MAF table dialed in!!! *Tune almost complete!

[txhorns281]

OK guys, here's the final result! Download the MAFCalibration.cfg file so your scanner will be setup to observe the results! I layed my MAF table right over my dynamic airflow and I sure am impressed... Now my only question is this.... Why are my LTFTs still so negative around low rpm and idle??? I tuned out my table via Wideband, SD/Open Loop pretty dang close to 13.0 as WS6Snake-Eater's write up suggested... Then, I aligned my MAF table to my dyanmic airflow as you can see on the LOG file here...

*EDIT* apparently these files might not work on older versions of HPT... a buddy of mine just tried to view then and could not... I used v1.5

https://webspace.utexas.edu/wl333/FinalMAF.hpl (Right-click save as please!)
https://webspace.utexas.edu/wl333/MAFCalibration.cfg (Right-click save as please!)

but check out the LTFTs in the histogram... I have a few ideas about what's going on here... but some seem inconsistent...

1) When i was tuning in SD/Open Loop i was targeting a 13.0 AFR, so my VE table is dead on when the VCM commands 13.0, but now that I'm back in closed loop and I'm commanding 14.628 so it seems natural that my motor will see me as running rich and subtract fuel... BUT if you look at the histogram my high MAP values the LTFTs are within normal swing... So if my whole table is tuned for 13.0 why wouldn't the whole table swing way off when trying to reach 14.628?

2) I did reset my fuel trims before logging but it seems that the LTFTs have somewhat already chosen which way they want to go... However the STFTs are still positive for now so the total corrections aren't to far of a swing... Does this mean anything?

3) Am I not waiting long enough for my fuel trims to learn themselves out? I mean I drove for about 20 min, maybe 20 miles but like i said you can already kinda see where they wanna go sooooo I don't think there's too much room for improvement... And it is running richer cuz they gas smell is worse than before... pretty much as bad as before I started tuning...

Thoughts and comments always appreciated! (Almost there dammit...)

[Dragman]

I like to drive 50+ miles to get to settle in.

[pknowles]

Well I logged the MAF freqency (mass air flow Hz), MAF airflow (mass air flow SAE), and Dynamic airflow. I had my MAF fail set to 0 and ran around with the MAF still plugged in. The graph shows that the MAF is reporting less air flow then the Dynamic flow calculation, which explains why I'm lean when I enable the MAF. The MAF is saying less air is coming in so less fuel is put in and I run lean. The best fit curve for the MAF falls on the data so it's hard to see the blue line. I'm working on building a new MAF table but I was wondering if the Dynamic airflow calculation takes into account for the fuel trims? If it does then the trims just have to be settled if not then I guess I have to get a little closer on the VE table; right now I'm between -6 and 0 on all cells.

graph (doesn't look great in IE6, save and look at in something else if you need too):
www.glue.umd.edu/~pknowles/MAFplot.bmp

[txhorns281]

Well I logged the MAF freqency (mass air flow Hz), MAF airflow (mass air flow SAE), and Dynamic airflow. I had my MAF fail set to 0 and ran around with the MAF still plugged in. The graph shows that the MAF is reporting less air flow then the Dynamic flow calculation, which explains why I'm lean when I enable the MAF. The MAF is saying less air is coming in so less fuel is put in and I run lean. The best fit curve for the MAF falls on the data so it's hard to see the blue line. I'm working on building a new MAF table but I was wondering if the Dynamic airflow calculation takes into account for the fuel trims? If it does then the trims just have to be settled if not then I guess I have to get a little closer on the VE table; right now I'm between -6 and 0 on all cells.

graph (doesn't look great in IE6, save and look at in something else if you need too):
www.glue.umd.edu/~pknowles/MAFplot.bmp

you probably ought to get Dynamic SD Airflow since I do think that trims will effect that somewhat... The trims partially make up for what inconsistency the computer sees b/w the MAF and VE (the other make up is for the inconsistency of the VE itself) so Dynamic Airflow for the motor might very well be the same or close, but I like to take the variable completely out of the question just to be sure...

[MNR-0]

My understanding is that the MAF will own all airflow calcs. providing it reports values within error margins of the VE table. The error margins are defined in Engine Diagnostics.

P0101 - MAF Performance
Maximum Delta Air Flow - P0101 Error

The VCM will constantly check the MAF operation by comparing the MAF calculated airflow against the calculated airflow using speed/density calculations. If the difference exceeds the threshold value at the particular speed/density calculated airflow then P0101 will be set.

I also understand that if the MAF fails with this code SD will take priority. But this should be confirmed by someone more experienced. No error then the MAF calc. will be taken as God.

[pknowles]

With my stock tune I was showing LTFT's around +5 all over the table, I then set MAF fail freq to 0 and my trims go around -3 changing nothing else and the only code I have is P0102.

[txhorns281]

My understanding is that the MAF will own all airflow calcs. providing it reports values within error margins of the VE table. The error margins are defined in Engine Diagnostics.

P0101 - MAF Performance
Maximum Delta Air Flow - P0101 Error

The VCM will constantly check the MAF operation by comparing the MAF calculated airflow against the calculated airflow using speed/density calculations. If the difference exceeds the threshold value at the particular speed/density calculated airflow then P0101 will be set.

I also understand that if the MAF fails with this code SD will take priority. But this should be confirmed by someone more experienced. No error then the MAF calc. will be taken as God.

I was always under the impression that in lower rpms the airflow calc was a hybrid of VE and MAF... But in higher rpms dictated by the MAF... In any case, you still would want the MAF to be reporting the same airflow as what the motor needs/wants/is using. The only way to truly see what airflow the motor is utilizing, is in SD mode (ie. without MAF or trim interference) If you command a particular AFR and the VE outputs it with relative closeness, then the VE can be trusted as a reliable airflow calculator... at least well enough to use for MAF recalibration... Does this sound right?

[BQuicksilver]

txhorns, what ever came of this? I'm in that same spot right now with my VE supposedly okay and then seeing a similar log to what you posted above (neg in that area, okay elsewhere) after re-enabling closed loop.

[txhorns281]

txhorns, what ever came of this? I'm in that same spot right now with my VE supposedly okay and then seeing a similar log to what you posted above (neg in that area, okay elsewhere) after re-enabling closed loop.

Well... it seems that playing with O2s and closed loop configuration will end up being the answer to all this... I've been a little busy moving lately but I haven't forgotten about this. (It's been like a splinter in the back of my mind!) MNR-0 seems to have figured this out, and i kinda get where it's going, but haven't had much time to test anything...

[WS6snake-eater]

Hey Will, when are you coming back into town? We need to put our heads together and get this figured out. Also I have a new MAF method that I want to give a shot. Hit me up if you will be in this weekend.

Matt

[txhorns281]

Hey Will, when are you coming back into town? We need to put our heads together and get this figured out. Also I have a new MAF method that I want to give a shot. Hit me up if you will be in this weekend.

Matt

Matt I won't be all settled till Monday but like you said, definitely need to figure this one out! I should probably go ahead and do vette O2s too huh... the more I was looking at some of my logs, I see that negative trimming crap going on when my O2 oscillations thin out. They're supposed to be tight oscillations right? Sometimes mine get all "gappy" and what not and that's where often i see rediculous trimming...

[WS6snake-eater]

I just created a tune with MNR suggested o2 settings...I'm not sure how it is going to act with a cam as big as mine, but I am going to give it a shot for the next couple of days and see how it turns out. The theory seems to be correct, and I can follow what he is saying now that I have a visual to look at. I really need to get a hold of a wideband to re-check some things. When you get down you need to let me borrow yours ;D

Matt

[txhorns281]

strangely today my trims have been doing VERY VERY VERY well... alot of that rich stuff is gone.... ???

[MNR-0]

Let us know how you go Snake and Horns.

Snake, a big cam will make things more difficult for the O2 sensors to stabilise, mainly because ot the extra air and cooling effects on O2s that result from increased durations. A choppy idle will really make life hard, but that's the price we pay.

However, the principle still applies.

Remember, stay with the stock MAF cal until right at the end, then rescale the entire MAF cal. by an appropriate % after the TRIMS have stabilised. You don't want to go changing too many things at once.

I was talking to a guy the other day about what I had done. He seemed a little miffed that my TRIMS were stable throughout all the inclement weather we been having lately. He builds race engines for 24hr enduros and Targas for some of the most elite pro-circuit drivers in OZ.

[txhorns281]

Remember, stay with the stock MAF cal until right at the end, then rescale the entire MAF cal. by an appropriate % after the TRIMS have stabilised. You don't want to go changing too many things at once.

i went the other direction and just ditched the sunnufabitch... lol After tearing my hair out trying to put together a MAF recalibration procedure... I did come up with something good though!

[WS6snake-eater]

Well I went ahead and loaded MNR's o2 values in tonight and things seem to be a little more stable. FWIW, I too am MAFless... I"m going to give it day or two for the trims to settle in then make one final VE adjustment.

Hey Will I think I might be on to something with the MAF, where we can blend it into the VE table in a linear fashion, and not loose any of the postivie characteristics of the SD tune. In fact I think that it might allow for better gas milage, and more stabability during weather changes. MNR thanks for sharing your insights into the o2 thing. I hope that it really helps out in the long run, and right now it seems to be working great. I'll keep you guys posted.

Matt

[checkmate]

Well I went ahead and loaded MNR's o2 values in tonight and things seem to be a little more stable. FWIW, I too am MAFless... I"m going to give it day or two for the trims to settle in then make one final VE adjustment.

Hey Will I think I might be on to something with the MAF, where we can blend it into the VE table in a linear fashion, and not loose any of the postivie characteristics of the SD tune. In fact I think that it might allow for better gas milage, and more stabability during weather changes. MNR thanks for sharing your insights into the o2 thing. I hope that it really helps out in the long run, and right now it seems to be working great. I'll keep you guys posted.

Matt

What were the o2 values? I seemed to have missed that somewhere. ???

[MNR-0]

Did you just load my Airflow Mode and Base Airflow Gain data? Cause that should be in the ballpark regardless of SD or size of MAF.

[MNR-0]

Oh, and another thing, I wouldn't recommend adjusting your LTFT boundaries either. They are set good for what closed loop was designed to do.

Reason is because if you log MAP vs RPM you will see that the preset boundaries cross-hair quite well under light throttle conditions.

In the end all you want is stable fuelling.

[WS6snake-eater]

Oh, and another thing, I wouldn't recommend adjusting your LTFT boundaries either. They are set good for what closed loop was designed to do.

Reason is because if you log MAP vs RPM you will see that the preset boundaries cross-hair quite well under light throttle conditions.

In the end all you want is stable fuelling.

I have changed my map and rpm ltft boundries...primarly because of the size of my cam, and my map at idle. It seems to have helped trimming. What are your LTFT boundries set at in stock form? Perhaps they are different than ours.