VVE epic struggle.

[Hib Halverson]

Well...maybe "epic struggle" is a little hyperbole but it got your attention.

Ok...we've got an LS7 which is stock except for an MSD Atomic Air Force intake manifold. That manifold's volume is 1000 cuin. vs the stocker's 692 cuin. and the runner lengths inside the manifold are different, so it sort of goes without saying that the engine's volumetric efficiency has to have changed. I entered 1000 CID in the calibration file for intake manifold volume.

I've done most of what's recommended to force open loop and I'm in OL because I have an instant MIL due to the MAF failing high and all the FTs are zero.

The first thing I did was just take a WAG and add 15% to the VE numbers at high-load and from 4000-7000 rpm. I did that and the engine was lean from 3000. Commanded EQR in the cal is .86-lambda and in scan data it's .859, yet, from 3000-up the delivered lambda goes from .86 or so to progressively more lean numbers. By 4500 it's .97 and I'm getting significant KR so I aborted the test.

Next I took some data using a lambda error map of RPM vs MAP and saw as much as 1.15 in the high-rpm/high-load areas so, I added 20% in those areas. No change. Then I added 40% and, again, no change.

So far, this happens regardless of what changes I make to the VVE table. I've gone as much as 40% rich in the high-rpm high load with a swing of as much as 55%...I tried one time with 15% more lean just to see what happened...nothing.

This is gonna drive me to drinking.:beer:

There has to be something I'm missing...the noob that I am tells me that.:confused:

Let me know what you think.

Here's a scan of a couple WOT runs from 2000 RPM or so to 4500.

18783

[joecar]

...

I've done most of what's recommended to force open loop and I'm in OL because I have an instant MIL due to the MAF failing high and all the FTs are zero.

The MAF DTC indicates you're in SD... this is regardless of OL/CL... to see OL/CL you have to watch the LTFT's or view the pid FUELSYS A/B.

[joecar]

Are you editing the Virtual VE surface and then Generating the Equations...?

[Hib Halverson]

The MAF DTC indicates you're in SD... this is regardless of OL/CL... to see OL/CL you have to watch the LTFT's or view the pid FUELSYS A/B.
My LTFTs are 0.
The PID "FUELSYS A/B" is not listed in my scanner, either as "selected" or "not selected".
While we're on the subject of PIDs, any suggestion of how to create one for injector duty cycle? I can't make the PID for duty cycle on the "not selected" list work.


Are you editing the Virtual VE surface and then Generating the Equations...?
No.
I'm entering corrections in the "Adjust" window, generating the coefficients, then generating the VVE and, finally, saving the file. The differences appear when I compare the finished file with a previous cal which is not modified for VVE.

Let me know what you think.

[joecar]

Post screenshot of your map.

Post your .ctz file.

[joecar]

My LTFTs are 0.
The PID "FUELSYS A/B" is not listed in my scanner, either as "selected" or "not selected".
When you uncheck Supported on the PIDs tab, and click on the column heading "Parameter" (to sort on this column), and then look down this same column, you don't see SAE.FUELSYS...?



While we're on the subject of PIDs, any suggestion of how to create one for injector duty cycle? I can't make the PID for duty cycle on the "not selected" list work.
Post a log of the injector pulsewidth, I'll find/make a calc pid for DC%.


Ok, don't do this step:


No.
I'm entering corrections in the "Adjust" window, generating the coefficients, then generating the VVE and, finally, saving the file. The differences appear when I compare the finished file with a previous cal which is not modified for VVE.

[joecar]

When you enter the correction in the adjust box, which button do you press to apply the adjustment...?

When you do this, does the highlighted portion of the VVE surface move...?

[joecar]

If you edit the equation, then you click Generate VVE to obtain the surface.

If you edit the surface, then you click Generate Equation to obtain the equation.

i.e. don't do both.

[Hib Halverson]

When you enter the correction in the adjust box, which button do you press to apply the adjustment...?
Multiply


When you do this, does the highlighted portion of the VVE surface move...?
Ah well...I don't know because I have the VVE table full screen. The numbers in the cells change when I apply the adjustement.


If you edit the equation, then you click Generate VVE to obtain the surface.

If you edit the surface, then you click Generate Equation to obtain the equation.

i.e. don't do both.

Up to now, I have not edited the surface in the 3D map.


Post screenshot of your map.

Post your .ctz file.Here ya, go...

1878718788

18789

[joecar]

Multiply


Ah well...I don't know because I have the VVE table full screen. The numbers in the cells change when I apply the adjustement.Ok, so long as the highlighted cells obtain blue dogears (indicating changed cell)...

[joecar]

If you edit the equation, then you click Generate VVE to obtain the surface.

If you edit the surface, then you click Generate Equation to obtain the equation.

i.e. don't do both.Instead of "Equation" I meant to write "Coeffs".

[Hib Halverson]

Ok, so long as the highlighted cells obtain blue dogears (indicating changed cell)...
Yep, they got the blue, dog-ear highlights.


Instead of "Equation" I meant to write "Coeffs".
Figured as much.

Also...I got to thinking: the way that engine is acting when I force it into OL is like it's running out of injectors but...no way. If I put a proper cal back in, i.e.: no forced OL and with the MAF working, indicated lambda from the WB is .86 until I tag the rev limiter...and that's as it should be.

[joecar]

Do this when you get a chance:



Post a log of the injector pulsewidth, I'll find/make a calc pid for DC%.

[Hib Halverson]

On the duty cycle thing, I looked again and did find a PID for duty cycle by bank but it's X'ed out. When I click on more info it says that PID is not valid becasue a required parameter, {GM.IBPW.1} is not selected. When I go find {GM.IBPW.1}, it's X'ed out but, in the more info box, no reason is given. Interestingly, there is another injector pulse with PID with a slightly different name {GM.INJPWB1} and it's listed as a valid PID. Any idea why that situation exists?

I'm thinking I can use {GM.INJPWB1} in a similar expression and get duty cycle so I tried it and got this error when opening Scanner with that PID in my calc_pid file..


List index out of bounds (4).

What does that mean?

[Hib Halverson]

Ok. I figured out the problem and, expectedly, it's one of those "well....duh" situations which comes out of trying to learn both new software and some new tuning methods in a short time.

Turns out, I had the two airflow tables, intake manifold throttle valve (IMTV) closed and intake valve open, reversed in my mind. I was making changes to the IMTV open table when I should have been making them to the closed table. While was changing the former, it was only the latter would result in change in delivered lambda.

Well....duh.

I knew that an engine with no intake valve only used one of those two, but I had the two confused. How did that happen? I could blame it on too much good beer...:cheers:
....but, it was just a dumb-assed mistake:doh2: born of lack of experience with EFI Live and learning to tune the VVE calibration in an E38.

Yesterday, out of curiosity, I started to compare the data in the two tables and that led me to research IMTV tables on this and other forums and that's when I learned I had them "backwards" in my mind.

I quickly did a tune with the closed table having 15% more fuel up top and that made a big difference.:good: Today, I start smoothing out the VVE to better match the MSD Atomic Air Force intake manifold, I'm using.

Not only did I straighten out my confusion on the IMTV valve issue but I also figured out how to make the PID for injector duty cycle. For those interested, it's:


*CLC-00-003
% 0.0 100 0.1 "{GM.INJPWB1} * {SAE.RPM} / 1200"

CALC.INJDUTY F110 CLC-00-003 % Fuel "inj duty cycle"

Lastly, thanks to "joecar" for helping me along with the thought process. I learn slow, Joe, but I'll get it....eventually.

[joecar]

On the duty cycle thing, I looked again and did find a PID for duty cycle by bank but it's X'ed out. When I click on more info it says that PID is not valid becasue a required parameter, {GM.IBPW.1} is not selected. When I go find {GM.IBPW.1}, it's X'ed out but, in the more info box, no reason is given. Interestingly, there is another injector pulse with PID with a slightly different name {GM.INJPWB1} and it's listed as a valid PID. Any idea why that situation exists?

I'm thinking I can use {GM.INJPWB1} in a similar expression and get duty cycle so I tried it and got this error when opening Scanner with that PID in my calc_pid file..



What does that mean?GM.IBPW1 and 2 is for Gen3 LS1/LS6.

GM.INJPWB1 and 2 is for Gen4 and beyond... if there is no duty cycle pid for this I can make one.

( And, I'll find out about that error message )

[joecar]

Turns out, I had the two airflow tables, intake manifold throttle valve (IMTV) closed and intake valve open, reversed in my mind. I was making changes to the IMTV open table when I should have been making them to the closed table. While was changing the former, it was only the latter would result in change in delivered lambda.

Well....duh.

I knew that an engine with no intake valve only used one of those two, but I had the two confused. How did that happen? I could blame it on too much good beer...:cheers:
....but, it was just a dumb-assed mistake:doh2: born of lack of experience with EFI Live and learning to tune the VVE calibration in an E38.

Yesterday, out of curiosity, I started to compare the data in the two tables and that led me to research IMTV tables on this and other forums and that's when I learned I had them "backwards" in my mind.


Ok, I'll make a note of this and make sure the descriptions get reviewed.

[joecar]

*CLC-00-003
% 0.0 100 0.1 "{GM.INJPWB1} * {SAE.RPM} / 1200"

CALC.INJDUTY F110 CLC-00-003 % Fuel "inj duty cycle"


Lastly, thanks to "joecar" for helping me along with the thought process. I learn slow, Joe, but I'll get it....eventually.

Yes, that is correct, good job :cheers:

No worries... looks like you have most of what you need (there are still a few more "features" to learn, but you have enough for tuning).

[Hib Halverson]

Ok, I'll make a note of this and make sure the descriptions get reviewed.

Ah, well...I really don't think there's anything wrong with the descriptions. It was my getting confused about stuff.


GM.IBPW1 and 2 is for Gen3 LS1/LS6.
GM.INJPWB1 and 2 is for Gen4 and beyond... if there is no duty cycle pid for this I can make one.[quote]

Got it.


Yes, that is correct, good jobIt was nice to get something right on this.

[quote]No worries... looks like you have most of what you need (there are still a few more "features" to learn, but you have enough for tuning).Indeed. I just got back from some WOT testing. I added 15% more fuel at high-rpm in high-load cells and lambda error went from 1.15 or .98-1.02 in most of the cells I logged. Gonna write one more version of the cal and go test some more.

Thanks again for your help.

[joecar]

No worries... post some log files so I can play along.

[Hib Halverson]

No worries... post some log files so I can play along.

Uh, well....I tried to upload the log file twice and each time it failed, so I sent it to you in an email. Is there a file size restriction for attached files in a forum post?

[Blacky]

I'm thinking I can use {GM.INJPWB1} in a similar expression and get duty cycle so I tried it and got this error when opening Scanner with that PID in my calc_pid file..


List index out of bounds (4).

What does that mean?

It means there's a fault in the EFILive software that is not handling the text in the calc_pids.txt file correctly. Do you still have the calc_pids.txt file that causes the error? If so can you please send me the file (paul@efilive.com) and I'll try and figure out why that error occurs (and fix it).

Regards
Paul

[Hib Halverson]

It means there's a fault in the EFILive software that is not handling the text in the calc_pids.txt file correctly. Do you still have the calc_pids.txt file that causes the error? If so can you please send me the file (paul@efilive.com) and I'll try and figure out why that error occurs (and fix it).

Regards
Paul

I fixed it. I found out I had some of the text on the wrong line. First, I had it as:

*CLC-00-003 0.0 100 0.1 "{GM.INJPWB1} * {SAE.RPM} / 1200"

%

...but it should have been...

*CLC-00-003
% 0.0 100 0.1 "{GM.INJPWB1} * {SAE.RPM} / 1200"

:doh:

While I got ya, I have another geeky quesiton. Twice now, when I've updated the software, I forget to reload my file length spec into properties. As a result, when I data log, I run to the end of the specified file length and that locks-up my machine, I get the "not responding" message and I have to force Scanner to quite.

Isn't there a way to allow the user to save then quit when he/she hits the file length limit?

Of course...the solution is to change that in properties, but :doh:I keep forgetting.

[joecar]

The scantool should remember the properties across software updates.

[Hib Halverson]

The scantool should remember the properties across software updates.

I'd have thought so, too, but last week, I had problem with the 50,000 frame limit and lost a whole test session. So, I set properties at 250,000 frames. No problems after that, until today, which is after I installed the 27 Aug update. I lost another test session of data when it, once again, hit the 50,000 frame limit. This time I set it at 500,000 frames.

Hopefully, the next time I update the software, I'll remember to check properties before I take data.
(sigh)

[joecar]

I'll bring this to Paul's attention.

[Blacky]

I'd have thought so, too, but last week, I had problem with the 50,000 frame limit and lost a whole test session. So, I set properties at 250,000 frames. No problems after that, until today, which is after I installed the 27 Aug update. I lost another test session of data when it, once again, hit the 50,000 frame limit. This time I set it at 500,000 frames.

Hopefully, the next time I update the software, I'll remember to check properties before I take data.
(sigh)

Updating the software does not (or should not) alter the property settings. If you change any settings and the software crashes (or you kill it using Task Manager or by switching off or rebooting the PC) then the changes won't be saved because the changes are only saved when the software exists cleanly.

FYI the maximum data log size setting is stored in the registry here:
18795

Regards
Paul