EFI Idiot

[Rockgolem]

I'm a self acknowledged EFI idiot. I'm a long time EFI live forum reader, first time poster. I'm surrendering because my brain is bleeding trying to figure out wtf I'm doing. Let me give you some background.
- L92 from an 08 Cadillac in a rockcrawler, stand alone (modified) engine harness.
- 6l80
- Engine has
-- LS3 intake/injectors
-- L92 stock MAF/MAP sensors
-- VVT delete kit
-- yella terra 1.8 ratio roller rockers
-- vinci torker cam CC236xl .551”/.551/212/224/114+0A
-- shorty headers, mufflers, no cats or rear o2 sensors

I had a mast HO VVT cam in it and took it to the dyno earlier this month, it made 0 power under 3500 rpm, the tuner recommended I change the cam and delete the VVT. So, I deleted the vvt and installed the vinchi cam. It seems to run fine but you can tell it needs adjustment. I want to tune the Cam on my own, I have a V2 and a LC-2 that I have up and running. I tried following the virtual Ve tutorials but can't get it to log right since the "commanded AFR" pid is unavailable (x'd out). I did the best I could come up with but have absolutely no idea what to do with the data I harvested on my test drives. Would anyone mind giving my tune and logs a view and telling me what I should be doing? I'm not sure what is considered normal, abnormal, or dangerous at this point. The Wideband O2 seems to be reading a lot rich at idle (around 20-25 AFR).
Attached are the Tune. Log 1 which was taken with the MAF disconnected and the autove tune I made (would hardly run unless loaded when the engine was already over 180 degrees) Log 002 is with the MAF hooked back up and the attached tune loaded. Suggestions? Thanks ahead of time -EFI Idiot.

[joecar]

Log E38.AFRATIO and SAE.LAMBDA instead of GM.AFR... (if SAE.LAMBDA works properly then this is the pid to log).

Is your LC-2 connected to your V2 using serial comms (it should)...?

So your pids would be SAE.LAMBDA and EXT.WO2LAM1... so your BEN pid would be CALC.WO2BEN = EXT.WO2LAM1 / SAE.LAMBDA, we would need to edit your calc_pids.txt to do this.

[joecar]

the injector pids are GM.INJPWB1 and GM.INJPWB2 (there is no DC% pid, we'll have to edit calc_pids.txt);

if GM.KR does not work, log GM.KNKRET or GM.KNKRET_B;

there are some other E38 pids you should be logging also, see the VVE Guide.

[joecar]

Ok, in Log_0002.efi I see your LC-2 connected using serial comms...

I would also log SAE.LAMBDA as a sanity check.

You need more pids to see what is going on.

[joecar]

See post #83 here: E38-VVE-Setup-Guide/page3 (https://forum.efilive.com/showthread.php?8961-E38-VVE-Setup-Guide/page3)

[Rockgolem]

Ok, its set up like that I added a few more PIDs to the log. The vehicle is warming up now, once its hot, I'll load the vve dial in tune (set up like that guide) and go for a ride to see what I can capture. What is still unclear to me after reading all of those guides is how I use the information gathered on the log to make adjustments to the VE tables. Will have new logs in a few.

[Rockgolem]

Okay, here is the log from the drive. Tune set up for logging with all of the minimum ids from the E38-VVE setup guide plus a few extras. Map disconnected, loaded then had a nice long drive but didn't get a whole lot in the upper RPMs.
18870

[Rockgolem]

Now I'm pretty sure I didn't do this right because it leaned out way to much but......I copied the Map with labels and then pasted it into B8101 using past with "multiply and labels" After I drove it with this tune, it seemed to lean (perhaps dangerously) It felt like it lost a lot of power Tune and log file from that drive attached. I'm going to stop and wait for some extra advice from you experts before I do something wrong enough for it to get expensive.

Modded VVE dial in tune: 18872
Log of that tune: 18871

[Rockgolem]

On a side note, since my vvt is gone, and I'm running the Ls3 intake/injectors, my L92 has essentially become an LS3. Would it be easier to copy/paste the LS3 VE tables into my tune as a better place to start for VE tuning?

[joecar]

List of pids to log (we may have to cut the list down a little):
- E38.SCALEDUNVE_DMA
- E38.SCALEDCORVE_DMA
- E38.INVLVTMP_DMA
- E38.INJPDELTA_DMA
- E38.APCYL_DMA
- E38.AFRATIO_DMA (you already have this) this depends on B3601
- SAE.MAP (you already have this)
- SAE.MAF
- SAE.LONGFT1
- SAE.LONGFT2
- SAE.SHRTFT1
- SAE.SHRTFT2
- SAE.LAMBDA (you already have this)
- SAE.RPM (you already have this)
- SAE.IAT (you already have this)
- SAE.ETC
- GM.FTC
- GM.HO2S11
- GM.HO2S21
- GM.ETCTP instead of GM.APP
- GM.KNKRET or GM.KNKRET_B whichever one works
- GM.SPARKADV (you already have this)
- GM.MAFFREQ2 (you already have this)
- GM.INJPWB1 (you already have this)
- GM.INJPWB2
- EXT.WO2LAM1 (you already have this)
- EXT.WO2AFR1 (you already have this)

We'll have to define CALC.WO2BEN as "{EXT.WO2LAM1}/{SAE.LAMBDA}".

EXT.DYNLAMBDA is for external lambda from dyno, the value logged is a 0-2V squarewave, looks like it's not valid.

[joecar]

On a side note, since my vvt is gone, and I'm running the Ls3 intake/injectors, my L92 has essentially become an LS3. Would it be easier to copy/paste the LS3 VE tables into my tune as a better place to start for VE tuning?
Yes... but there may be some other tables we need.

[Rockgolem]

I put together the PID selection as requested. I'll create a tune with the LS3 Ve tables with the VE tune (fuel trims/PE/MAP disabled) and run it first thing after work tomorrow with those PIDS you asked for. Thanks a lot for your help so far.

[joecar]

I put together the PID selection as requested. I'll create a tune with the LS3 Ve tables with the VE tune (fuel trims/PE/MAP disabled) and run it first thing after work tomorrow with those PIDS you asked for. Thanks a lot for your help so far.No worries :cheers:

Disable CL/LTFT/STFT and MAF (this is what you meant, right?), but do not disable PE, leave it enabled, it shows up in SAE.LAMBDA and GM.AFRATIO_DMA.


Post your new tune and log file.

Also:
- is the tune in post #1 above the original tune from GM...?
- post your TCM tune file.

[Rockgolem]

Okay, read a step........do a step.......eat a banana. I'm slowly getting this. I took this tune which is the original tune from post #1 (a modified Cadillac Escalade L92 tune) modified with LS3 VE tables 18874 and applied the coefficients I got from this run/log (paste and multiply) 18875. After some pretty novice style blending, I ended up with the VE map in this tune 18876. Then I took that for a half hour drive today and got this log 18877. The thing that boggle me is my WB02 read low (12-13s sometimes dipping into the 10s when cruzing) even though its running richer, the vehicle seemed like it had even less power than it did when running super lean. When I looked to apply that log to the table all of the coefficients were 0.0 =(

Also, it would die at stop lights if the transmission was in gear.

Also, per request here is my transmission file (6l80e with T43 TCM) Its a stock tune with gear ratio and tire sizes changed. 18878 When I had it at the dyno the tuner played with it a bit and ended up locking it in second gear (shift tables for 2<3 were changed to 403mph). He did that so my 40" tires didn't tear apart his dyno. The problem is he never unlocked it so I had to re-load the stock tune when I got back.

Any input to the tune would be well appreciated at this point.

[joecar]

So you did this: from Log_0004.efi (https://forum.efilive.com/attachment.php?attachmentid=18875&d=1443480645) you edited table B8101 in VVE dial in_Camaro VE tables_0001.ctz (https://forum.efilive.com/attachment.php?attachmentid=18874&d=1443480616), generated coeficients, and saved as VVE dial in_Camaro VE tables_0002.ctz (https://forum.efilive.com/attachment.php?attachmentid=18876&d=1443480772) from which you logged Log_0006.efi (https://forum.efilive.com/attachment.php?attachmentid=18877&d=1443480908)...?

ok, I'll take a closer look at it tonite.

Which wideband do you have...?

What elevation are you located at...?

[Rockgolem]

That is a correct statement. I'm running an innovate LC-2 my current elevation is 1,459.7 m or 4800ft. I figured out why my coefficients were 0.0, I wasn't logging my W02 Base Equivillancy ratio. Is there any way to make a map to apply to my VE table from the information I logged or do I need to take another long drive?

[joecar]

Ok, this gives me a good big-picture... tonite we'll cut down on the pid count to get better resolution (more frames per second).

[joecar]

That is a correct statement. I'm running an innovate LC-2 my current elevation is 1,459.7 m or 4800ft. I figured out why my coefficients were 0.0, I wasn't logging my W02 Base Equivillancy ratio. Is there any way to make a map to apply to my VE table from the information I logged or do I need to take another long drive?Should be able to...

make a calc pid for this: CALC.WO2BEN = "{EXT.WO2LAM1}/{SAE.LAMBDA}" <-- I can show you how

if you compare GM.AFRATIO and SAE.LAMBDA you will see that GM.AFRATIO lags behind;

you can add calc pids to a log afterwards (define calc pid in calc_pids.txt, reopen scantool, select the new pid, save log);

did you remember to apply a transient filter (ETCTP varying by less than 5% in 100 ms or similar)...?

[Rockgolem]

I didn't remember the filter. But I built one (research), created my calc pid (more research) and finally got the results of my log file to paste/multiply into my tune. It resulted in some massive spikes. When blending the graph is it generally better to smooth UP to the top of the spikes or to tame the spikes down and then blend into them?

[joecar]

Depends... post current tune/log files.

[joecar]

Ok, we'll have to reduce the pid channel count... I'm thinking over what pids to eliminate...

[statesman]

Joe, you might want to take a closer look at his wideband readings... particularly in Log_0002. There's some bat-shit crazy numbers in there.

[Rockgolem]

I haven't moved from the tune/log in post #14 except to try and apply the coefficients from the log to the tune....which is where I got the giant spikes. Since I didn't know the best way to blend I was waiting on input. The spikes are particularly at low RPM/low MAP they had higher factors.

[joecar]

I haven't moved from the tune/log in post #14 except to try and apply the coefficients from the log to the tune....which is where I got the giant spikes. Since I didn't know the best way to blend I was waiting on input. The spikes are particularly at low RPM/low MAP they had higher factors.Ok, let me play with your files a little...

[joecar]

Joe, you might want to take a closer look at his wideband readings... particularly in Log_0002. There's some bat-shit crazy numbers in there.Are you getting any misfires...?

[Rockgolem]

The engine has not detected any misfires It has backfired twice when coming off WOT though (possibly due to DFCO being disabled for logging).

[Rockgolem]

Ok, we're going to disregard some of those past tunes. I realized the changes to my idle range were due to my huge cam overlap throwing my detected AFR very lean. As I kept modifying the idle kept getting worse (to where it wouldn't even start without my foot on the pedal). I did a bunch of research on blending and redid my VE from the first log/and LS3 tables. What I got was a much better run. Here is the current tune I'm working from and a long log of that (just with VE pids).

18881

18882

When I try to paste the coefficients from my WB02 map I get more crazy spikes in the low RPM/low map and high map/high rpm. also some serious dips in the mid RPM/low map range. This thing looks impossible to blend/smooth. Any suggestions?
18883
18884