My AutoVE doesn't look right??

[Mr. P.]

OK this is my first time going through the AutoVE process, I need extra sets of eyes to tell my if I am doing the right/wrong thing here.

This is what we are tuning:
http://www.hp-motoring.com/Paxton/shiner.jpg

2003 Silverado SS AWD, with Paxton Novi-1500 (think Vortech T-Trim, they share the same compressor), factory bone-stock LQ9 6.0L. Also added is a decent-sized A/A intercooler, custom E-Fans, 1-7/8" ARH headers, Corsa Touring exhaust, PLX WB02, & totally built 4L60. OS is COS3.

So last night was the maiden drive with the Paxton, and I specifically chose to drive on the freeway to AutoVE and use the cruise-control on level road to get as clean a dataset as possible. I never drove over 2500-RPM, and tried not to get heavy on the throttle this first time out. I recorded 2 log files, almost an hour of driving. Today I spent the afternoon examining the logs, came up with BENS myself using Excel, then also used the map feature in EFILive Scan Tool to come up with BENS and for the most part they lined-up ok except where the truck went into overrun conditions (MAP < 40-kpa).

So I applied my BENS and the resulting VE table looks... weird. Is this resulting VE Table right?

http://www.hp-motoring.com/Paxton/1stAutoVEBENS.gif


Also, can someone verify that I don't still have DFCO, PE, etc still enabled that will screw-up datalogging?

* Here's the tune I flashed to the truck before logging (http://www.hp-motoring.com/Paxton/Paxton_COS3_0006_AutoVE.tun)
* First 1/2 of driving data (http://www.hp-motoring.com/Paxton/Paxton_Log_AutoVE_1.efi)
* Second 1/2 of driving data (http://www.hp-motoring.com/Paxton/Paxton_Log_AutoVE_2.efi)
* Here's the tune with the BENS applied (http://www.hp-motoring.com/Paxton/Paxton_COS3_0007_BEN_Apply.tun)

Any other suggestions? The WB02 data looks really clean especially in the second log file, am I right in this judgement?

Mr. P. :)

[5.7ute]

What filters did you use on the data?
Links dont work. Attach the tune & log files directly to your post.

[Mr. P.]

OK attached the files here. "Paxton COS3_0006 AutoVE.tun" is the file before logging, "Paxton COS3_0007 Manual BEN Apply.tun" is what I arrived at after applying BEN adjustments.

I used the pre-defined BEN filter -
* TPS changing < 5% per 100ms
* ECT > 172 (*F)
* RPM between 500 and 3000
* TPS > 1%
* VSS > 5-mph

But there were a few spots where I looked at the data 'manually' because I was able to cruise at 1600/2000-rpm with the map at 45/50/55/60 for extended periods and got really good data.

I guess my question is this - there is no head or cam change in the engine at all, so why is the low-end of the VE table require 25% more VE and the mid-range MAP areas not need any change at all? Doesn't make sense but the logging data clearly shows this is what the motor wants...

If this all looks good, do I go ahead and start logging 2400+ RPM now???

Mr. P.

[Chevy366]

Are you using serial or analog with the PLX ? You should see the BEN factors go to 1.0 or close to 1.0 , anything within .2 is good , (.98, 1.2) , fill cells with at least 30 count , 50 is better but not always doable .
I do 2nd gear pulls for most of the AutoVE .
A stock 6.0L with just CIA , COS3 tune will track real close to the stock VE numbers .
Weather sucks right now for AutoVEing , seeing it Snowed .
Just a observation , AutoVEing at lower temps the tune does not do well in warmer weather (COS or MAF) but at low to upper 70s produces the best results .

[Mr. P.]

Thanks Gilbert, just the kind of advice I was looking for. Yeah upper-70's, come to think of it *my* VE is best at those temps too LOL.

I am using the analog output of the PLX; I took the PLX apart to do the serial mod but never finished, I went to Frys looking for the RS232 adapter but no dice. :( You can see in the original post that VE table tracks nowhere close to stock, the shape of the graph is radically different. The engine does have different injectors in it, 640cc @ 3-bar, I increased the IFR {B4001} to 9.125 before driving/logging.

I didn't do any 'pulls' at all, I drove steady-state at a cruise-control set RPM on the freeway and it captured some really clean data - the RPM, DYNAIR, and MAP stayed constant and the WB02 also tracked very cleanly and when I highlighted/mapped those spots in the log I interpreted those cases as 'golden' BENS. Here's an example -

http://www.hp-motoring.com/Paxton/ManualBENS.jpg

Here's a spot of 100-frames where not only the RPM & TPS are steady, but the commanded AFR, DYNAIR, and IPW are as well - the motor is as close to what I would consider 'dyno loaded' as you can get and I take the 0.98 BENS shown here as golden at 1910-RPM. I read through both logs looking for these 'windows' of data and made my own Excel spreadsheet with these 'perfect' cases.

Having read the log by hand, I am now wondering if the default BENS filter isn't specific enough - I don't think it is good enough just to filter where TPS is 'stationary', to me it makes more sense to focus where MAP is stationary as well as RPM; afterall the VE table is MAP vs. RPM, wouldn't you want to find cases in your logged data where these situations were constant and get a solid baseline?

Am I approaching/analyzing this right?

Anyways, I did do as the tutorial suggested and applied the BENS filter, and the result the AutoVE process came up with in the 50-60kpa cells was very close to the BENS I mapped by hand, but it wanted to increase the VE table 30-35% in the low-MAP areas of the table, BENS were like 1.30-1.35! I don't see how this could be anywhere close to correct. I dunno.

Mr. P. :)

[Mr. P.]

Actually, now that I think about it, I could easily write a log analysis utility that finds these 'steady state' conditions and reports them out using a sliding-window kind of approach... I might just do that.