Calibrating VE with fuel trims.

As Andre mentioned in this particular course that its possible to calibrate the VE table using fuel trims, I'm particularly interested in this as i only have access to a wideband meter through my dyno and at this stage don't have the capability of inputting a wideband meter into HP. At a later date i will be getting the gear to do this but for the mean time i have to work with what I've got.

I have a few questions relating to using the fuel trims rather than a wideband to show the errors VS VE as it wasn't covered deeply in this course. If there is a separate course covering this and if someone has a link to that would be awesome. Would i be correct in saying that to create a histogram for fuel error vs VE i would be setting the main Parameter as short term fuel trims, then having my column axis as manifold pressure, and row axis as engine speed/RPM? Would this work to be able to calibrate the VE table similar to the wideband setup?

I'm aware that this will cease to work once open loop is enabled. I just want to make sure my thought process on using a histogram for fuel trims to find the errors in the VE table. Would i need to be enabling LTFT Aswell for this? i am under the assumption i would not as these are not instantaneous feedback like STFT.

Thanks again for all the help.

Just to clarify, that's possible in closed loop only, while fuel trimming is active.

This is done just like the closed loop portion of MAF scaling, except you alter VE instead of MAF calibration, so the process of making adjustments to reduce total fuel trim (short + long term trim), is the same. You do need to log both short and long term trim.

To tune higher load areas in open loop, where no fuel trimming is taking place, you'll need the wideband and to switch over to VE adjustment based error between lambda target and lamber actual (as monitored by the wideband).

Thanks for your reply Mike. So just to clarify you need to log both short and long term trims for this to work?, I see the math Channel for STFT + LTFT. I currently have ltft disabled but for the purpose of calibrating the VE I’ll re enable them.

i had a go at it abit today but was only logging STFT so ill try again with both active. seems ultra time consuming to try populate each cell with data even with looking for trends to hand blend them.

One other thing I might add, upon my attempt at calibrating the VE today. I was gathering data in the histogram STFT + LTFT against VE. Now I did have LTFT disabled so that might be the issue but I had good cell counts (over 300) per cell and was using the paste special function and it appeared to be working, after pasting that data in the error was less and I ended up narrowing it down to +/- 3% but after looking at the 3d option I started noticing the slope for the VE table was getting erratic, to me that indicated I’m doing something wrong? Any in-site into that?

Thanks again

Yes, log both STFT and LTFT which it sounds like you've done and made some great progress.

At a certain point yes using this method may cause a VE table to be less than smooth, and some smoothing is generally suggested.

Feel free to post the table data and 3d form so we get a feel for just how erratic it is. Some variance may be normal, but beyond a certain level it often indicates there's either room for improvement via some smoothing, or perhaps an issue such as erratic fuel pressure.

I really appreciate your reply’s Mike, you’re a wealth of knowledge.

at the moment I’m running a speed density patch on the car I’m tuning. Now I did run into a slight issue, upon loading up the original base stock file to copy the VVE table I loaded it into excel and transposed it to swap the axis. But I noticed on the stock VVE table the manifold absolute pressure sites differ from the table that’s created when the speed density patch is applied. Also the RPM axis stops at 6200rpm where the tables (VE IMRC CLOSED & OPEN ect) go to 6800. So once I’ve copied the table over to get the car started theirs two rows of cells that are still 0. Along with the manifold absolute pressure sites being different. Is this normal? Is there a way to edit the sites so it will match the VVE table I’ve transposed? I thought it was due to the 2 bar speed density patch as that’s the only option I have, I’ve made sure to change the map sensor tables back to 1 bar but still no luck.

HPA wants everyone here to succeed and I'm happy to help!

That's a great question. As a starting point you can keep data where the axis values are the same, and if lets say the new axis break point is between two old ones, you can use an average of the two.

For example perhaps you had break points at 1800, 2200 RPM, and now have one at 2000 RPM.

The average of the values at 1800 and 2200 may not give you the perfect value for 2000 RPM, but it should be good enough to start the process.

When it comes to areas that didn't exist prior, like your 6800 RPM example when the table used to end at 6200, that's uncharted territory. At high RPM VE will often taper down a little, so if you start out by keeping the 6200 RPM value in the new 6800 RPM section, it will likely be in the ballpark or perhaps rich.

Then as you're going through the tuning process, gradually working your way up in RPM and load, you can "tune ahead" some, meaning if you find you needed to add 5% at 5500 and 6000 RPM, perhaps add 5% to your 6800 RPM values before you run the engine up that high.

When using a 2 bar speed density patch, with a 1 bar MAP sensor, yes you'll want to make sure the MAP sensor scaling remains appropriate for the sensor you have.

The VE table load values will extend beyond what you need for an NA engine, but as a precaution I copy the values from the highest load area your engine does hit during tuning, into all higher load areas you don't expect to hit.