My commanded AFR is changing rich on me!

[Chris81]

I've noticed that at the track (and street) about the middle of 3rd gear and all of 4th my commanded AFR changes from what i've commanded (which is 12.7-13.0) down to commanding 11.77. This is verified by my wideband as well. Why is it doing this and how am I missing this table? I've tuned for years and I have never ran into this.. I've looked through everything I can think of. HELP!

Some background:

2002 Z06 Corvette
Mods: Vararam, ported TB, B&B Bullet catback, my tuning and drag radials.

What is going on?

[rhit_rs]

I had a similar problem with cat protection causing the commanded AFR to drop in a stepped fashion. Does your ecm have this? On the stock solstice LE5 e67 tune it can command up to a .67 reduction factor in afr. I disabled it and ridded myself of the problem.

[Chris81]

I don't see any tables that change AFR, but now that you mention about the Cat Converter I see some tables that have to do with o2 transition times and lean limits under Engine Diagnostics -> Cat Converter Parameters (tables C3201-C3210)

I'm wondering if that might have something to do with it, I guess it might be seeing a lean condition and its adjusting the commanded AFR to cool off the cats.

I need to play with them and see what happens. Or, I guess I can turn off Cat Converter Protection and see if it stops. It would make sense I guess.. i'm used to tuning cars with no cats so I always turn it off.

[joecar]

Hi Chris,

When cat overtemp protection kicks in, the logged commanded AFR does not include the fuel component from COTP.

If you open your log file and your tune file with scantool->tunetool (ST->TT) linking enabled, you can position the scantool cursor on the place where you see richness, and then in the tunetool you can see the operating cells highlighted in the various tables...

To do this, you have to log the same pids as listed on the axes of those tables, and you have to use the same units for those pids between ST and TT, and it makes it easier if you set the TT fueling units to EQ and log both AFR and EQUIVRATIO in the ST.

Post up some log files and your tune file.

[rhit_rs]

Hi Chris,

When cat overtemp protection kicks in, the logged commanded AFR does not include the fuel component from COTP.


Really?? Hmm, it seemed to me it did on the e67 ecm. Perhaps it does on some and not on others...I can post some logs of the effect if you would like. Unfortunately I was not logging cat temp at the time, but I noticed the drops in commanded afr, speculated it was cat protection then disabled the protection. After that, I logged cat temp and no more change in commanded afr even though the temp it would have kicked in at was exceeded.

[joecar]

Oh, Chris has a Gen III PCM which doesn't include COTP in the commanded AFR...

I don't know what the E67/E38 does with COTP.

[rhit_rs]

Oh, Chris has a Gen III PCM which doesn't include COTP in the commanded AFR...

I don't know what the E67/E38 does with COTP.

Oh, okay. For the record I'm pretty sure the E67 does.

Well, it must not be COTP if the commanded AFRs are changing then.

[Chris81]

Hi Chris,

When cat overtemp protection kicks in, the logged commanded AFR does not include the fuel component from COTP.

If you open your log file and your tune file with scantool->tunetool (ST->TT) linking enabled, you can position the scantool cursor on the place where you see richness, and then in the tunetool you can see the operating cells highlighted in the various tables...

To do this, you have to log the same pids as listed on the axes of those tables, and you have to use the same units for those pids between ST and TT, and it makes it easier if you set the TT fueling units to EQ and log both AFR and EQUIVRATIO in the ST.

Post up some log files and your tune file.

I use the linking feature between the TT and ST often.. i'll go ahead and log EQ RATIO tomorrow on a drive and i'll make sure all my units match up and we'll see what happens.

I included my latest tune file that is in the car now, as well as the log for one of my passes from the track the other night. I didn't have the wideband on it then.. but commanded and actual are dialed in. Even with it going rich on that pass, the car still ran 114.39 in the 1/4 at over 4200 ft DA.. so I know its making good power (that MPH corrected is 120.69 for comparison).

[joecar]

Some comments:

Log these
DYNCYLAIR_DMA
DYNAIR
(those will take 2 channels each, so you will have to drop some other pids to keep the channel count no bigger than 24).

What happened at the beginning of your run (see cursor in attached pic of your log)...?

LOL... you have some serious wheelspin...:cheers:

The HOS2x1 voltages 850-905mV indicate a little lean at WOT... 920-930mV would be good (but do use a wideband).

As you run up the rpm at wot, the VE table hits that hole in the upper region (see the white cell in the 2nd pic)... fix the hole (raise it up a little).

[Chris81]

LOL good question, I didn't really notice that.. i'm not sure.. It didn't feel like it was spinning bad that I can remember.. my 60' was a 1.77 on that pass.

The splits were:
60' 1.77
330' 5.17
660' 7.95
MPH 89.2
1/4 12.33
MPH 114.39

I did notice the random drop in timing and my TP% in 1st gear.. I don't remember letting out or anything and all my traction control/active handling was off. Weird..

What do I need to be looking for using those 2 PIDS?

[joecar]

DYNCYLAIR_DMA indexes into B5913 and B5914 to see where you are on the spark map.

DYNAIR is calculated from PCM using MAP, RPM, Temp, and can be used to compare VE and MAF.

[Chris81]

Ok cool.. i'll do some logging and post up.

Thanks for the help!

[ScarabEpic22]

Random question, joecar how do you get the scan tool to link up both RPM columns like that?

[joecar]

Random question, joecar how do you get the scan tool to link up both RPM columns like that?Erik,

In the scantool chart display, drag the vertical cursor to select a portion of the chart trace (i.e. say a few seconds worth, covering a range of RPM say)... then in the tunetool you will see those RPM rows/cols highlighted...

This feature allows you to correlate a very specific part of the chart trace to the tunetool tables cells/rows/cols... very handy as you can see.

Cheers
Joe
:^)

[eficalibrator]

The HOS2x1 voltages 850-905mV indicate a little lean at WOT... 920-930mV would be good (but do use a wideband).
Don't fall into the trap of believing narrowband O2 data at WOT. It's absolutely useless, especially if you don't know the temperature at the tip of the HEGO. You're on the right path when suggesting to just use a good wideband here.

I haven't had a chance to pour through the files, but you may also want to look at the AIR or IVT gains to see if one of those is responsible for your change in commanded EQR. The PCM will always take the richest of its EQR commands in an attempt to remain safe. So, even if PE isn't that rich, something else very well may be...

[joecar]

Greg, very good point... "use a good wideband".

[joecar]

Greg,

Unrelated to this thread, a question that I'm still not sure about:

If running with CLMAF, what happens at WOT... does the PCM still use the MAF or does it revert to VE...?


</hijack>

[ScarabEpic22]

Erik,

In the scantool chart display, drag the vertical cursor to select a portion of the chart trace (i.e. say a few seconds worth, covering a range of RPM say)... then in the tunetool you will see those RPM rows/cols highlighted...

This feature allows you to correlate a very specific part of the chart trace to the tunetool tables cells/rows/cols... very handy as you can see.

Cheers
Joe
:^)

Cool, didnt know that! Learn something new everyday thats for sure.:cheers:

[Chris81]

Don't fall into the trap of believing narrowband O2 data at WOT. It's absolutely useless, especially if you don't know the temperature at the tip of the HEGO. You're on the right path when suggesting to just use a good wideband here.

I haven't had a chance to pour through the files, but you may also want to look at the AIR or IVT gains to see if one of those is responsible for your change in commanded EQR. The PCM will always take the richest of its EQR commands in an attempt to remain safe. So, even if PE isn't that rich, something else very well may be...

Good deal, I always use a LC1 when tuning.. I just didn't have it in for that outing. My actual typically is a little richer than commanded so I shouldn't be lean.

Joecar, thanks i'll richen up the VE a little in those spots, but doesn't my PE rely solely on info from the MAF at that rpm?

Greg, thanks for the insight. Could you possibly give me the ID tag for those tables?

[joecar]

...

Joecar, thanks i'll richen up the VE a little in those spots, but doesn't my PE rely solely on info from the MAF at that rpm?
...
That's what I used to think, but I'm not sure of that anymore.

[eficalibrator]

If running with CLMAF, what happens at WOT... does the PCM still use the MAF or does it revert to VE...?
If you have it set to only use filtered MAF (max rpm to disable dynair set below idle), then it should never look at VE. Period. Going WOT would simply remove the ECU's ability to perform measurements and corrections to AFR, while still applying any positive LTFT's that may have been learned earlier.

[joecar]

Ok, so it still honors the threshold RPM (B0120) for airmass... thanks Greg...:cheers:

[Chris81]

I'm going to bring this thread back up..

I'm going to do some more logging tonight to see if I can figure anything out.. but I did find this thread: http://forum.efilive.com/showthread.php?t=3840 that sounds alot like what is going on with my car. Anyone have any thoughts on this?

Thanks, Chris

[Chris81]

Has no one else had this problem?

I still need to log a little more, I had a couple of computer problems.. but its always the same thing.. commanded AFR will just drop from what i've adjusted in B3618 down to commanding 11.77.

Surely there has to be someone that has had to deal with this problem before.. I've been through every parameter I can think of numerous times and I can't find a table or multiplier or anything that would even remotely cause the Commanded AFR to change. Hopefully i'm just missing something..

Does anyone have thoughts on the post I made above with the link?

Would changing to a custom OS possibly fix this? (I had planned to go to COS5 soon for my nitrous setup)

[WeathermanShawn]

What type of tune are your running?

Did you verify if B0701 is set to 'Disable'?

If running off MAF if you're Commanded AFR does not match actual then you are overestimating CYLAIR. If running off SD, you are running off DYNCLYAIR.

Short of some programming bug, or if your method of tuning is giving you inaccurate data, it appears your DYNCLYAIR is being overestimated. Perhaps COTP, wideband error, or bad log data.

Post up a log and a tune.

[Chris81]

I still have B0701 enabled since the cats are still on.. it was turned off for AutoVE though, I wondered about that in the beginning but was told: "When cat overtemp protection kicks in, the logged commanded AFR does not include the fuel component from COTP."

I'm running a MAF tune and used AutoVE with a wideband for the VE tuning as well as the MAF table. My Commanded and Actual AFR's are usually very close, even when the commanded AFR drops to 11.77, my actual follows it down like it should. I just don't want it to! LOL

I have included my current tun. file and a log from the track in post #8 on the 1st page.

[WeathermanShawn]

I recall a somewhat similar problem once. Seems like it had to do with my Pid selection for commanded AFR and my units.

Somewhere in my scan I did not have the units reading the same parameter. I did a complete reload of the tuning and scan software. Re-validated pids etc.

Seems like it is a conversion problem. I can not find anywhere in your tune where 11.7 AFR is commanded. You could try disabling COTP and see what happens.

[Chris81]

O.K., now I understand what your saying.

I recall a somewhat similar problem once. Seems like it had to do with my Pid selection for commanded AFR and my units. Somewhere in my scan I did not have the units reading the same parameter. I.E. AFR vs EQ.

I think a complete reload of the tuning and scan software. Re-validated pids etc.

Seems like it is a conversion problem. I can not find anywhere in your tune where 11.7 AFR is commanded. You could try disabling COTP and see what happens.

Need Joecar or expert to jump in.

..WeathermanShawn..

I have wondered about my units being off in the scan tool (something like that has happened to me as well, but mine was a scaling issue), but I figured that if that would have happened that my commanded AFR being logged and what my actual AFR is wouldn't jive.. for example my actual would be correct but my commanded would be off since all the scan tool is doing is "scanning" and not able to change things up. does that make sense? LOL

I will for sure do some more logging tonight with COTP off as well.

[WeathermanShawn]

Chris, your log is the first one I have ever seen where you have +Ltrims between shifts. Maybe you just shift better than me. Every log I have at WOT I enter at negative trims, and between shifts stay at 0.0 Ltrims.

Yours go positive. Really your entire Ltrims never go negative. Are you carrying +LTrims into WOT? Now why that would change your commanded AFR is beyond me. Just wondering what is happening to those +Ltrims? The software says 0.0, but when you shift they are right back there.

Is it possible your +LTrims are getting added to your commanded AFR? That percentage of Ltrims would make your AFR at ~11.7.

[Chris81]

My trims are goofy because I need to dial it in some more, I changed a couple of things and haven't adjusted for them yet, as well as the Vararam CAI I have always seems to sway trims more so than F-body stuff.. usually my avg. LTFT's are +2 so.. they were going positive between shifts because its trying to command 14.68 for that time in between. When it starts to command 11.77 all the trims are zero.

[joecar]

Make sure that the units agree between the pids in these places:
- scantool Data tab (edit these on PIDs tab by rightclick->Metric/Imperial);
- tunetool table axis (edit these by going Edit->Configure Display Units);

Check also this:
In tunetool, go Edit->Properties->Display and check that fuel units are not Lambda.

I like to use these units: VE units g*K/kPa and fuel units EQ.

[Chris81]

Ok, looks like I found the problem! I will do some more logging soon to make sure, but it looks like it ended up being the COTP after all.. I guess it does alter the commanded AFR.

I'll include some logs to show.. I went ahead and adjusted my fueling to get rid of the positive trims, so now my avg's are .4 and -1.4 on the last log and 0 at WOT :)

The only thing I did between logs is turn off B0701, the first log shows it going rich again, as always.. and the 2nd log it doesn't do it at all.. I did get a bit of KR though, so it looks like I might need to add a little fuel. I wasn't around the shop so the WB wasn't in, but I promise its right.

Thanks for everyone's help!

Chris

PS, the files included are the 2 logs and the current tun. file in the car.

[WeathermanShawn]

Wow, good job!

Kinda going fast there..Just kidding. You ought to see some of my logs. We have some pretty wide open roads here. I always count my 1 second frame(s) Vss acceleration to compute 'road HP'. Saw you are doing two-second frames, so nice to compare.

Noticed you maxed out at 90 Kpa. Are you at elevation? I can barely hit 82-84 Kpa on a good day up here at 'mile high'.

You may have discovered something worthwhile here. It appears when you hit about your 3rd WOT in succession, your COTP kicks in. Never realized it actually changes your commanded fuel. Perhaps there is a 'timer' involved that calculates how many seconds your narrowband O2's stay above 800mv. You seemed to have hit it.

I have ran headers/w cats for over a year now with COTP disabled. Always noticed before disabling COTP I would hit ~11.7 on the wideband, but never logged commanded AFR at the same time.

[joecar]

Chris,

That's a good log... see how EQIVRATIO follows the PE table (which is richer than the OL commanded fuel table at those points)...

As WeathermanShawn said, note that you're logging at 5 frames/second since your pid channel count is above 24.

[Chris81]

Thanks guys..

Weatherman.. i'm out in west Texas where it is around 3000 ft actual elevation.. and there are PLENTY of open roads..haha. I have noticed that the car would usually start to go rich in the middle of 3rd gear it seems.. I wondered if it took that long on each run to get "hot" enough to kick in COTP or there was some sort of timer involved.

Joecar, I see that! I'm glad I got this figured out.. it does pull a little better on the top end now that i'm leaned out.. i'm going to have to hurry and make another trip to the track soon to see what I gained.

And, sorry about the pid channel count.. I guess I forgot to get rid of something.. I usually never go over 24.. I guess I was in a hurry! I'm going to log it again in a day or 2 (I wait until night time since we have been over 100 degrees here already) just to make sure.

[joecar]

I was wrong, it looks like COTP may be reflected in commanded AFR...

see this: showthread.php?p=97048#post97048 (http://forum.efilive.com/showthread.php?p=97048#post97048)

It's a little confusing because it doesn't always show up and there's no discovered COTP AFR table.

[Chris81]

Yup, its a weird deal.. I've logged mine many more times with COTP turned off and the AFR has always commanded what I wanted it to.

[5.7ute]

I was wrong, it looks like COTP may be reflected in commanded AFR...

see this: showthread.php?p=97048#post97048 (http://forum.efilive.com/showthread.php?p=97048#post97048)

It's a little confusing because it doesn't always show up and there's no discovered COTP AFR table.

I had a quick look through the PCM map I have for the ls1 and there is a few tables in regards to COTP that we dont show in efilive. From the small descriptions evident there is a filtered multiplier at play & not a base AFR that I could have easily added in a cax file. This has put it well out of my range of skills to easily implement.
I have emailed you a copy with the addresses of the tables to have a look.
Cheers Mick