VVE ZONE tune question.

[Williamblake]

Im digging into VVE to try and get a grasp on it and i had an idea for adjusting and keeping the the 3D VVE smooth. Let me start by saying im not a math guy and ive read some on coefficients but i don't fully grasp it and doubt i ever will.

With that said obviously the VVE has zones and there is a coefficient for each zone. To my simple understanding of it that means that number represents the VE for each cell in that zone. What i have noticed when using the Hptuners built in VVE editor or Cringers editor is that after adjusting (Copy> Paste Special % half > Calculate coefficients or in cringers tool upload the histo and hit calculate coefficients in the tool) it changes shape of the zones and has sharp rises or dips between zones like in the pic below.

VVE after edit.png

Now here is my idea and what i wanna know if it may work or is way off base. In cringers tool i can copy over my histogram from the log. Then highlight each zone and get stats for the zone. And if i take the mean error for the zone and then apply that to the entire zone it moves the zone up or down (in the 3D view) accordingly but keeps the general shape of the zone the and the entire VVE map ending in a much smoother looking map.

This is a pic of the stock VVE before:

vve stock .png

And this is after adjusting each entire zone by the median error. You scan see where several zones have dropped down but mostly kept their shape. (This one could use a little smoothing on the edges i didn't do.)

after adjusting .png

And here is a screen shot of cringers tool where im getting the median error for each zone.

median error .png


Hopefully the way im explaining what im thinking makes sense If anyone has any input on this idea weather it seems decent or im way off track please let me know. It popped in my head and i just want some other opinions on it.

And im just using my stock 2014 Silverado to data log i know the fueling errors arnt much at all. IDK if its even enough of an error to be able to truly test this idea and see what happens.

[Alvin]

That is reasonably smooth. I wouldn't sweat it too much. I could make it smoother if you post your tune file if you'd like but I doubt you'd notice the difference.

Some factory tunes have TERRIBLE looking VVE tables.

[TriPinTaZ]

I haven't found a need to have the VVE table smooth on the Gen V stuff. As long as the zone boundaries are accurate and the zone area is close, the car should drive great and trims(if you use them) should be in line.

[Williamblake]

Here is my tune after i edited the VVE. I appreciate you taking a look!

[Alvin]

This looks sufficiently smooth.

[CYN_CRVR]

Regarding the original post, I have also used this same method for VVE tuning. I think it is the best way to make the first 1 or 2 passes (which should get you most of the way there in a stock-ish car if you did a good job logging and filtering data).

[Williamblake]

Thanks for yalls input i feel a lot more comfortable with how VVE should look now and what is smooth enough.

[charcoal03silvy]

I have found adjusting the zone breakpoints from factory allows it to calculate each zone significantly better going back to coefficients and remove the need for any zero by zone tuning or special tools.

[danmw2003]

I think becasue a lot of newer vehicle's have prediction coefficients, its difficult to near impossible to get a smooth looking VVE. At least for me anyway... Some of the more seasoned veterans around here could likely work it like magic, but its not uncommon to see a "not smooth" VVE table.

[Williamblake]

i had some good result with the "not smooth" VE table. i found another post that explain the zone shapes and that helped me understand it better. I did try adjusting each zone buy the mean percentage to see how it did and it took me from about -2 percent on average to 5-6 on average. Im doin this on a stock truck with very small changes but ill say that my idea was not a very good one lol.

[danmw2003]

I personally, wouldn't worry about it, if your vehicle has VVE and prediction coefficients. Just make sure you dont have huge spikes.

[Cringer]

Im digging into VVE to try and get a grasp on it and i had an idea for adjusting and keeping the the 3D VVE smooth. Let me start by saying im not a math guy and ive read some on coefficients but i don't fully grasp it and doubt i ever will.

With that said obviously the VVE has zones and there is a coefficient for each zone. To my simple understanding of it that means that number represents the VE for each cell in that zone. What i have noticed when using the Hptuners built in VVE editor or Cringers editor is that after adjusting (Copy> Paste Special % half > Calculate coefficients or in cringers tool upload the histo and hit calculate coefficients in the tool) it changes shape of the zones and has sharp rises or dips between zones like in the pic below.
...

For the base VVE (no VVT), there are 6 coefficients (not one). When you add a V8 VVT cam you add 4 additional coefficients (10 total). When you go DOHC you add 5 more (15 total). Plus the equation also takes into account RPM and MAP (or Pressure Ratio) for two additional variables.

The complete math behind this is covered in detail here:

https://youtu.be/pZ7i9c0sCN8

The method you describe (using the per zone average error) and modifying a single zone by that value absolutely works the way you describe. It will raise or lower the entire zone and NOT change the overall shape or curves. This is the same thing as going to the [ECM] 13400 - VE Normal - Constant Coeff table and modifying the zone coefficient directly (by adding/subtracting to raise or lower the entire zone). Of course I don't recommend direct manual manipulation of the coefficients, the VVE Editor (or my tool) makes this chore much easier and reduces the changes of error.

This per zone averaging method will work reasonably well assuming the following two conditions:

1) The initial shape is correct (or close enough to correct).
2) The range of fuel trim error values are also reasonably close.

2a) By this I mean if a single zone has a tight range of fuel trim errors say from -5% to -2%, then the per zone averaging method is fine.

2b) However, if a single zone has a wide range of fuel trim error data, say from -20% to +10%, then you really need to fix the overall shape first (and most likely this means to rezone your VVE).

Directly modifying any of the other coefficients is definitely a NO-NO as these control the curves and overall shape. You will definitely get unexpected results by doing this.