Making progress - need some logs reviewed

[rb1]

Hi all, need some guidance on these logs. Have an LS motor swapped into an older truck, and got it running, but just scratching the surface with the tune now. My background is heavy into carburetors and distributors... so alot of the learning curve here is knowing what parameters to log and work on, and for that I will need your help!

First off, reading a TON and have made some adjustments.

I had decided to starting with the RAFIG and also then did the CALC.VET.

The truck is on a chassis dyno, and using an NTK AFX wide band for the logging. It is an M6 combo, and has an unknown camshaft, but idles around 66 KPA (I can get the KOEO KPA, but we are at 4500 ft).

I did one round of RAFIG, and three of CALC.VET.

The LTFT's look odd to me on the logs, can you guys give me some opinions on that and anything else that you see on the logs?

Posting them shortly,

THANKS!!

[rb1]

15784This first log is the third round of a CALC.VET.

[rb1]

The AF ratio looked OK at WOT, so put a bit of load on it with a couple of pulls, this is the log. Also some idle time in there.. any advise would be GREAT!15785

[rb1]

Here is the tune file, slight changes with the RAFIG, and then a couple of tweaks on the VE table... yeah it needs some smoothing still.
15786

[joecar]

Post screenshots of your Calc.VET scantool maps (MAF and VE).

Post your calc_pids.txt file.

[rb1]

15792
15793
15794

Ok, here are screenshots of what I think you are asking for, and the attached calc_pids.txt file, - just have to resize screenshots,

thanks!

[Woodchukka]

You have given Joe screen shots of your EFILive Tune maps. What Joe wanted to see (as I read it) are the maps from the v7.5 EFILive Scan Tool. The maps you use when data logging.

[rb1]

You have given Joe screen shots of your EFILive Tune maps. What Joe wanted to see (as I read it) are the maps from the v7.5 EFILive Scan Tool. The maps you use when data logging.

duh!

Sorry about that. Here they are. THis is the second round VET on the dyno, I already see I should have smoothed (or rather filled in the missing cells) things out better on the map before copy and pasting in to B0101.

Please let me know if I am off track here, or what you guys think of the data.

THANKs!

1579515796

[Woodchukka]

There is a spread sheet that I have used from the forum. Have a read of this thread.

http://forum.efilive.com/showthread.php?4399-VE-Table-Interpolation-SpreadSheet&highlight=interpolation

The adjustments to the VE table will be erratic also if you have not hit a particular cell enough times and obviously that will make the table erratic. The values should be an average of the accumulated data (not a one off number) which will give more accurate numbers for the adjustments.

As for how it looks I am not familiar with the numbers as I use Grams*Kelvin/kPa so my numbers look different however the numbers will depend on my point above.

[rb1]

There is a spread sheet that I have used from the forum. Have a read of this thread.

http://forum.efilive.com/showthread.php?4399-VE-Table-Interpolation-SpreadSheet&highlight=interpolation

The adjustments to the VE table will be erratic also if you have not hit a particular cell enough times and obviously that will make the table erratic. The values should be an average of the accumulated data (not a one off number) which will give more accurate numbers for the adjustments.

As for how it looks I am not familiar with the numbers as I use Grams*Kelvin/kPa so my numbers look different however the numbers will depend on my point above.

THANKS! I will check that spreadsheet out, sounds just like just what I need.

I spent more time on it today, and it runs quite well so far, tinkering with idle tricks and the cam is unknown but was told it is pretty big. Idles rough but stable so far.

Regarding my values on the VET map... I probably should change that over to the G*K/kPa because I constantly have to go in and enter the engine displacement... I will have to look into how to change that.

My main concern is that I am doing the CALC.VET correct... it seems I am, but I need some confirmation. The thing that I am worried about right now is the goofy LTFT I am seeing on bank 1.. Not sure what is going on there. Always seems to be in the positive side of correction. Any tips or tricks there? Here is the latest log and the latest tun

15799
15800

[rb1]

When doing the CALC.VET, with a Wide band.... are the LTFT from the narrow band sensors contributing? SUch as in my case?

I think the info (tune, calc.vet) should be in my first few posts in this thread.

[joecar]

Yes (look at the pid SELBEN).

[rb1]

Yes (look at the pid SELBEN).

Ahhh, will have a look at that closer.

So I had better tidy up my LTFT on bank 1 that is showing +25 before going any further I bet... The O2 sensors do switch, but they get lazy, especially on bank 1. This is a new install, with short tube headers and the bungs are right after the header... I should swap the sensors bank to bank and see if the errant LTFT follows?

I also have a wide band in each pipe also, about 12" downstream of the narrow band sensors. I have only been using the wide band on the passenger side bank, perhaps that is bank 2. The numbers all look OK that wide band.

thanks

[rb1]

OK, just reviewing my printout from the CALC.VET Tuning Tutorial. I was under the assumption that I was using my wide band for all the corrections during this step... I guess my question in post 11 was just to confirm that. But you are saying that it is in fact using the narrow band data?

[rb1]

I just swapped the o2 sensors side to side and bank 1 still shows a lazy o2 number... the thing is it is mostly just at idle. THe LTFT and the o2 switching speed seems to be alot better at off idle engine speed.

So I went back in the tune, and I had modified the misfire data (c5621 and others around there) to pretty much avoid the 0300 code, but in my data log I could not see the per cylinder misfire counts.... so I went back in and put the stock values back in there and logged a couple of minutes of idle and revving the engine a bit. Seems like the misfire counts are going up super fast at idle, and it does not seem to be just one cylinder, it is a bunch of them adding to the counts.

I think this has a ton to do with my bank 1 LTFT issue, thoughts? How can I get around this? The idle is very choppy, but it is pretty stable.

Here is a couple of logs
15805
15806
15807

[joecar]

...

So I had better tidy up my LTFT on bank 1 that is showing +25 before going any further I bet... The O2 sensors do switch, but they get lazy, especially on bank 1. This is a new install, with short tube headers and the bungs are right after the header... I should swap the sensors bank to bank and see if the errant LTFT follows?

...Fix this, the LTFT should not be pegging like that (it is a symptom of an airleak). Yes try swapping the O2's across banks and see if the problem follows or stays.

[joecar]

OK, just reviewing my printout from the CALC.VET Tuning Tutorial. I was under the assumption that I was using my wide band for all the corrections during this step... I guess my question in post 11 was just to confirm that. But you are saying that it is in fact using the narrow band data?Yes, SELBEN uses LTFTBEN while in closed loop and WO2BEN otherwise.


You could also do this: disable CL/LTFT/STFT/SOL and edit calc_pids.txt so that CLC-00-305 uses CALC.WO2BEN instead of CALC.SELBEN... and then use SELBEN in your MAF scantool map, and use CALC.VET (which is defined by CLC-00-305 which you just edited) in your VE scantool map.


If you can, use g*K/kPa for VE units (set tunetool to display VE in g*K/kPa).

[rb1]

Thank you Joe, you certainly are an incredible source of knowledge here!!!!

"Yes, SELBEN uses LTFTBEN while in closed loop and WO2BEN otherwise."

that is a huge part of my missing puzzle here. thanks.

So, while doing more research and thinking since my last post, here is a thought I had about this ltft issue:
When watching my logs, I see that all my misfires and LTFT trouble (+25% bank 1) are at idle. Once rev'd up, the misfire counts go away, and the LTFT's come in line again. I believe this idle only issue is caused by the camshaft overlap causing the misfires, and perhaps also the locations that the Narrow band sensor on bank 1 is just a tick off where bank 2 is located. Both are about 20 inches after the shorty header collectors, but perhaps the clocking in the pipe is a bit different making bank 2 react different? (not as sensitive?). I did switch the 02 sensors from side to side and the problem stayed at BANK 1.

So, because of this, maybe I will never be able to get the LTFT issue handled at idle, and I was doing some reading about just turning the ltft off for good, seems this is called SOL, semi open loop? And uses STFT only? Good idea to try?

[rb1]

double post sorry

[rb1]

You could also do this: disable CL/LTFT/STFT/SOL and edit calc_pids.txt so that CLC-00-305 uses CALC.WO2BEN instead of CALC.SELBEN... and then use SELBEN in your MAF scantool map, and use CALC.VET (which is defined by CLC-00-305 which you just edited) in your VE scantool map.


If you can, use g*K/kPa for VE units (set tunetool to display VE in g*K/kPa).

OK, I am going to try this...

[rb1]

and then use SELBEN in your MAF scantool map, and use CALC.VET (which is defined by CLC-00-305 which you just edited) in your VE scantool map.

Can I get clarification on that please, I did open up my calc_pids.txt and replaced CLC-00-305 as you instructed, but a bit lost on the above quote...

many thanks!!!!

[rb1]

OK, I also am following another post where you said to do this, which I think is to eliminate all trims (LTFT and STFT), and I am going to do this to mine also:

CL: B3801 disable.
LTFT: B4205 disable by setting temperature to max.
STFT: if B4206 exists, disable.
SOL: if COS, then in B3647/B3649 set stoich cells to non-stoich (if stoich EQ 1.00, set to EQ 1.01 or EQ 0.99).

[rb1]

edit - I my OS does not have the B3647 or B3649, do I need to change something else insteadÉ

[joecar]

No problem, that just means you have a non-COS.

[joecar]

Thank you Joe, you certainly are an incredible source of knowledge here!!!!

"Yes, SELBEN uses LTFTBEN while in closed loop and WO2BEN otherwise."

that is a huge part of my missing puzzle here. thanks.

So, while doing more research and thinking since my last post, here is a thought I had about this ltft issue:
When watching my logs, I see that all my misfires and LTFT trouble (+25% bank 1) are at idle. Once rev'd up, the misfire counts go away, and the LTFT's come in line again. I believe this idle only issue is caused by the camshaft overlap causing the misfires, and perhaps also the locations that the Narrow band sensor on bank 1 is just a tick off where bank 2 is located. Both are about 20 inches after the shorty header collectors, but perhaps the clocking in the pipe is a bit different making bank 2 react different? (not as sensitive?). I did switch the 02 sensors from side to side and the problem stayed at BANK 1.

So, because of this, maybe I will never be able to get the LTFT issue handled at idle, and I was doing some reading about just turning the ltft off for good, seems this is called SOL, semi open loop? And uses STFT only? Good idea to try?Lol, thanks for your kind words.


Note that misfires show up as lean (O2 sensors detect excess oxygen) which is what might be causing your bank1 LTFT to peg positive...

since it is on one bank only, there is a problem (either misfiring for real, or leaking air)...

this has to be fixed before proceeding...

(are you losing coolant, do you have air leaks on bank1, do your injectors leak on bank1... i.e. what can be causing a misfire on bank1 at idle)

you may have to jump in and do RAFIG/desired air.

[joecar]

and then use SELBEN in your MAF scantool map, and use CALC.VET (which is defined by CLC-00-305 which you just edited) in your VE scantool map.


Can I get clarification on that please, I did open up my calc_pids.txt and replaced CLC-00-305 as you instructed, but a bit lost on the above quote...

many thanks!!!!

First, edit CLC-00-305 in calc_pids.txt as you have done (using Windows Notepad).


Then, in the scantool, on your MAF map, in the map properties, goto the Data tab, use WO2BEN instead of SELBEN as the pid... (and remember to save the map)...

i.e. the MAF map will now use WO2BEN to correct the MAF table;


and the CALC.VET now uses WO2BEN (from the edited CLC-00-305) to calculate a new VE table from the corrected MAF values.

[rb1]

Lol, thanks for your kind words.


Note that misfires show up as lean (O2 sensors detect excess oxygen) which is what might be causing your bank1 LTFT to peg positive...

since it is on one bank only, there is a problem (either misfiring for real, or leaking air)...

this has to be fixed before proceeding...

(are you losing coolant, do you have air leaks on bank1, do your injectors leak on bank1... i.e. what can be causing a misfire on bank1 at idle)

you may have to jump in and do RAFIG/desired air.

Oh, you deserve those kind words!!! I have been searching day and night getting up to speed (or at least trying to!!) and your name is on this forum back I think to one post I saw in 2005! That my friend is commitment!!

Back to this deal.... No coolant leaking (level has not changed), injectors... well I don't "think" they are leaking, but the thing that is getting me here is how just off idle the misfires go away and all is well in it's world.

I did do several RAFIG's (each day I fire it up I let it idle and log the RAFIG for 25 mins or so), it starts very well, idles very stable (at 60 kPa), I was just thinking that the placement of the o2 sensor (done very unscientific regarding its placement) may just put it in a spot where it is not reading the same "stream" as bank 2, making it more / or less? sensitive?? AM I crazy... wait, don't answer that.

[rb1]

First, edit CLC-00-305 in calc_pids.txtas you have done (using Windows Notepad).


Then, in the scantool, on your MAF map, in the map properties, goto the Data tab, use WO2BEN instead of SELBEN as the pid... (and remember to save the map)...

i.e. the MAG map will now use WO2BEN to correct the MAF table;


and the CALC.VET now uses WO2BEN (from the edited CLC-00-305) to calculate a new VE table from the corrected MAF values.

OK, onto this deal... THANKS for the walkthru... I actually figured out the SELBEN one this afternoon, I changed the data tab to "BEN from WB (factor)" and that gave me some believable numbers to correct the B5001 with.

But I am still stuck on the CALC.VET one... Sorry, I may be having a real dense moment here, I have two maps set up to do CALC.VET, the SELBEN one, and the other is labeled VET and uses RPM on the rows and MAP on the columns, and the parameter is "calculated VE, BEN corrected (g*k/kPa)

I don't see WO2BEN or anything else like that in the parameter drop down menu...

Where am I going wrong here?

THANK YOU!!!!!!!!

[rb1]

OK, forgot to attach the calc.pids.txt file just to make sure I have not messed that up somehow

[rb1]

the MAG map will now use WO2BEN to correct the MAF table;

OK, was re-reading this and I am not sure what a "MAG" map is?

thanks

[joecar]

Calc.VET procedure consists of these two things

(in your case modified to be in OL and to use WO2BEN instead of SELBEN):


- B5001-like MAF map: corrects existing MAF using the pid CALC.WO2BEN,

- B0101-like VE map: calculates a new VE table using the pid CALC.VET (CLC-00-305)(which uses corrected MAF to calculate the new VE).

[joecar]

OK, was re-reading this and I am not sure what a "MAG" map is?

thanksMy fat-fingered typing... it is supposed to say MAF.

[rb1]

My fat-fingered typing... it is supposed to say MAF.

10-4, thanks!

Did my calc_pids.txt look OK?

I must have something squirrel'd up here.... I did a couple of CALC.VET logs today, now with Open Loop, and using only the wide band. It seems to me like the after a few corrections, I was getting leaner and leaner, into the mid 15's on the wide band. (not WOT)

I really am trying to understand the math involved here, and have come a LONG way in the last week, sorry for the dense questions.

Thinking that my maps may be wrong somewhere, so I have some screenshots of them to get some opinions on,

thanks!!

[rb1]

Here are the resulting log files....

[rb1]

So when trying to figure out the math, I take this screen shot of a frame in the CALC.VET log I was working on today.

I see the Wo2BEN says .9732 factor... is that the key number in the correction?

Is the idea to look at the GM.AFR (14.68) and compare to CALC.AFR_NGK1 (14.22), but that does not equal .9732. (14.22 / 14.68 = .9686)

Where am I going wrong here?

[joecar]

10-4, thanks!

Did my calc_pids.txt look OK?

I must have something squirrel'd up here.... I did a couple of CALC.VET logs today, now with Open Loop, and using only the wide band. It seems to me like the after a few corrections, I was getting leaner and leaner, into the mid 15's on the wide band. (not WOT)

I really am trying to understand the math involved here, and have come a LONG way in the last week, sorry for the dense questions.

Thinking that my maps may be wrong somewhere, so I have some screenshots of them to get some opinions on,

thanks!!Yes, your calc_pids.txt is correct (I see your have NGK wb).

The math goes like this (in steps):
1. use WO2BEN and/or LTFTBEN to correct the MAF table;
2. use the corrected MAF table to calculate a new VE table (using the Ideal Gas Law);
3. apply 1. to the MAF table by multiplying it into B5001;
4. apply 2. to the VE tables by copying it into B0101.

As always, and use paste-multiply-with-labels and paste-with-labels so that the axes get aligned correctly.

As always, hide the low count cells, and apply the transient filter.

Also note that VE kicks in for transients below B0120 (rpm)... you could also set B0120 to 400 rpm, but take note that the transient filter happens to remove any VE contribution.

NOTE: in your maps, on the Data/Col/Row tabs, checkmark Show Units (it helps when you see the units on the axes).

[joecar]

Here are the resulting log files....MAF map looks great.


The VE map looks reasonable (it rises in the right places), I thought it could be a little higher... can you paste it into your B0101 and compare with the pre-Calc.VET B0101 to see how different it is.

Questions:
- what engine and mods...?
- what injectors...?
- is FPR manifold-referenced or un-referenced...?
- post an image of your B4001.

[joecar]

So when trying to figure out the math, I take this screen shot of a frame in the CALC.VET log I was working on today.

I see the Wo2BEN says .9732 factor... is that the key number in the correction?

Is the idea to look at the GM.AFR (14.68) and compare to CALC.AFR_NGK1 (14.22), but that does not equal .9732. (14.22 / 14.68 = .9686)

Where am I going wrong here?Yes, 0.9732 is the correction factor (multiplier) at that operating point.


WO2BEN is defined as {GM.EQIVRATIO} * {CALC.AFR_NGK1}/14.57

where:
- {GM.EQIVRATIO} = commanded fueling = B3601 / {GM.AFR},
- {CALC.AFR_NGK11}/14.57 = wideband Lambda'
- EQR (equivalence ratio) = 1/Lambda,
- Lambda and EQR are not dependent on stoich AFR (PCM, WB, fuel all assume different stoich AFR's),
- {GM.EQIVRATIO} comes directly from the fueling tables (when you view them in EQR units),
- {GM.AFR} sometimes has minor little modifications to it (i.e. does not always seem to follow {GM.EQIVRATIO}).

[joecar]

What is your B3601 set to...?


Note:
- in CL your NGK wideband should be reporting AFR 14.57 (which corresponds to wideband EQR 1.00 or wideband Lambda 1.00).
- in CL your PCM should be commanding B3601 as the AFR (which corresponds to commanded EQR 1.00 or commanded Lambda 1.00).

[rb1]

Replys below, did not realize caps lock was on... not yelling, LOL.


Yes, your calc_pids.txt is correct (I see your have NGK wb).

The math goes like this (in steps):
1. use WO2BEN and/or LTFTBEN to correct the MAF table;MAKES SENSE TO ME
2. use the corrected MAF table to calculate a new VE table (using the Ideal Gas Law);I FOLLOW THAT TOO, BECAUSE THE PARAMETER BEING USED IS CALLED CALCULATED VE, BEN CORRECTED G*K/kPa
3. apply 1. to the MAF table by multiplying it into B5001; GOTCHA
4. apply 2. to the VE tables by copying it into B0101. GOTCHA

As always, and use paste-multiply-with-labels and paste-with-labels so that the axes get aligned correctly.YUP

As always, hide the low count cells, and apply the transient filter.YUP

Also note that VE kicks in for transients below B0120 (rpm)... you could also set B0120 to 400 rpm, but take note that the transient filter happens to remove any VE contribution. WILL LOOK INTO THIS

NOTE: in your maps, on the Data/Col/Row tabs, checkmark Show Units (it helps when you see the units on the axes).AHHH, THANKS FOR THAT

[rb1]

Regarding B0120 - it is/was set to 400 right now

[rb1]

Answers in bold below


MAF map looks great.


The VE map looks reasonable (it rises in the right places), I thought it could be a little higher... can you paste it into your B0101 and compare with the pre-Calc.VET B0101 to see how different it is.

WHen I compare B0101 original to after this round of CALC.VET, with it in OPEN LOOP now, the map is significantly lower. When I run the engine with this map, the Wide Band numbers are quite lean, like in the mid to high 15's. Not sure why that is...

Questions:
- what engine and mods...? - purchased used, so a bit unknown, but 5.7L (2001 LS1 out of corvette), with a cam. Cam numbers are unknown but I have made a few pulls on it and it still has not made peak power by 5900 RPM. I do not have much LS experience, so maybe that is not out of the ordinary
- what injectors...? they are stock from what I was told, they say 24508208
- is FPR manifold-referenced or un-referenced...? there is no line to the fuel pressure regulator
- post an image of your B4001. b4001 as below

[rb1]

what is your b3601 set to...?


Note:
- in cl your ngk wideband should be reporting afr 14.57 (which corresponds to wideband eqr 1.00 or wideband lambda 1.00).
- in cl your pcm should be commanding b3601 as the afr (which corresponds to commanded eqr 1.00 or commanded lambda 1.00).

b3601= 14.68

[rb1]

Is this a good way to show the calibration difference?

I like to compare things by percentage, so this is comparing B0101 on the stock tune to the CALC.VET applied tune, which it seems is pulling alot of fuel out.

I used the drop down - CALIBRATION menu to pick "show absolute percentage differences" and here is a screen grab:

Just to be clear, with doing the calc.vet in open look, no LTFT or STFT, I seem to be getting too lean of a correction, moving the wide band into the mid to high 15' when the tune was applied,

THANKS

[joecar]

Post your most recent log file and most recent tune file (and calc_pids.txt again so I don't have to go back looking for it).

[rb1]

HI, thanks for hanging in there.

This tune 15826 (0011) is too lean, by about 1.0 AFR (running mid to high 15's on the wide band)

Here is the VET file that got me there 15827

A bit of smoothing applied, was going to run the logging a few more times for sure.

Here is the calc_pids.txt file15828

thanks

[joecar]

ok, I'll look at your files when I get home shortly.

[rb1]

ok, I'll look at your files when I get home shortly.

Cant thank you enough for sticking it out here Joe. It is a tough deal to get up to speed on.

[rb1]

Update. I just could not get this ltft discrepency out of my head so i was going to move the O2 bungs closer and had to drop the exhaust and guess what i found.... I guess the exhaust shop missed a spot!!! I am surprised that i could not head anything, but ironically i has grabbed a stethescope to listen any small header leaks!

So the ltft are pretty equal now, but my progress in the tune is all lost because this was the bank my wide band was in.

[joecar]

Cant thank you enough for sticking it out here Joe. It is a tough deal to get up to speed on.I'm running behind, I could only take a quick glance, I'll take a deeper look later today.

[joecar]

Update. I just could not get this ltft discrepency out of my head so i was going to move the O2 bungs closer and had to drop the exhaust and guess what i found.... I guess the exhaust shop missed a spot!!! I am surprised that i could not head anything, but ironically i has grabbed a stethescope to listen any small header leaks!

So the ltft are pretty equal now, but my progress in the tune is all lost because this was the bank my wide band was in.That weld looks very poor (poor bead, poor deposit, burn thru pipe).

[rb1]

That weld looks very poor (poor bead, poor deposit, burn thru pipe).

Yeah, pretty crazy. I have not had the wide band back on it yet, just cruised around while catching up on some other things in life LOL. Plan on working on it some tomorrow I think. Now the ltft's are in the negatives, and tracking somewhat close to one another. No errant +25% anymore that is for sure.

Just to be clear, I did move the o2 location, as well as welding the hole up around that one bung.

I am hoping that now, I have gotten more familiar with process, that I should be able to make some real headway in a much shorter amount of time.

I am getting some Knock Retard at part throttle, so I will have to map the KR, MAP, and TPS to see where that is happening also.

stay tuned, no pun intended, lol.