getting things straight: genIV LMG, E38, wideband O2 and the VVE table

[harascho]

what a long title... but it tells it all

I installed a LC-1 wideband did some logs and tried to have a close look at the VVE table.
From the first log it seems I am running in the area of 14,2:1 average, pretty rich...

I did as described in Swingtans E38 VVE dial in guide.

Some comments:

I did step by step as described, disabled the LTFT, STFT, DFCO and I set the cruise and PE AFR as in the tutorial.
Then I disabled the MAF and started looking at the IAT compensation. As it is still cold in Germany I only could manage to get an IAT log with temps from 5deg Celsuis up to 30deg C. I could not see any drift in the WB AFR so I was ready for the next step.

I could not find parameter B8101 as on page 9. (V2 plugged into Laptop) Only B8001 showed in the search.

Then next thing after preparing the .tun file as in the tutorial and disabling the MAF.. the engine started and died right away.

So I am confused.. ( as usual.. don't worry..)

Any ideas what I might have mixed up?

Harald

[swingtan]

If the engine started then died, it's probably still using the MAF readings for air flow metering. By disconnecting the MAF, you are now telling the ECM that there is no air flow, so it doesn't add any fuel. To fix this, ensure that the MIL code settings are correct....


Under DTC Type, set P0101 to P0103 to "C: Non Emissions"
Under DTC MIL, set the same parameters to "No MIL"


If you have physically disconnected the MAF, then the stock settings for C0301 to C0312 should be fine to fail the MAF. If you disconnect the MAF and alter the MAF test settings, you may end up with a condition where the ECM thinks it's still working fine. Do a sanity check on the MAF parameters and ensure they will work OK.

For the reference to B101 on page 9, I think you have an old version of the guide. The latest version is 1.9 and has a few corrections in it.

[harascho]

just in time.... I will check that out tomorrow.
I'll keep you posted.

Harald

[harascho]

in the latest version I found:

Parameter: B8020
The stock settings here seems to assume that the MAF will work for all RPM ranges over 3000 RPM. This won’t work when the MAF is turned off. Reset these boundaries to something that gives you a series of more sensible zones for your expected RPM range.
Look at where you would normally be driving and set the zones accordingly. Something like 2000, 3500, 4500, 5500 RPM may work well. This table has a big effect on the VVE shape.

Should I use here the rpm's I touch most in normal driving situations?

that would be 1000 -1800 - 3000 - 3800 for example. It's a rare occasion that I touch the 3000's


Parameter: B8021
As with the previous setting, this helps to shape the VVE table. Adjusting the setting to make the zone match expected driving conditions will help a lot. The stock settings are pretty good, but you may want to drop the “MAP 1” values to help set the idle airflow.

NOTE: At this point, ensure that the “UNITS” set up for B8021 match with B8001. In other words, make sure both tables are using kPa or PSI. Mixing these up will result in the VVE zone boundaries not being where you expect them to be.

Before moving on, you should do 3 things…
1. Go to B8101, still searching that B8101.. press <F4>, select all cells and “copy with labels”
caught B8001 and tried to copy with labels and had an error message...

Am I looking in the wrong places?

[swingtan]

B8020: I'd set those so that the RPM ranges you normally drive at, are inside the zones. Don't make the "normal RPM range" the boundary as you want those ranges as close to the middle of the zone as possible.

B8101: Have you started Tune Tool with the V2 connected? Note the paragraph just above the section titled "Parameter: B8020". If you do not start the Tune Tool with the V2 plugged in, the VVE tables will not show up.


Simon

[harascho]

hi, I worked a little more on the VVE Table, but I am still confused with the preparation.

As in the tutorial I copied the B8101 table then modified B8020 and B8021 and saved that progress in a new file name. Next step was I opened that new tune file , went to B8101 and pasted the original B8101 table back in place.
here are pics of the original table and from the then different table. From a first look I don't see any difference at all. Did I miss a step ?
Next was generate Coeffs.....
I recieved an error message like:
" Zone #12 is too small. Zones must ne at least 3x3 cells"
where is my fault? Ideas?

Harald

[swingtan]

It looks like you have set the lowest RPM range too small. The way I chose the zone boundaries was to look at what idle RPM I have, and make that near the centre of the first RPM zone.

So say my idle was 700 RPM, I set the lowest RPM boundary at 1000 RPM.

Then I did the same for cruising, say I cruise at 1600 RPM, so I'd set the next boundary at around 2500 RPM.

I then jumped to the end of my RPM range, so say I shift around 6200, then the highest boundary should be below my shift point, so I might make that 6000.

the last boundary I just did an average to make the graph look straight. as I said in my previous post, do not make the zone boundaries at your normal engine operating points.

Another tip is not leave the MAP boundaries the same across the RPM zones. Unless you have a reason to have changes between RPM zones, there should be no reason to change them. As mentioned in the guide, the stock settings are pretty good and may not need any changing.

Finally, you need to copy the VVE data in B8101 from the tune before you altered the zone boundaries. Basically you are going to put the stock VVE data back into the changed zones to correct the shape. the final result of this step should be a VVE table that looks the same as the stock VVE.

Simon.

[harascho]

swingtan, with your help I was able to get it right.. logging will be today..

What response speed of the LC-1 do you use normally? The instant response seems to bee too fast and the signal is pretty fuzzy with this setting (I wired the LC-1 with a good GND source....only the cable shield from controller to my V2 connection is not connected at the V2 side. Should there be a connection from shield to GND or is this connected in the controller?.. I bought my LC-1 used so I don't know how the wires ended in the OEM version..)

Harald

[swingtan]

I don't use an LC-1 so I can't comment. I use a TechEdge and run the serial connection into the V2. Post up a log file and I'll have a look at it for you.

Simon

[harascho]

I will post some logs later I am still stuck in my office...
It will take me some more hours to come back with some logs.
What time difference do we have ( Germany - Melbourne/Australia)?


Harald

[harascho]

I did the first logs MAF-less and here is what I got.

I had a hard time filling the cells of my BEN table... not an easy task to perform



The Tahoe ran fine with the settings of the VVE tutorial thanks to the good work of Simon. The only thing I noticed was a very harsh 1 -2 shift.

Harald

While preparing the attachment I took a look at the log... I still have some LTFT activity? Only a small amount, but still there..... To be on the save side I also posted my tun file...

[swingtan]

OK, here's what I'd do.


First, fix the knock. Drop the timing in those cells to stabilise the spark. Is that a stock spark table? It looks very high. Then when you look at your IAT spark table you are adding yet more timing.
Next, when logging TP% select "ETCTP" and you should then get a true 100% WOT reading rather than 83%.
When logging for VE corrections, be gentle with the throttle and try and hold it steady for as long as possible. Don't go stomping on it all the time as it doesn't give good data.
The VVE looks rather bad in that tune. Is that how the stock one looked? If not, then copy the stock values over again, don't forget to generate the coefficients and VVE before saving.


You should be right to try adjusting the VVE after that.

Simon

[harascho]

Ok Simon,

first the knock...I will take care of this first. I advanced the timing and I missed a high load testdrive after that. I only adjusted the high octan table and kept the low octan table stock for safety if I ever advance too far.

Any comments on the LTFT which are still about 4% in that log...I expected them to be zero??

The VVE in that tune.. I copied the original one in that table with the changed boundarys. For me the original VVE looks odd in the beginning..??

Harald

[harascho]

Here are pics of the original VVE and the VVE according to the tutorial which will be used for data logging. I checked all of the tutorials points a second time and I ended with that shape of the VVE.

Any ideas why my LTFT are still active?

To the spark.. I worked on the high octan spark table and reduced the timing in the cells with knock.

[swingtan]

Hi Harald,
I'm not sure why the LTFT's are displaying like that but it may have something to do with them not being reset. Try clearing the fuel-trims from ScanTool...


Connect the laptop to the car via the V2, ignition on.
Open ScanTool and connect to the vehicle.
Go to the DVT tab ( F12 ) And click on "Activate".
In the DVT window, click on "Fuel System"
On the top row, click on "Fuel Reset" under "Fuel Trim Reset".


The stock VVE tables are also rather bad to start with. I've had a play and tweaked them a little to reshape them ( read over written them with near stock Holden 6Lt data ) which should be a bit better. Being the 6Lt version they will be a bit rich, but it's better to be rich than lean.

Last point here, you have AFM active on the tune, so you will want to turn that off in your tune. I think that setting B0921 to "NO" should do that. Once you have set up the normal VVE, you can turn it back on and then work on the AFM VVE.

Regards,

Simon.

[harascho]

You are a very big help for me. I will try the modified tune this evening after work. The LTFT are strange 'cause they changed in my log althought they should be disabled. I will check if resetting them will the trick.

Harald

I am curious how the VVE table looks now..

[swingtan]

It looks very much the same as in the VVE guide, as I used the same one for that ;)

[harascho]

I did a nice testdrive today. After some adjustments to the High octan spark table KR is alomost 0 but with the modified tune file from you (basicly my file with your VVE map..) my Tahoe starts and dies immediately.

I compared the two files, the only differences are B8001 and B8051 (I copied your VVE map into my tune file before..). I flashed my tune file and the Tahoe ran..

To be on the save side I post the two tune files here together with my logged file.

Harald

[harascho]

I took a second look at the two tun files. In the Tuning tool I could take your values (VE coefficients B8001 and B8051) and store them in my tuning file which works in my Tahoe.
Here is a log I took a few minutes ago and also the tun file

Harald

[swingtan]

That's looking pretty good now.

In the log file you can see it's running a bit rich, but it's pretty consistent. Try the Auto VE process with that data and see how the VVE drops in the logged areas. Then smooth out the big spikes and dips, as well as the borders were your data ends and the original VVE remains. It'll be fairly obvious once you have applied the data. You don't want any big steps in the VVE.

Another point to remember is that the generation of the co-efficients will do a fairly heavy smooth across the changed cells and zones. If you have large changes in single zones, it will effect all the data in that zone. So you can cheat a little and alter any cells that have had no data applied, by setting them to similar values to the other cells in the same zone.

It looks like you're on the right track though.

Simon

[harascho]

Hi Simon,

which of my logs do you mean? The last one with ALL the data from your modified tun file ( VVE table plus the coefficients) or the first log with my coefficients and your VVE table....

Harald

[harascho]

Could it be that I just forgot to generate the coefficients after I pasted your VVE table in my tun file....??

So in the second tun / log I had your coefficients, ok not the normal way by hitting the coeff button in the VVE table but rather by taking them from your tun. The result should be the same I guess.

I ok pls forgive my my silly questions from time to time..... I am on my way..

Harald

[swingtan]

You always need to "generate Coefficients" after modifying the VVE. If you don't the changes don't get written to the file. I was talking about the last log file. It looks pretty good for VVE work.

Simon

[harascho]

Hi Simon,

then I better start over again.

this is what I will do later the day:

I will take my tun file, copy and paste your VVE table into it.
Then after I generate Coefficients the B8001 and B8051 should look like the ones in your file.
This will be my starting point.
Next will be a test drive with data logging. I also will try to get some data at higher rpm ranges by driving in lower gears ( I don't live in an area where high speed runs are possible..).

Then we will see what results this brings.

You think I could try the AutoVE process.. ok will give it a try.

harald

Thanks for your patience with me Simon..... I hope this thread will help other starters..

[swingtan]

No problems. Using lower gears for VVE tuning "may" result in a slightly richer mixture in higher gears. It's not a big problem though if you mainly drive in those gears and are willing to run a tad rich on the odd occasion you may go to a track. It's all tuning for what you expect the car to do.

[harascho]

Hi Simon,

I did a quicker run today.
Ok what is the next step? Should I apply the BEN_LC11_E38 Map to my original VVE map or should I try to modify your VVE table?
Is there a need to apply any filters to the BEN LC11 E38 map first as in the AutoVe tutorial?

Harald

[harascho]

last question:
should I change the rpm boundaries of any of the tun files (like in the prep for the logging tune file..) before I work on a specific VVE table?

Harald

[swingtan]

Apply the BEN data to the tune that was in the car when the data was logged. It's referenced to the tune in the car at the time of logging so there is no use applying it to any other tune.

Don't go changing the RPM boundaries unless you see a need to. As you have seen, altering the boundaries will mess up the shape of the VVE so if you go changing them, you will need to fix the VVE shape again.

[harascho]

Good morning all together

here are some results....
Simon, as the VVE map was a bit rich you can see very clear where I collected data.
Here is a pic of the VVE after I applied the BEN map (the BEN with the average values..) and before I generated the coeffs.

It's quite visible where my data is applied and not. I will work a bit towards the edges and see what I get after generating the coeffs and the VVE.

Harald

[swingtan]

That's looking good. Reduce the cells where you have not applied data to blend them into the area where you have applied the BEN data. The generate the coeffs and the VVE.

Simon

[harascho]

That's what I've got so far.

I will try this modified VVE table in my "normal" tun with MAF

Harald

[GAMEOVER]

Thanks Swingtan & Harascho, :thumb_yello: This thread really helped me alot with my Tune...:D

My VVE is a work in progress! It a little too rich at WOT.

[ringram]

Yeah, the VVE is a surly sod and you need to work him over by hand as the generate function does some serious messing as Swingtan says.
Needs much more manual intervention than the old LS1.
Ive just started mine as well. I used NB02 and got it in the ballpark so I could drive it, managed 454rwhp on dyno dynamics. But now aiming for more accuracy with a Serial WB to hit 500 if possible :)

Swingtan, how many repeat cycles did you do to yours? Or are you still going :) ?

[swingtan]

I did a few, but not too many. I found in the end that unless you had good access to a dyno, you needed long drives on the street to get sufficient data. I've done a few 2hr drives and the data from those worked very well. As it sits now, I can probably do a VVE in half a dozen goes and get it pretty close. The thing is though, that you need to do a lot of work on them before you start "seeing" where the changes need to be made. Learning little tricks like preempting the smoothing function and over correcting some cells and changing other so the desired result comes out.

Currently though, I've gone back to running the MAF and CL. The results now are quite stunning....

http://farm4.static.flickr.com/3462/3398392156_29def37d1b_o.jpg

Thats averaged over 30km of driving and with a 220/224 @ 114 cam in it. I reckon I could squeeze a little more out of it if the O2 switch point tables worked.....

[harascho]

Simon, great numbers....

But I don't give up...after working or better to say start working on the VVE table I will do more runs. First I have to re-check the IAT compensation as we have higher temps now and I will be able to get the IAT to higher values.

Then I have to choose a better route for my log runs. It's hard to find streets and long stretches where runs at steady speeds are possible.

Next I have some trouble understanding your MAF adjustment advices in the VVE tutorial. When I log and try to hit as many cells as possible in the BEN map how do I get input on the necessary changes of B10976 and B10987?

Harald

[swingtan]

Just put aside a day when you can and go for a good long drive. Trying to tune the VVE by doing short repeated runs will make things very hard for you. I found that the absolute minimum time would be about 30min of driving but really you want a good 2hr's of constant driving, getting the engine into different RPM and MAP zones. When I do mine, I go for a drive and change gears to get the revs I need and then do a few 0-100KMH runs ( or 0-160+ in your case, I'd love to do some testing on an Autobahn ) to do the WOT stuff.

Last weekend I went to the 2009 Australian LS1 Drag Nationals and fine tuned a friends car. It's about a 2 hour drive and we able to quickly dial in the tune given the amount of data we logged. Admittedly though, it was an LS1 so it was always going to be easier. If any one is interested, some pics can be seen HERE (http://www.flickr.com/photos/swingtan/sets/72157616945850157/) and HERE (http://s734.photobucket.com/albums/ww346/Nidzdotnet1/LS1%20Nationals%202009/).

As for B10976 and B10987, I have no idea what they are, I didn't even know we numbered the parameters over 10,000. Any chance you added a digit in there somewhere?

Simon

[harascho]

Yes Simon, you are correct, the parameters are B1097 and B1098. I messd up some of the digits....sorry. How does the MAF adjustment work?

My MAF didn't come from a 5.3 engine it's one out if a 4.8 but I was said that they are the same type..

To the testdrive. Yes I will do longer runs including autobahn and WOT. I do not know if it is better to start all over again or to take the first mods to the VVE as a baseline. As I startet with your VVE from a 6.0 which is on the rich side maybe I should continue with the VVE I posted in this thread.

Harald

[swingtan]

OK, For MAF adjustment, just log MAFFREQ2 as well as what you have now. Then look at where the WBO2 differs from the commanded AFR and read the MAF frequency from those areas. Then adjust the settings in B1097 and B1098 up or down to suit. The settings are for airflow vs MAF Frequency, so increasing the value will richen the mixture while decreasing will lean it out. If it seems a little confusing, open the "MAF Calibration - LC-1 WO2 BEN" MAP and use it like the Auto VE BEN MAP.

As for starting from scratch, it all depends. If you are not too far off now, it's probably worth just moving forward. Do a good drive and apply the corrections, then drive home and see the difference. Once home, again apply the corrections and it should be pretty close.

Simon

[GAMEOVER]

Wow!!! Getting the VVE table in the neighborhood is kinda easy... But having WBO2 match Command AFR is NOT!!!! I'm having trouble getting my WBO2 to match Command AFR....I'm always either .5 too rich or .5 lean..... And this is from cruise up to about 3500 RPM's! Some spots are 1 point off!!! But the more data I gather(i.e. longer trips) the better my VVE table improves...:)

[GAMEOVER]

Still working on it...

[ringram]

You might find the afr variation is down to changes in IAT. Hence why using that mod table against injector flow might help you reduce that error. It used to on the old LS1 setups with the COS. Log BEN's vs IAT and see if you see pattern, if so try that.

Personally once you are close Id just to a long drive like Swing says and sort out the VVE, then consider it done and look elsewhere to clean up the variation. Otherwise you can end up chasing your tail all the time. I doubt you will ever get it to track 100% all the time.

[GAMEOVER]

I did the IAT compensation about 2 weeks ago and it's working out perfectly. I agree, I need to do a 2 hour long drive so I can have sufficient data and consider it done...:)