermanShawn]

I'm not saying your wrong, just challenging the accepted rule.


We know one bank runs leaner than the other, with negative trims on both banks, the difference is still preverlant under WOT & PE.

I'm thinkin that if you autove and aim to get your rich bank values 0% to -1% and the lean bank +2% to +3% or were ever they need to be, when you go into PE or WOT, THE PCM will add the xtra fueling to the lean bank only and balance the AFR in each bank.

That would be an interesting log.

While I think I know where you are going on this..if your wideband is on the + LTFT bank..you can not get your Commanded AFR to match actual. Edit.. I do agree with you that entering PE with a negative value may not match either..

How do you get around that?

[Gelf VXR]

That would be an interesting log.

While I think I know where you are going on this..if your wideband is on the + LTFT bank..you can not get your Commanded AFR to match actual. Edit.. I do agree with you that entering PE with a negative value may not match either..

How do you get around that?

The problem is the MAF value is going to be an average of the values of the two banks.

If the VE map is dialed in to the rich bank as close to zero or slighty negative and the lean bank is positive, the MAF needs to read the average of the two banks which will be slightly higher than the dialed in VE value for the commanded AFR to match actual. Im not sure how to over come that?


Edit: a slight adjustment to the MAF calibration could overcome this offset?

[mr.prick]

Use an average of both banks for VE&MAF.

[WeathermanShawn]

Gelf:

I have run into a few more issues with a MAF closed-loop tune than a CLSD tune (in terms of what your are describing).

There are many, many points of MAF resolution vs MAP/RPM for airmass calculation (disregard interpolation). So in my case 7500-8000 MAF (Hz) may be used both in PE mode and non-PE mode. If you apply LTFTS at 7500-8000 Hz that change the calculated airmass..under a different RPM in PE Mode you may have an airmass calculation that is incorrect for PE Mode.

Again, whether CLSD or MAF/SD Closed-Loop..I try to aim for zero Trims prior to entering PE. Like Mr. Prick, I use an average of both banks..

[stigmundfreud]

+figures for ltft mean the car is adding fuel yes so tune is lean? So long since I did any of this that confusion is creeping in. If the LTFTben is 1.25 is that rich or lean? I know EQ 1.13 = 12.9 or there about so would indicate rich but is that before or after fueling is sorted. IE if you log 1.25 LTFTBen is that what the car is adding to the mix?

[joecar]

LTFTBEN 1.25 means that the LTFT's are adding 25% fuel to trim to stoichiometric.

(LTFTBEN is the BEN derived from the average of LTFT1 and LTFT2)

so the computed airmass (from either VE and/or MAF) is lower than actual, causing fuelmass (required to meed desired/commanded AFR) to be computed low... this causes NO2 to indicate lean, PCM determines how lean (by looking at how much time the NBO2 signal spends low) and determines LTFT correction which it then applies to the next iteration.

[stigmundfreud]

Cheers Joe, its coming back and that matched what I thought (of course ;)) about the 1.25 being extra being applied. Of course we have the wideband in now so will log purely on that (gave up on the idea of trimming with the ltft bens and waited). Should be a fun night assuming the replacement LC1 works

[joecar]

Sanity check the LC-1 by checking Lambda it returns while PCM is commanding CL.

[SSpdDmon]

My thoughts on LTFTs from 2 years ago - see post #6.

http://www.ls1tech.com/forums/pcm-diagnostics-tuning/890259-fuel-trims.html

Sounds like the OP has stumbled upon some similar questions. If I ever tune a car again to use the fuel trims, I think I might actually target a +5% across the board...not the -1% to -4% that I remember seeing everywhere I looked back in the day. :)

[mr.prick]

LTFTs are impacted by the dynamics tables as well.

After installing some new fuel injectors (ASA 38lb) with their "known" offsets,
LTFTs under decel were highly negative.

I reduced the impact & evap factors by X.XX% and got them inline.
This may have been needed because of the extra CI my engine has now. :nixweiss:
Otherwise the "known" injector settings may have worked.

High positive & high negative LTFTs can cause severely incorrect fueling.

[Gelf VXR]

I have found that slightly positive trims gave me some knock pretty evenly across the map, I'm at the limit of the knock threshold for g cyl. What this has showm me that the GM Hi octane map is the correct shape for my set up, for the solution I have moved the Hi octane map all one cell to the left. I will add back incremenatally until I see knock in any cells.
8555
8556
8557