Can someone clear up some STFT/LTFT confusion

[stigmundfreud]

Its been 3 years since I really last looked at tuning, since then I've upgraded to an e38 system and toying with just leaving things as is out of the box (other than some compensating PE mode commanded fuel as setup on the dyno).

Currently I have no widebands in the car and no plans to do anything in the short term so for now just logging the status quo. My old LC's are both probably dead now and were as unreliable as a stag party in a brothel

The confusion I am hitting is STFT vs LTFT. If the average cell in the STFT map is showing it pulling fuel why does the LTFT in the same cell appear to be adding in virtually/approximately the same amount as being pulled?

Further when I should be entering PE mode it seems the tune holds on to closed loop. Correct me if wrong but when in PE shouldn't it change to OL-Drive? It does hit OL-Drive but not as instant as I'd expect.

So for a stock tune from the factory it seems to run lean according to my LTFT's but STFT's look pretty good. Of course this is logging with MAF, CL, LTFT's enabled so everything that can interfere is switched on. I just find it hard to believe out of the box VE is causing ltft to add in fuel - is this MAF scaling perhaps? I'm adversed to getting to the VE tuning stage on this one as the powers enough as is and so what if its not optimal. However, I'm altering the exhaust next week so if I'm in positive LTFT now and with better exhaust cfm rates I dont want to be relying too much on trims to keep things right.

tune = http://www.hermantoothrot.co.uk/car_0001.tun
log = http://www.hermantoothrot.co.uk/scan1.efi

[joecar]

Hi Mike,

When the STFT hits the B4115/B4116 boundaries then they get reset and the LTFT gets updated...

i.e. when the STFT tries to exceed the boundaries it spills onto the LTFT.

[joecar]

I'll look at youe files later tonite...

[stigmundfreud]

cheers joe! Those tables you quote cannot find in my tune.. Its just kind of not expected to see positive ltft's as I'd expect them to err on side of caution for the tune. Obviously the trims are adjusting the fuel but to be putting fuel in, I'd expect to see it attempt to pull out but then that could be my picky mind. What does worry me is if I do anything to improve airflow, even slightly then it would be relying on trims far too much for my liking.

Maybe the easiest way is to scale the maf by using the crude method of using the trims?

[joecar]

Mike,

I just now had a chance to to look at your files, I didn't realize you had an E38/E67... virtual VE table... I'm not practiced on E38/E67 (I haven't played with any).

In your calibration file I don't see the STFT max/min correction parameters... but it works the same way: the STFT moves up/down in some manner, when it hits a +/- threshold it spills over into the LTFT, i.e. the LTFT is bumped and the STFT is reset to zero...

your STFT/LTFT look like they are working as intended... your HO2Sx1 voltages are switching correctly and sufficiently frequent...

You can scale the MAF table using LTFT's... of you can make PE richer (see next paragraph).

Your HO2Sx1 voltages indicate you might be a little lean at load... your PE B3618 table is a little leaner than I would like... I would set it to EQR 1.16 thru-out the range, and if you want you can set it to 1.12 from after peak TQ to redline... (i.e. below peak TQ you want to be sufficiently rich, after peak TQ you can lean it a just little to make higher peak HP... but thru-out you have to be sufficiently rich to avoid sharply raising combustion chamber temperature).

I see some knock, I can't tell if it's real, but see what I said about PE.

Now, having said all that, a wideband will come in very useful, this will help to show if the MAF really needs rescaling.

Have you seen this: http://forum.efilive.com/showthread.php?t=8961

[stigmundfreud]

cheers Joe! I've leaned the PE out to those figures on the dyno to account for the really rich condition it otherwise hits on standard settings. On the dyno the PE tracks true at around 12.95 using those figures. The knock I was getting sarah to read out when on the road yesterday and it mostly is only occuring for an initial part at throttle tip in then settles back down around 0.2. Unfortuantely didnt grab cylair on that run so will relog with cylair for more info on cells and load but it was typically on throttle tip in from constant to WOT, nothing real at high rpm.

Agree on the wideband but not wanting to get into too much changes on this one, the exhaust is about the only mod and over the old LS1 shipped on the 5.7 we got here, the tune it seems to be pretty efficient.

Thanks for taking a look, I might try some crude maf rescaling but its been so long (and I know you are the oracle on the forum for thread locations...) dont suppose you have a guide to MAF scale via trims? I cant find anything

[joecar]

Ok, I see about PE.

For MAF scaling using LTFT's, see this: showthread.php?p=120121#post120121 --> uses LTFT's to correct/scale MAF.

Use the MAF portion... I don't know how/if you would apply the VE portion to E38/E67 VVE.

[gmh308]

Its been 3 years since I really last looked at tuning, since then I've upgraded to an e38 system and toying with just leaving things as is out of the box (other than some compensating PE mode commanded fuel as setup on the dyno).

Currently I have no widebands in the car and no plans to do anything in the short term so for now just logging the status quo. My old LC's are both probably dead now and were as unreliable as a stag party in a brothel

The confusion I am hitting is STFT vs LTFT. If the average cell in the STFT map is showing it pulling fuel why does the LTFT in the same cell appear to be adding in virtually/approximately the same amount as being pulled?

Further when I should be entering PE mode it seems the tune holds on to closed loop. Correct me if wrong but when in PE shouldn't it change to OL-Drive? It does hit OL-Drive but not as instant as I'd expect.

So for a stock tune from the factory it seems to run lean according to my LTFT's but STFT's look pretty good. Of course this is logging with MAF, CL, LTFT's enabled so everything that can interfere is switched on. I just find it hard to believe out of the box VE is causing ltft to add in fuel - is this MAF scaling perhaps? I'm adversed to getting to the VE tuning stage on this one as the powers enough as is and so what if its not optimal. However, I'm altering the exhaust next week so if I'm in positive LTFT now and with better exhaust cfm rates I dont want to be relying too much on trims to keep things right.

tune = http://www.hermantoothrot.co.uk/car_0001.tun
log = http://www.hermantoothrot.co.uk/scan1.efi

If you are looking at LTFT and STFT in cells in a histogram you will most likely get this kind of result where they seem to oppose each other. The relationship is best viewed on the graphical display.

You will see STFT's run a sinusoidal curve of sorts as they go rich/lean and back again.

Period is around 2 x per 3 seconds at idle and maybe 10 x per second under cruise. So the STFT's once LTFT settles will generally show both + and - results on the histo but if they are correcting they may be more weighted to + or to -.

As Joecar notes the STFT's push the LTFT's so the STFT correction required is minimal and probably weighted around a moving average of zero. An E38 gets this done very quickly (less than a minute or two) under steady state running in a cell.

The VE has no effect while running in steady state closed loop with MAF.

The MAF calibration and fuel type will tend to affect LTFT the most if all else is stock.

[swingtan]

A few more tips....

1. The E38 is not like the LS1 PCM. While general goals are the same, how you get there is a bit different.
2. You need to look at specific E39 PIDs that replace the LS1 PIDs.
   for example...
   [insert]
   
   • Log GM.ETCTP% rather than GM.TP%. the former gives the actual measured throttle opening while the later gives the % of the 5v signal applied to the electronic throttle. Generally, around 88% of 5V gives 100% throttle opening so the latter only gets to around 88% peak.
   • The same goes for E38.AFRATIO_DMA and GM.APCYL_DMA
   [/indent]
3. The E38 has some internal TP adjustment tables that we don't see as well. check this link with my comments on PE activation and throttle position. http://forum.efilive.com/showthread.php?t=6461. So in your tune, a TPS setting of 40% is really more like 65% and 20% in the tune is more like 52% when you log ETCTP.
4. Another tip on PE. The Gen IV seems to like a slightly richer PE than the LS1 did. So as Joe says, it might be an idea to add some more fuel in there.
5. Finally, if you set up the VVE and MAF then you can turn off LTFT and just run with STFT. Then you don't need to worry about any differences. I think you might find that the with changes in running conditions, the STFT will regularly be different to the LTFT. This is because as the LTFT's are learnt, they will always lag the correction of the STFT. So at one point, the STFT may be adding fuel, slowly increasing the values of the LTFTs. But then, conditions might change ( different fuel, change in weather, change in altitude, etc. ) and the STFT might want to start removing fuel. At this point, the LTFTs will still be trying to add fuel, as this is what was learnt previously.

Simon.

[magic919]

I see you've got B0157 quite low to trigger PE mode. However, You might want to pull B0156 down too and then it'll switch in at lower throttle area. Mine is at 4 throughout.