VE Tuning - What am I doing wrong?

[Shdwdrgn]

I have a 2004 Trailblazer with a P10 PCM, and am 'blessed' with having 5 different VE tables to tune. At the moment I am focusing on the low RPM Dbl/Sgl tables. These tables span 600-3800 RPM across the top, then call for a Load% down the side. I set up my histogram for LTFT and tried using 'Calculated Load', but it barely registered any data below the 50% mark (which means only the last three rows of the table have any data in them). I switched to using 'Absolute Load', and now I get a nice spread of data across the center of my table which seems to correspond to the actual driving conditions.

For the past week I have logged some data, then applied the numbers from the histogram to the VE tables (Paste special, Add), driven around 2-3 days and scanned again. Unfortunately something is not right. Every time I make changes, the histogram keeps giving me back a set of data with even higher numbers, and the average MPG reported on my dash has been steadily going down.

OK, I'm new to this, but I thought I understood the basics... Build a histogram that matches the sensors and values shown in the editor, log the appropriate data, then apply the LTFT adjustments to the VE table. Obviously I'm doing something wrong here. My best guess is that 'Load%' in the editor is not the same as 'Absolute Load %' in the scanner -- so what would be the correct item to use here? Is there a trick I can use to find the specific sensor used in the editor that should be applied to the histogram so I'm not trying to just 'guess' what I should use?

I've attached three files below. The text file shows the current VE table I am running, and you can easily see the band just under the center-line where the values have increased dramatically. Above the center-line, the values have been decreasing from stock, so overall I have adjusted by as much as -9 above the center, and +16 below the center. I have also included a log file from this morning (after driving for the last 3 days with the previous tables), and the config file for the scanner with all of the histograms I have set up. Hopefully someone can tell me where I failed.

[Shdwdrgn]

Can anyone help with this? The main problem I'm facing is that in the editor, the Dbl/Sgl-fire VE tables are RPM vs Load%, but there's no clear indication what Load% is supposed to map to. Is this some kind of calculation that we need to mimic? For every histogram I have tried so far, when I add my scanned LTFT values to the DBl/Sgl-fire VE tables, my new scans show even higher values for the trims. I'm at a loss what needs to be done...

[ScarabEpic22]

Use the Low and High load vs TPS tables to determine what TPS = Load, its frustrating but it's because these are a hybrid alpha-N/SD from GM. Be glad you dont have a P12, they have 5 VE tables to tune based on camshaft position.

[Shdwdrgn]

OK I found these tables, but what am I plotting against? Sensor=Load? PID=Absolute Load? I guess I'm completely lost as to what I'm supposed to do with this data.

In the meantime, while looking for the tables you mentioned, I also discovered tables for Dbl/Sgl Fire... I guess you have to be hooked up to the vehicle in order to get those tables to show up in the list. Unfortunately, when I try to use these built-in tables, there is no unit value available for 'Load', and when I start scanning, all of the data shows up on the 0% row. When I try to manually select the Load sensor, its name is in black in the list... I believe this means that I don't have all the required values listed in the table? But how do I figure out what values *should* be in the table to fully read the Load sensor?

[planethax]

Pretty sure it is TPS you want in your histogram

[ScarabEpic22]

Honestly Ive been doing it long enough I just interpret the log and tweak it in the appropriate areas. The P10s dont vary that much unless you want to run a different octane, CAIs and exhausts really dont change the VE tables that much on these once you get it dialed in.

I havent tuned one in HPTuners yet, done all of them so far in EFILive. The next one I get in Ill use HPTuners to continue to get more familiar with it. Im really tempted to make the TPS Load tables match the VE load axis so it's closer to a true SD setup, might be worth a shot if you want to play with it.

[Shdwdrgn]

Maybe I'm starting this wrong... From my reading, my understanding is that you want to do VE tuning from SD mode. However I thought I read somewhere that since the 2002-2005 Trailblazers don't have a MAF, they run in SD mode all the time. Is that correct, or is there something I should be doing to force my system into SD mode before I start logging?

I'm still not having any success here. Even with trying to tune the VE idle table, which is just a simple RPM vs MAP, every time I scan I get a band of LTFT numbers that are calling for high positive values (starting out around 5-6), and when I add the chart from the histogram to my VE table, the next scan shows even higher numbers through that same area. I guess I could try subtracting the values, but that doesn't match anything I've read, and sounds counter-intuitive.

Currently I just hook up my computer and start logging once the engine is warm, but since the results (especially my recorded MPG) indicate that my changes are making things worse, it makes me think either that I need to be changing something before I start recording, or something is just completely screwed up in my methodology.

[ScarabEpic22]

Post your tune and a log.

[Shdwdrgn]

From which point do you want the info? My starting point, and a log/config showing what I'm seeing? Or do you want an adjusted tune file based on the scan I got the first round, and the changes I made based on that log?

[Shdwdrgn]

I finally got some time to start working on this again, so I wanted to start by posting my starting point. Below I have attached my current tune (which I have been running for the past 4 weeks), and a log I took this morning. I am re-posting my .cfg file, although I believe this is the same config I was using in my original post.

When looking at the histograms on tabs 1 and 3 (Idle LTFT and Sgl/Dbl LTFT), it appears there are clear regions where the default values need to be raised. In the past, I have copied the log tables, then in the editor I would Paste-Special -> Add. After reprogramming, I would drive around for a week, then scan again. I would expect that after these steps, my next scan would show LTFT values much closer to zero, but instead, the results I have been seeing are typically even larger adjustments... Any suggestions?

2004TB-20130508.hpt
2004 Trailblazer VE-Spark.cfg
20130604-01.hpl

[jbac]

Hey,

So you're reading slightly positive LTFT? So that means the engine is running lean. As I've understood, to get the correct ratio you should reduce the VE for that fuel trim cell.

[Shdwdrgn]

It's worth a shot to try it that way... except everything I've read says the opposite -- that if the numbers are positive, it means the computer had to compensate by adding more fuel, so I would want the default values to be higher than they are. Of course it doesn't seem to be working that way, so I'm willing to try going the opposite direction next time I update my PCM.

[cobaltssoverbooster]

It's worth a shot to try it that way... except everything I've read says the opposite -- that if the numbers are positive, it means the computer had to compensate by adding more fuel, so I would want the default values to be higher than they are. Of course it doesn't seem to be working that way, so I'm willing to try going the opposite direction next time I update my PCM.

yes, and you would increase base values by the trim amount to achieve your goal. if your values are small enter them by hand using a half value or enter it into excel and make it halve everything for you.