My Car Thinks Stoich is 15:1

[RevGTO]

Yep, multiple tuning experiments have proven it. I have played with VE and MAF calibrations and the only thing that changes is the LTFT's. No matter what I do, the logged AFR stays the same. I have gone from wildly negative to wildly positive LTFT's by altering the VE and MAF tables. But logged AFR stays around 15:1 while commanded is 14:63.

I don't really mind as I'm not running cats and being a little bit lean isn't going to hurt anything. But the net effects are that PE and lean cruise are off, so that I have to compensate by commanding richer AFR's than actual.

I thought that I had jacked my VE and MAF tables both so far out of whack that they were finding equilibrium at 15:1. But that theory didn't test out.

So I'm left with two theories: either the .cax files corrupted my tune so it wants to go lean, even when not commanded (unlikely), or that ethanol in the gas misleads the O2's regarding stoich.

Any opinions?

[5.7ute]

You will find that the fuel trims will trim until a stoich AFR (according to the narrowband sensors) is met when in closed loop, regardless of the fuel used. There is obviously a difference in what your wideband reports as stoich in comparison to your wideband.
What is B3601 set to?

[mr.prick]

What WBO2 are you using?
The pre-defined PIDs for all of the WBO2's are for the manufacturers stoich AFR and
that is almost always 14.7:1
Serial WBO2 is the same.

It maybe just a little bit of error between the calc_pid for AFR and commanded AFR.
If that's the case it's .03 AFR leaner than commanded not .07AFR leaner

[joecar]

Did you disable lean cruise, and do a full flash...?

[RevGTO]

B3601 is 14.63. I'll have to check my LM-1, but IIRC, stoich is 14.7 on there, so that could be a source of some of the discrepancy.

Lean cruise is not enabled; my current tune is built from one that has never had it enabled. But I have not done a full reflash with it.

[WeathermanShawn]

PE and lean cruise are off, so that I have to compensate by commanding richer AFR's than actual.

or that ethanol in the gas misleads the O2's regarding stoich.Any opinions?

RevGTO:

I thought you were running a lean cruise of sorts ( going from memory on your tune). I'm with Joe. I would make sure my B3605 is 'stock' and/or eliminate the EQ's <1.0 and make sure B3639 is all zero's. Then do a complete re-flash. I assume all your B3618 has EQ > 1.0?

I changed all my B4105 setting to 550mv. It just seemed to line up with an EQ of 1.0 (14.63 AFR) much better. Thats a thought.

My hunch is that it is still your wideband reading. I know you went through all the voltage offsets (another thread), but lately I have been glad I went with a serial connection through V2. Other than occasionally changing out the O2 sensor, it has all went flawlessly.

I believe the ethanol will change stoich more toward AFR values approaching 14.1 (10% ethanol). So it does not make sense it would swing you toward higher AFR's.

Just my thoughts. Good luck..

[mr.prick]

Ethanol in the gas won't change the AFR reading the sensor displays.
For serial WBO2 when stoich is seen it will display the AFR that it is programmed to show and for Lambda 1.00 and for analog it will depend on the calc_pid expression.

You can correct this by checking the Lambda value the WBO2 gives and multiplying it
by {B3601}.
Serial or analog?
If analog what is the analog output programmed to?

[joecar]

RevGTO, if you believe something got corrupted, do this sequence:
1. view all tables, make sure they're all sensible/sane and in-range.
2. full flash in a known good stock file for your PCM/OS/year/model.
3. cal-only flash in the file from 1. above.

Doing that will fix anything that may have gotten munged by an incorrect .cax definition.

[RevGTO]

RevGTO: I thought you were running a lean cruise of sorts ( going from memory on your tune). I'm with Joe. I would make sure my B3605 is 'stock' and/or eliminate the EQ's <1.0 and make sure B3639 is all zero's. Then do a complete re-flash. I assume all your B3618 has EQ > 1.0?

I changed all my B4105 setting to 550mv. It just seemed to line up with an EQ of 1.0 (14.63 AFR) much better. Thats a thought.

My hunch is that it is still your wideband reading. I know you went through all the voltage offsets (another thread), but lately I have been glad I went with a serial connection through V2. Other than occasionally changing out the O2 sensor, it has all went flawlessly. 1. I WAS running lean cruise ... this is one of the reasons I wanted to insure my AFR's were on target. Same thing with PE at the track. I had already edited all the tables you mentioned ... zeros in B3639, all cells in B3605 at stoich or below, except for a few that are 14.77, PE in mid to low 12's.

2. I have my nb's at 450mv ... that's what your tutorial suggests, but I've done that for a while ... I think AutoVE may recommend that. Do you now suggest 550, or is that a typo?

3. I think my wideband reading is good. Per the other thread, I knew it was off because logged consistently read leaner than metered. Since reprogramming, the logged and meter read exactly the same. I reset the stoich calibration to 14.6. We'll see if that makes a difference. This is V1 analog btw. All the specs can be found in the final post here: http://forum.efilive.com/showthread.php?t=13380&page=4

[WeathermanShawn]

No, I don't not recommend 550mv as the O2 voltage. I am only suggesting it as one way of manipulating stoich AFR. I also recommend that running the factory O2 settings are acceptable..they vary per airflow, and do include 550mv at the higher airflows.

I don't know what to say. My car runs at 14.63 AFR..so if someone else runs at 15.0..and they are 100% sure their wideband is totally accurate..I don't know what to say other than it is very unusual.

I guess you could change B3601 to 15.1 as your EQ and tune for that.

Other than trying another wideband..I guess you have to tune around your new stoich. Tough call. If it was me, I would go to a dyno or try another wideband.

Anybody have any ideas?

[mr.prick]

I found this post confusing. (http://forum.efilive.com/showpost.php?p=119399&postcount=39) :confused:
Are you using the LM-1 PID ({EXT.AD}*10) with different analog output settings?
If this is the case these settings:
0.88333v @10:1 AFR
4.21667v @20:1 AFR

are the same as:
0.00000v @7.35:1 AFR
4.21667v @20.00:1 AFR
Which is the same as ({EXT.AD}*3)+7.35 and matches the LC-1 PID not the LM-1 PID of ({EXT.AD}*10).

[RevGTO]

Well, you can never be 100% sure, but I reprogrammed the wideband for 14.6 and did a free air cal ... came out just right at 20.9.

If all else fails, I'll do the full reflash as you and Joe have suggested. The thing used to read 14.6-14.7 before my old laptop died at the end of track season last year. When got a new one and resumed tuning this spring, one of the first things I did was download and install the .cax file for lean cruise. I can't be sure if it was cause or coincidence, because that's when I started logging and getting stoich in the low 15's.

That's when I started with CalcVE which allowed me to get my LTFT's dialed in like never before. But logged AFR stays on the same baseline no matter what I do.

[RevGTO]

I found this post confusing. (http://forum.efilive.com/showpost.php?p=119399&postcount=39) :confused:
Are you using the LM-1 PID ({EXT.AD}*10) with different analog output settings?
If this is the case these settings:
0.88333v @10:1 AFR
4.21667v @20:1 AFR

are the same as:
0.00000v @7.35:1 AFR
4.21667v @20.00:1 AFR
Which is the same as ({EXT.AD}*3)+7.35 and matches the LC-1 PID not the LM-1 PID of ({EXT.AD}*10).If I understand your question, the answer is "yes." They are similar to the LC-1 settings. I followed 5.7 Ute's advice and it got my logged AFRs to match the meter readings on the LM-1. It's a long thread but you can follow his logic there. All I know is that it worked.

[WeathermanShawn]

Yes, I am confused also.

Since tuning is so 'sensor dependent', when things get illogical or too confusing..I try to go back to basics.

More than once I have restarted my entire tuning program from step 1. Flash in the stock tune, do AutoVE or WeathermanShawn's Tutorial, put in new sensors, check plugs, vacuum, exhaust leaks, etc.

We are so wideband dependent (a good thing) that if it is not accurate..we have a bad tune.

I also use Innovates LM-2. I calibrate it (reads 20.9). But I still replace the sensor about once a year. I try to match it up with IBPW at various RPM's and MAPS to check my overall fueling.

I'll stick with the theory that the wideband sensor is not giving a representative reading. If you are commanding 14.63 AFR and reading 15.0, I would put a 75-80% probability that the wideband reading is not accurate.

I am not an expert on it..just my opinion. It would be a shame to not tune correctly due to an inaccurate sensor.

I would trust Mr. Prick's calculations. He has done good work on this area of tuning. Let us know how it goes..

Tuning Enthusiast..

..WeathermanShawn..

[RevGTO]

There's definitely truth in what you're saying. Just resetting the LM-1 to 14.6 from 14.7 brought me down to a 14.85 average. It's an .03 error to the rich side vs. .07 to the lean side, but closer overall.

Went to the track tonight. I'm going download and post up a pass where you can see how crazy I am in PE.

I think you're right and I need to start all over again and do AutoVE. Thanks to all you guys for your good advice, and sorry to be a nuisance on this board with my jacked tune.

[mr.prick]

The easiest way to fix this is to restore the default settings of the LM-1 and
use the WBO2 spreadsheet to edit the sae.generic.txt

[WeathermanShawn]

I think you're right and I need to start all over again and do AutoVE. Thanks to all you guys for your good advice, and sorry to be a nuisance on this board with my jacked tune.

Your not a nuisance. If I or someone else sounds annoyed, it is because we share your frustration.

I actually think you have shown good persistence and understanding. An area 'we' do not talk about much in tuning relates to injectors, fuel pressure, etc. Its possible that a lot of irregular fueling during PE comes from even the stock injectors not 'being perfect'. I do not address it in my Tutorial as it seems to be an advanced topic. But if airmass is spot on and fueling is not..one wonders..:sly:..

Later..

[RevGTO]

I just realized that unless you have my LM-1 settings in your sae_generic.txt file, the logged AFR's won't display properly, so no point in posting up a log.


The easiest way to fix this is to restore the default settings of the LM-1 and use the WBO2 spreadsheet to edit the sae.generic.txtBefore changing my LM-1 and sae_generic settings to what I have now, my logged AFR's differed considerably from what the meter read on the LM-1. At least they match up now. But I can play with your system. Certainly it seems like the right way to match WBO2 output to B3601. I may need to PM with questions ... I am a math and excel idiot ...

[mr.prick]

Yep. :hihi:

[joecar]

Hey, if you don't persist you'll never solve the problem... :)

[ScarabEpic22]

Hey, if you don't persist you'll never solve the problem... :)

Sounds like something a programmer would say...:hihi:

[5.7ute]

I found this post confusing. (http://forum.efilive.com/showpost.php?p=119399&postcount=39) :confused:
Are you using the LM-1 PID ({EXT.AD}*10) with different analog output settings?
If this is the case these settings:
0.88333v @10:1 AFR
4.21667v @20:1 AFR

are the same as:
0.00000v @7.35:1 AFR
4.21667v @20.00:1 AFR
Which is the same as ({EXT.AD}*3)+7.35 and matches the LC-1 PID not the LM-1 PID of ({EXT.AD}*10).

Whats so confusing? This method removes any ground offset & puts the scanned values in synch with the unit. By using the lc-1 settings in the lm-1 you also get much better resolution in the scanned values. Pretty much a win,win situation IMO.

[mr.prick]

What I got from the post was that he was using ({EXT.AD}*10) with the output programmed for ({EXT.AD}*3)+7.35

Using the LC-1's output with the LM-1's PIDs. :nixweiss:

[5.7ute]

LM-1 sae_generic.txt pid:
*CLC-00-928
V 0.0 5.0 .1 "{EXT.AD1}"
AFR 10.0 20.0 .2 "(({EXT.AD1}-({EXT.AD2}-2.5)*3)+7.35 "

The generic txt was modified to use the LC-1 - ground offset equation.

[RevGTO]

My understanding is that mr. prick's WBO2 spreadsheet is set up to cancel out the discrepancy in stoich between {B3601} and the WBO2's default 14.7.

I can say that 5.7 Ute's PID equation brought my LM-1 meter reading and logged AFR's into harmony. Changing the LM-1 default AFR to 14.6 brought it even closer to commanded stoich - my {B3601} is stock for LS1 at 14.63.

So my thought was that by using mr. prick's spreadsheet I could get my logged AFR's even closer to commanded.

This all has to do with CalcVE tuning. The more I play with it, the more I recognize that you can stretch the VE and MAF tables to compensate for one another and attain to good LTFT's at various settings. But proper AFR's don't follow ... which in my mind means that you have to start with a verified VE table and then tune the MAF accordingly ... in other words, back to AutoVE and AutoMAF.

[RevGTO]

What I got from the post was that he was using ({EXT.AD}*10) with the output programmed for ({EXT.AD}*3)+7.35
Using the LC-1's output with the LM-1's PIDs. :nixweiss:But you program both the LM-1 and edit the LM-1 sae_generic.txt pid to reflect the same values ...

[5.7ute]

My understanding is that mr. prick's WBO2 spreadsheet is set up to cancel out the discrepancy in stoich between {B3601} and the WBO2's default 14.7.

I can say that 5.7 Ute's PID equation brought my LM-1 meter reading and logged AFR's into harmony. Changing the LM-1 default AFR to 14.6 brought it even closer to commanded stoich - my {B3601} is stock for LS1 at 14.63.

So my thought was that by using mr. prick's spreadsheet I could get my logged AFR's even closer to commanded.

This all has to do with CalcVE tuning. The more I play with it, the more I recognize that you can stretch the VE and MAF tables to compensate for one another and attain to good LTFT's at various settings. But proper AFR's don't follow ... which in my mind means that you have to start with a verified VE table and then tune the MAF accordingly ... in other words, back to AutoVE and AutoMAF.

RevGTO, as you dig further you will notice that many parts of the tune need to be accurate for any of the auto tuning methods to work accurately. Also whenever you use any fuel with an AFR away from 14.7:1 things will get complicated real fast unless you start using EQ or Lambda units.

[WeathermanShawn]

This all has to do with CalcVE tuning. The more I play with it, the more I recognize that you can stretch the VE and MAF tables to compensate for one another and attain to good LTFT's at various settings. But proper AFR's don't follow ... which in my mind means that you have to start with a verified VE table and then tune the MAF accordingly ... in other words, back to AutoVE and AutoMAF.

I have to respectfully challenge that assertion. If your have functional narrowbands that properly switch at stoich, and your wideband is functioning properly and accurately..all CALC.VE Table is doing is computing the airmass (from the MAF), applying an LTFT correction, and applying the Ideal Gas Law and expressing it as a VE %.

See it has nothing to do with a 'verified VE Table'..if you do AUTOVE and your wideband is not correct, your results will not be accurate either. It is not the method that is giving you an improper AFR..If you do an AUTOVE and your wideband is reading .2 lean..all your values will be .2 lean.

While I will agree that everyone should do an AutoVE (preferably on a dyno) if you are going back to closed-loop and keeping your MAF..you will have to tune with LTFTS and MAF frequencies all over again. Your MAF will be driving the majority of your Trims and your spark, along with your Airflow Calculations (steady-state).

I have done about 6-8 AUTOVE's along with about 75 logs using the CALC.VE TABLE Method. Here is an example of a 1 1/2 hour Log, using an LM-2 Wideband with Serial Connection and two 1 year old narrowbands.

B3601 is 14.63 AFR (EQ=1.00).

I filtered out low ECT's, and throttle transients. The remainder shows the average AFR both in closed-loop and WOT. I have over 50 logs that all average 14.63 AFR. It is richer in PE Mode (as commanded) and the low MAPS and higher RPMS (rich decel). I still think it is your wideband. I would just upgrade to serial connection and then see how it goes..:)

[RevGTO]

Yes, thinking more clearly in the light of day, I recognize that conclusion was not warranted. What I have observed, though, is CALCVE will correct MAF settings such that you can get LTFT's in the desired range with a variety of VE settings. But the logged stoich varies very little ... back to the original problem.

I'm going to work with the WBO2 spreadsheet a little more and see if I can make progress. But before I do, is there any reason not to change {B3601} to 14.7 for test purposes? That would be the simplest way of canceling out the discrepancy between {B3601} and LM-1 programming.

I want to do all I can to try to solve the problem before purchasing any new product.

[WeathermanShawn]

Rev..I have wondered about that too..changing B3601 to read 14.7 AFR (EQ=1.00) to match the LM-2 programming (since you can not technically program 14.63 to the exact digit).

But, I always wondered if the PCM 'expects' its original stoich reading. Like it is hard-wired. A couple of times logging I would get a Commanded AFR of 15.35 at Idle..Still in closed-loop..could not find a table anywhere that commanded it.

So the short answer is I do not know. As a F.Y.I., I just put in a fresh wideband sensor..the only time I could get anything near leaner than stoich is if I was accelerating with a lot of Spark Advance (non-PE). So I guess it is always possible that something mechanically could cause a higher than average 'stoich'.

But, I am guessing. The best solution may be a Calculated PID that allows you to log LTFTBENS using 14.85 or higher as your stoich. Then everything would match. I agree LTFTBENS are not the most 'glamorous' way to tune..but once you get them nailed down..they barely drift and your car runs very smooth with very little variation of AFR.

It is about the best you can do in closed-loop. Now hitting Commanded AFR in PE Mode is the real art. Thats what takes some skill to consistently hit!

Later..

[mr.prick]

Analog is a little shakey compared to serial so don't sweat it too much. :doh2:
http://forum.efilive.com/attachment.php?attachmentid=6496&d=1256054929


Shawn, is AFR richer when in the overdrive gears?
I stay in fourth gear when doing WOT pulls. ;)

[WeathermanShawn]

Mr. Prick I have seen AFR vary per gear..normally my 4th gear (M6) is .1-.3 leaner at lower RPM's and by 4000 Rpms is within +/- .1-2.

But, like you I stay out of overdrive gears..1st & 2nd occur too fast..3rd (M6) is where I do a lot of tuning..I can keep my Rpms and speeds somewhat sane.

My AFR problems tend to come at Rpm's below 3200. If the airmass (esp MAF) has too low of a frequency for the Rpm..you can see the resultant IBPW is 'too low' Same if I get any pinging or KR..Lower MAF=lower airmass..not enough fuel.

But, to directly answer your question I stay out of overdrive gears for the most part. 3rd is my main tuning gear..With the 3.90 gears I can do a 40-95 mph pull pretty quick..and stay within 25 mph of the speed limit.

Disclaimer..Don't try this at home. For off-road use only..:grin:

[RevGTO]

I can't see any harm in trying 14.7 in {B3601}, but I'm out of town for a week and so no logging. But I can continue to work on ideas to try to isolate the problem.

I have V1 and I presume there's no way to connect a serial cable to it. Right?