VVE Tuning Closed Loop STFT versus AEM Wideband

[scottt28]

Happy Veterans Day

Question... I've been reading up and watching the forums on VVE tuning. Today, I put the truck in SD mode, and left the truck in closed loop (I know Ed M prefers open loop, and I intended to do it in open loop, but forgot to reset it with the scanner). The STFT errors were quite a bit different than the AEM wideband errors. Both sheets show conditional formatting - for errors > 1, values are in RED, for values < -1, values are in green. I never got more than about 42% TPS and never hit PE. Graphs have a 50 hit cell minimum...

STFT Errors 11.11.JPG AEM Errors 11.11.JPG

Could not upload the scan because it is too big. I look at the wideband numbers and feel pretty good. I look at the STFT numbers and wonder if I'll burn the damn thing up. Does anybody want to comment on the differences?

[schpenxel]

Was it in closed loop this entire log?

If that's the case then that's expected.

One big and common misunderstanding is what fuel trim values really represent

Fuel trim values are the % of fuel that is being added/subtracted so that it's NOT off. If you had say a fuel trim value of 10%, that doesn't mean it's 10% lean. It's mean it WOULD have been 10% lean if it weren't adding 10% more fuel to make it not lean. Because of this the wideband should show very little error in closed loop areas given enough data. That seems to be the case.. 4, 6, 8% positive fuel trims but AFR error is next to 0 for the most part, because fuel trims are doing their job...

So in closed loop the error on the wideband should be very close to 0 (with enough data, of course). That appears to be the case for the most part. Obviously the higher RPM / low MAP area is a bit further off for some reason but without a log I'm not really sure why..

[scottt28]

Was it in closed loop this entire log?

If that's the case then that's expected.

One big and common misunderstanding is what fuel trim values really represent

Fuel trim values are the % of fuel that is being added/subtracted so that it's NOT off. If you had say a fuel trim value of 10%, that doesn't mean it's 10% lean. It's mean it WOULD have been 10% lean if it weren't adding 10% more fuel to make it not lean. Because of this the wideband should show very little error in closed loop areas given enough data. That seems to be the case.. 4, 6, 8% positive fuel trims but AFR error is next to 0 for the most part, because fuel trims are doing their job...

So in closed loop the error on the wideband should be very close to 0 (with enough data, of course). That appears to be the case for the most part. Obviously the higher RPM / low MAP area is a bit further off for some reason but without a log I'm not really sure why..

Yes, it was in closed loop for the entire time.

I agree with you that it WOULD HAVE BEEN OFF 10% (for example) if the ECM did not add/subtract the fuel. That makes sense to me. Based on that assumption, it seems like the fueling is off 10%, otherwise the ECM wouldn't need to add/subtract fuel. And if the ECM is changing fuel 10%, I should probably be using that STFT error to tables. RIght?

Shouldn't the goal be to make the fuel trims as close to zero, and the wideband readings also as close to zero as possible?


Next question... the truck seems to "break up" at higher rpm. I stomp on it and it accelerates to a certain point, then it sounds like it hits a rev limiter and pops and sputters and doesn't accelerate further. Is this a sign of higher RPM / low MAP issue you mention? Is it too much advance?

[Jggregory99]

Yes, it was in closed loop for the entire time.

I agree with you that it WOULD HAVE BEEN OFF 10% (for example) if the ECM did not add/subtract the fuel. That makes sense to me. Based on that assumption, it seems like the fueling is off 10%, otherwise the ECM wouldn't need to add/subtract fuel. And if the ECM is changing fuel 10%, I should probably be using that STFT error to tables. RIght?

Shouldn't the goal be to make the fuel trims as close to zero, and the wideband readings also as close to zero as possible?


Next question... the truck seems to "break up" at higher rpm. I stomp on it and it accelerates to a certain point, then it sounds like it hits a rev limiter and pops and sputters and doesn't accelerate further. Is this a sign of higher RPM / low MAP issue you mention? Is it too much advance?

If you are in closed loop, you should NOT be using WB data AT ALL...In fact...Delete the channel for your WB in the scanner as to not get confused. You have to remember, in OL your err comes from the WB and not the Front O2 sensors, but in CL the O2 sensors are driving the fueling to stoich, so your WB is going to constantly show stoich because the O2's make it so... It's one or the other...never both... If you try to use WB data in CL it will constantly tell you to add fuel... I made this mistake once and ended up running a cruising afr of 10.8 because I used the WB data while in CL....

[BigDaddyCool]

You should really be tuning in open loop with wideband first...... as close loop is determined by many other variables and if they are wrong so will your fueling.

Use your "correct" wideband data to correct how the car functions in closed loop ........... assuming your wideband is setup and is fairly accurate.

[schpenxel]

Shouldn't the goal be to make the fuel trims as close to zero, and the wideband readings also as close to zero as possible?


Next question... the truck seems to "break up" at higher rpm. I stomp on it and it accelerates to a certain point, then it sounds like it hits a rev limiter and pops and sputters and doesn't accelerate further. Is this a sign of higher RPM / low MAP issue you mention? Is it too much advance?

Yes to first fuel trims close to 0 / WB AFR error close to 0 also

Not sure about the breaking up without seeing log. I doubt it's too much advance though

And I'm ignoring the rest of the posts.. in the end it's the same. Some people swear by using a wideband for non PE areas, some people swear by using fuel trims. Do whatever works for you. Either way you're just figuring out how far off fueling is. Bottom line if the car is going to be ran in closed loop anyways then I prefer using fuel trims.

[scottt28]

Yes to first fuel trims close to 0 / WB AFR error close to 0 also

Not sure about the breaking up without seeing log. I doubt it's too much advance though

And I'm ignoring the rest of the posts.. in the end it's the same. Some people swear by using a wideband for non PE areas, some people swear by using fuel trims. Do whatever works for you. Either way you're just figuring out how far off fueling is. Bottom line if the car is going to be ran in closed loop anyways then I prefer using fuel trims.

Thanks for the feedback. Here's what I did... First, I used the STFTs to adjust the VVE tables. Then I did some smoothing.

I also took out 5% in timing in the 2800 - 3200 range and added a couple degrees in the 4200+ rpm range.

I'll flash the tune and see if STFTs get closer to the AEM wideband readings. I'm thinking that the short term fuel trims will now have smaller corrections, I'll have a little less knock in the 3000 rpm area and decent power on top.

Tune and log attached. It'S SD mode in closed loop. Roll into and out of the throttle. To some extent, I feel like I am chasing my tail. I go back and forth over a couple percent. Overall, it runs well

[schpenxel]

Yeah, honestly once it gets to a few % I'd just call it good. Let fuel trims do what they do and soak up any small corrections and enjoy it.. I've done the chase a few % up and down thing and it gets old fast