2015 Silverado VVE Tuning issue when calculate coefficients

Attachment 91935

I am trying to tune the VVE table on my 2015 Silverado and having some massive numbers generated when I click the claculate coefficients button.

I am logging AFR error with my wideband, I copied and paste special by% and everything looks great until I click the calculate coefficients button. The cells around the 23 zones are all in the 2000 range but those two zones are over 55000.

If I change the cell hits in the histogram around and try again, it will be fine so it seems the math is doing something crazy there. I can get it to mess up the zones around it as well.

Any ideas on how to fix this issue, what I am doing wrong or just use the higher cell count from the histogram when pasting?

I started using 50 cell hits but it seems that 300 or higher eliminates this issue. It was a 2 hour log, so I have quite a few cells populated even at 300 so I can make it work. I was just curious as if anyone had this issue and I didn't find much from the search button.

I copied the labels from the tune VVE table and the zones are all stock. I used AFR error with engine speed vs. MAP/Baro ratio for the histogram.

Any help is appreciated. I am just learning so I have a long ways to go. I have the MAF logging less than =/-3% error and now i'm trying to work on the VVE and hit a road block with this issue.

Attachment 91937Attachment 91939

I have come across this a few times as well when working on my vve table for an e78 2014 chevy spark.

I assumed the issue for me was due to large injectors and possibly incorrect injector data. This was causing me to compensate in the vve table and the crazy large number was the result. I'm still not sure of the actual cause.

I think I just started smoothing the areas before calculating and I think that has prevented the issue for me. That's probably a band aid fix approach but my table is a ways off at this point anyways.

I'm interested in learning about this so thanks for asking the question and bringing it back on my radar.

View the 3D model before calculating coefficients and look for anomalies. Also, do a vertical smooth, column by column ONLY on the uncorrected data, between where you have applied afr error corrections on the table; once you get the whole thing close you can skip that.

Thanks for the reply JS264. I will try those steps out later today. After I posted, I figured I would go ahead and just adjust the VVE with the cells over 300 hits like I mentioned and it worked for the first table but when I went up to the next degree intake angle table it really messed up more zones. I tried all the way up to 30,000 cell hits copy and pasted and it still messed up on the next table. This is really crazy but I will try your advice later to see how it goes.

Thanks.

I got pissed at tuning the VVE on mine with VVT so currently the cam advance tables are zeroed out... Maybe after I get my cold piping redone and MAF put back in I'll do it.

I tried smoothing between vertical cells and smoothing any peaks to the point where they smoothed out decent and it still screws up the whole table as soon as you calculate coefficients.

Here is a pic of the table before and after I click calculate coefficients:

Attachment 91943Attachment 91944

Quote:

---

Originally Posted by MattofEarle

I tried smoothing between vertical cells and smoothing any peaks to the point where they smoothed out decent and it still screws up the whole table as soon as you calculate coefficients.

Here is a pic of the table before and after I click calculate coefficients:

Attachment 91943Attachment 91944

---

1. Try to zero out all the prediction coefficients tables. Engine -> Airflow -> Dynamic -> Prediction Coefficients.
2. Also, you may have to shift around the zone boundaries for the vve table.

I made sure all 11 prediction coefficient tables read 0 and it still didn't fix my issue. Is there any method to properly changing the zones to make them work better?

If the changes are huge, sometimes you just have to go zone by zone. You can help yourself out by evening out the zone sizes a little so that you have more an even grid to work with.

Quote:

---

Originally Posted by sevinn

If the changes are huge, sometimes you just have to go zone by zone. You can help yourself out by evening out the zone sizes a little so that you have more an even grid to work with.

---

I have been dealing with this for quite sometime. Was one of the first posters about this. This is obviously an HPT issue but they have been radio silent about this

So how do you change/even out the zones?

I know Higgs said something about decimals but i cant find the post.

Quote:

---

Originally Posted by Ghostnotes

I have been dealing with this for quite sometime. Was one of the first posters about this. This is obviously an HPT issue but they have been radio silent about this

So how do you change/even out the zones?

---

Under airflow>dynamic you can change the Zone RPM Boundaries and Zone PRatio boundaries until you are happy with the way your vve table zones look when looking at them in the editor. The values you use are going to change depending on whether you are NA or Boosted because you can get more resolution per zone if you're an NA car with the PRatio boundaries than you can with something that needs to go over 1.00 PRatio.

I don't think there's really a right or wrong way to do it, but personally I just make all the rpm and pratio zones even when I can.

Quote:

---

Originally Posted by Ghostnotes

I have been dealing with this for quite sometime. Was one of the first posters about this. This is obviously an HPT issue but they have been radio silent about this

So how do you change/even out the zones?

---

I did my first VVE change yesterday and initially I was changing the table zone by zone (hitting calculate after each zone change).. but it messed up when I got to zone 7 and gave negative numbers. So I did a full copy/paste special by % on the whole table and calculated to see which ones messed up. Luckily it was still zone 7, so I just took the original numbers for those cells and manually put them back in. It seems to at least have gotten me moving on being able to get more data.

Today I loaded up maf failed mode etc and went for a drive trying to populate all the cells I could by running different gears manually. Set my cell hits to 20 vs 5. I assume that will work ok.. If not someone let me know. I applied the changes tonight after work to the whole table at one time and calculated and all the numbers seem to be ok this time around. Although the 3d view isn't exactly a smooth picture. I will log again tomorrow to see where the numbers fall to see if what I changed helped any.

That works but one thing i might add is most do the vve and run through all the load cells in one gear. Otherwise the map wont represent the engine if the different loads from the different gears skew things

Quote:

---

Originally Posted by nicul15

That works but one thing i might add is most do the vve and run through all the load cells in one gear. Otherwise the map wont represent the engine if the different loads from the different gears skew things

---

Yeah, I did the normal single gear 2000-4000 500rpm increment/hold 5 secs, the 2000-4000 light throttle, 2000-4000rpm moderate(less than 50%) throttle method the other day. Just noticed that it did not populate as many cells as I would like.

I am going to take the car out shortly with the changes I made from the multi gear drive last night.

Ideally, I need to get this thing on a Mustang Dyno to be able to do some steady state stuff

Did another log today and worked on changes tonight. Copied/pasted special multiply by % 3 separate times with each time I calculated I would get off numbers in Zones 7, 13 and 19. Repeated like clockwork the 3 times. So I tried to Paste Special by 1/2 % and it seemed to work. May be an option to at least attempt if someone if getting bad numbers on a whole table copy/paste.

More logging tomorrow

I have been saying for years that there is a math issue in how HPT back calculates the coefficients from the VVE. But regarding a part of the swing in values is usually due to the zone boundaries. You can only make changes within the zones with any kind of success. The VVE is really a bunch of independent VE surfaces yet they still need to transition between each other. The calculation in the background is not as sophisticated as how its done with software as Metlab.....

Quote:

---

Originally Posted by Redline MS

I have been saying for years that there is a math issue in how HPT back calculates the coefficients from the VVE. But regarding a part of the swing in values is usually due to the zone boundaries. You can only make changes within the zones with any kind of success. The VVE is really a bunch of independent VE surfaces yet they still need to transition between each other. The calculation in the background is not as sophisticated as how its done with software as Metlab.....

---

I assume if you are making % error changes zone by zone and get a zone that calcs to bad numbers you just plug the originals in and try again? Zone 7 seems to be the weird one for me. Setting up a few hrs on a Mustang dyno sure would make all of the data gathering easier.

Something i checked, i always got that huge negative number whenever i was correcting the idle area of the VE, if i copy pasted error % it got screwed, so i tried to multiply the area by a single number instead of each number by a different % like it does when you copy/paste special, that let me actually advance in my tuning instead of just having to clean it all and try again.

See if that works for you

Quote:

---

Originally Posted by mJolnir

Something i checked, i always got that huge negative number whenever i was correcting the idle area of the VE, if i copy pasted error % it got screwed, so i tried to multiply the area by a single number instead of each number by a different % like it does when you copy/paste special, that let me actually advance in my tuning instead of just having to clean it all and try again.

See if that works for you

---

I appreciate the advice.I will give that a try too once I get back to working on it.It's frustrating for sure when it throws out the bad numbers.