Scanner table weighting?

[MikeOD]

I was curious if anyone has looked into adding this to the scan tool.

Every table has resolution, (except virtual VE), so when we log anything, be it fueling error against the MAF sensor, or a VE table (even with VVE, we have to setup a resolution to scan with), or knock retard...anything really, we're interpolating data.

When the PCM looks at a table to determine what to do, it interpolates a value between 2 cells. Say 2 neighboring cells in a MAF table are 8100 Hz and 8250 Hz and they have 273 and 302 g/sec in them respectively. If the MAF sensor is reading 8175 Hz, then the PCM is going to assume the average of the 2 cells for airflow, 287.5 g/sec. Now...if the 8100 Hz cell is spot on, but the 8250 Hz cell is way off...that's going to result in a fueling error....and as the MAF fluctuates above and below 8175 Hz, the error is going to affect a cell that's otherwise good.

Is there a way to weight the data? If the difference in frequency between 2 cells is 150 Hz (8250 - 8100), and the current reading on the MAF while logging is 8190 Hz, would we have the scanner automatically apply 60% of the fueling error to the 8250 cell, and 40% of it to the 8100 cell? I believe this is how the PCM is deciding what to do based on the sensor reading and lookup table...it's 60% of the way from one cell to the next, so it's weighing the data between those 2 cells in a fashion like that...

I currently try to log to get more and more and more data, and then I even try to find ways to filter the data...to ignore data that's halfway between cells and try to only apply data that's closer to the actual value of said cells...for example...I'll setup the software to log 8075 Hz, 8100 Hz, 8125 Hz, then skip to 8225, 8250 and 8275...then I'll export that to excel, and have a second table in excel that filters out only the 8100 and 8250 values...then use that to correct the tune...I'm just trying to come up with ways to do it faster.

[DSteck]

Custom histogram with twice the resolution
Highlight error
Copy with labels
Paste

Then review it and smooth it. The key is to get a large sample which can dampen out the effect you're talking about. Gen III VE tables are especially bad about this because of the 400rpm jump in that axis.

Realistically though, if the scanner says 8125Hz and there's +5% error, that error is actually from a few cycles back, so it's not like your error is truly live. This is why it's important to try and get the bulk of your data in steady state conditions.

[MikeOD]

Yeah, I do my best for steady state...and I know the error lags behind.

I did not realize copy with labels and paste did that...talk about making it easier for me!!!

Thank you Dave.