Ok, after perusing all the "MAF" threads I've managed to find in this forum, I'll toss in a new topic for the TGO community. I've a '6870 ECM with a Hypertech 155332 bin (from an HLK bin, I figure) and, before I make the jump to a '7165 ECM + ARAP bin (to start), I'd like to optimize the tune I have.

My question is how to find the conversion between "counts" and "volts" as they apply to the two MAF tables. For example, please see for MAF table #1, HLK vs ARAP:



Since it would appear that the overall MAF transfer function is modeled in the bins as a 6-segment PWL approximation to the real MAFS, is determining the HLK MAF Table #1 scalar as simple as performing a linear regression on the numbers?

Question #2: For anyone who might have experience with the $1F ECM code, how do I handle "counts" in the computation for BPW and LV8?

Thanks in advance for pointing me in the right direction...

 

To the readers of this thread and, after a few more targeted topic searches, I managed to answer my own question regarding how MAF tables in the $1F and $6E masks are handled by their corresponding ECMs - I learned that the MAF tables are identical just displayed differently in the bins!
These were the (3) additional threads I found that explained MAF math:

https://www.thirdgen.org/forums/diy-...uestion-s.html

https://www.thirdgen.org/forums/diy-...nges-rbob.html

https://www.thirdgen.org/forums/diy-...af-tables.html

My thanks to the thread starters and participants in these threads, especially RBob, funstick, gta_88_kicks, and Grumpy (RIP), for helping me understand better.

 

I'm not surprised that the MAF table usage is nearly (or is) the same as the later code ($32 & $6E). I've been trying to dump the ROM from a '6870 ECM with no luck. Ended up the ECM was bad, fixed that so I'll get it.

With that we'll have the full $1F source code. Which is really what is required to know what the ECM is doing.

RBob.

 

Thanks, RBob, and do keep us posted. Since I've a 6870 ECM + Hypertech 155332 PROM, I very much want to understand the ECM's arithmetic operations to optimize my power plant's tune. Now that I understand better how the ECM handles the 6 MAF tables, I can turn my attention to the variables that dictate BPW and LV8.

 

Ok, interested TGO readers: I've reviewed the (6) MAF tables from representative bins [HLK ('85), ABTT ('88), and ARAP ('89)] corresponding to $1F (6870), $32B (7165), and $6E (7165) masks. Interested in the shape of the curve, I plotted all (6) tables in Excel - please see attachment.

While the smoothest curve was found in the ABTT bin, it was only in the ARAP bin where the MAF table scalars appeared (the $6E mask). Btw, I've not exhaustively reviewed all $6E bins. Not familiar with the hard coding in the respective ECMs' firmware, this suggests to me that the MAFS's transfer function was modeled by GM engineers using piece-wise linear approximations to the MAFS's non-linear behavior. This also suggests to me that MAF scalars were hard-coded in the 6870 and early 7165 ECMs. If true, then I agree with some tuning experts in this forum that it's unwise to manipulate a bin's MAF table unless you can also manipulate MAF table scalars; otherwise, the MAFS's transfer function breaks apart.

It’s interesting to me that the HLK MAF curve exhibits deviations from ABTT’s smooth characteristic at MAF table transitions.