2007 6.0 Missed Something Major Here

[Big D]

Holy Smokes this has been a brain drainer!!

I worked on this for about 7hrs and while general practices make sense nothing worked properly.

2007 Silverado 6.0 LY6 K&N type intake system and aftermarket exhaust (no headers).

I used Swingtan's auto VVE guide, http://forum.efilive.com/showthread.php?t=8961 Followed it to the letter.

First idle was horribly like it was cammed. When starting it would fire then shut down with out correcting using the pedal.
When logging ben's and applying using the "copy with labels" then apply using "paste and mulitply with labels" the values would max out to 10000.

Other times using just paste with labels or any other function including what I just stated all values would decrease the table and not be a "vertical" change but a "hole" in the table. I would smooth this out using a 4% change mostly negative numbers and apply coeffs and generate vve. This helped out cruise area AFR's but idle sucked and WOT was way to lean like 18 aft

My spark table wouldn't always use the commanded amount either. what is holding this back?

Eventually I re-enabled the all the fuel trims except B1510 and re-enable the DFCO. This help idle situations out back to stock but the VE table B8101 was not even close to right.

The other issue was B8101. I would adjust, I would have a error saying zone 12 is too small for a 3x3 area ?? When regenerating B8101 I would save then flash the ecm. I would then open a different "mod" tune and the B8101 table would not resemble anything it should be. I dont have any screen shots of this but will try to get some. The table would generally be flat and have huge spike with a long high ridge through it.

With out doing some idle compensation work, according to the pdf tutorial I dont see how its possible to leave the ecm setup in a complete SD situation. The only way I could get the VE table to closely resemble what the tutorial explained was to manipulate it with out the right data but then after you generate the VE it wouldnt be close anyways. I also had to use "paste and subtract with labels" to have the VE table move the way the tutorial showed.

With all of this not going properly either the tutorial is not correct for my situation or I missed something really big, which I think is the problem.

I have 14 different mods to this stock tune and log files to go with each for anyone to help me out with.

I have about 13 million questions about this. Sit back, take a look and go wtf? and then start asking away.

[Big D]

I was able to get a copy of the bad VE table I referenced in 1st post.
Here is a photo of my horrible description

[Big D]

I altered the VE table using the BENS and this is what I come up with. I dont think this is correct using swingtans tutorial as a guide.

[Big D]

Ok after finding this thread http://forum.efilive.com/showthread.php?t=10296&page=5 this has cleared up a ton of my questions cleared up.

Looks like I am ( & was) headed in the correct direction.

[swingtan]

Looking at the VE table in post 2, I'm guessing you missed the section on page 10 of the tutorial regarding B8021, the bit that says "Before moving on, you should do 3 things". If not you may have done it incorrectly. Note that in this step, you are just copying over the original data to the new tune. You want to "copy with lables" and "paste with lables" for this step only.

Failing to do this step, or doing it incorrectly, after changing the zone boundaries, will give a result very similar to the image in post 2. It looks like you are on the way though, so that's a good thing.

Simon

[Big D]

Looking at the VE table in post 2, I'm guessing you missed the section on page 10 of the tutorial regarding B8021, the bit that says "Before moving on, you should do 3 things". If not you may have done it incorrectly. Note that in this step, you are just copying over the original data to the new tune. You want to "copy with lables" and "paste with lables" for this step only.

Failing to do this step, or doing it incorrectly, after changing the zone boundaries, will give a result very similar to the image in post 2. It looks like you are on the way though, so that's a good thing.

Simon

After spending some time away and coming back to this, it now seems to be making a lot more sense.
Thanks for the input!

[Big D]

Did some more testing on this and have some more thoughts.

BEN's, when applying to B8101 they create large downward spikes using "paste and mulitply with labels". The VVE Tutorial show the spikes going up ward. The reason I question this is when smoothing B8101 to match the downward spikes AFR numbers dont seem to get better and it seems to run worse.

How does the "labels" part of paste apply or work vs. paste and add or paste and mulitple?

Going for a long drive you can only log so much info before its "full". How are you compliling mulitple logs into B8101? Do you drive until logging stops, adjust b8101 and repeat?

Have you tried smoothing and blending large areas of B8101? That works well but I have not tried it to see the result. Bad idea?

My V2 box for some reason is dead after tuning today and waiting to hear from support. I wish I could see the B8101 to reference on replys to this but I will have to wait to see the table.

[swingtan]

The BENs are a correction factor "for your tune". They will mean nothing to another tune or vehicle. The tutorial has both up and down spikes and the images are just an example of what it "may" look like.

The "labels" means that the data will be correctly applied to the areas that need correction. You can just select some cells and "copy" and "paste & multiply" but you need to ensure you pate are the correct cell, otherwise you will see a world of pain..... Using "labels" ensures you get it right.

If you are filling the memory space, then you are logging to the internal memory of the V2. Get an SD memory card and go into "setup" in the V2 and change the memory type to the SD card. Full instructions are in the V2 BBL documentation.

When smoothing, you may need to do some localised smoothing around spikes first. Select a small area with the spike in the middle and apply a couple or more smoothing steps. This will average the spike and the surrounding areas, then smooth using larger selections that appear similar in value. Applying smoothing across the entire table is probably not a great idea as the generation of the VVE will do this for you.

Simon.

[Big D]

The BENs are a correction factor "for your tune". They will mean nothing to another tune or vehicle. The tutorial has both up and down spikes and the images are just an example of what it "may" look like. The BENs I have been using on avg is .6 to .8 and this drops the table down like in the second set of photos previously posted.

The "labels" means that the data will be correctly applied to the areas that need correction. You can just select some cells and "copy" and "paste & multiply" but you need to ensure you pate are the correct cell, otherwise you will see a world of pain..... Using "labels" ensures you get it right.

If you are filling the memory space, then you are logging to the internal memory of the V2. Get an SD memory card and go into "setup" in the V2 and change the memory type to the SD card. Full instructions are in the V2 BBL documentation. I have a Sandisk 512 mb install and set to record to it. Sometimes, it would log 8300 frames and other times 10000 frames.

When smoothing, you may need to do some localised smoothing around spikes first. Select a small area with the spike in the middle and apply a couple or more smoothing steps. This will average the spike and the surrounding areas, then smooth using larger selections that appear similar in value. Applying smoothing across the entire table is probably not a great idea as the generation of the VVE will do this for you. I have been using "localized" smoothing starting with -10% to get close then work with finer amounts to flatten the table out at the tip of the downward spike. The generate coeffs, and generate VVE.

Simon.

I take it you must have a 2 gig or something large for a SD card? I can see why this is important now.
My cell counts are 400+ in take off, cruise and deceleration areas but the BENs dont change. They average a number such as .68. So if I am able to hit many cells no matter what the count is (25 min), doesn't this justify enough BENs average to correctly smooth B8101? Is the goal here to get a good averaged BEN number somewhat shortly with less than 2hrs of driving. Is this why using a dyno to load and adjust RPMs using one throttle % is so easy.