12212156 Fuel dynamics questions

[pmcquay]

Here is a log with those changes. I dont have fuel trims on. The more I look at it, the more I feel like the VE/MAF tables are just not tuned in those dips. Maybe its just that simple?

This log is with the previous injector changes as well, and it doesnt seem to have made any difference. I still do see a bit of the hump going on and off of throttle, but the more I look at it, especially off throttle, it looks like theres a hump and then it settles into a new AFR, as if my VE or MAF tune is just wrong in those cells.

For Example:

Screenshot 2024-05-16 131412.png

And slightly less violent:

Screenshot 2024-05-16 131503.png

[kur4o]

K_Dyna_Air_Coefficient
make sure you swap x/y axis for better reading.

These are the airflow filtering. What is does is to predict 2 TDC before air reaches cylinder.
There is 18 zones,

managed to figure
1st zone is WOT
2nd zone is idle
3-18 zone is matrix based on map and rpm threshold defined in some tables, but can`t figure yet exact usage.
For example zone 3 will be from 0-1000 rpm and from 0-25 map, and so on. We need to figure which zone your in when the issue is exhibited. Some ram logging will be needed. And the scalars and usage is still unclear but with some testing it can be dialed.

Trims usually mask very good the conditions you have, In open loop these are very hard to tune out, if not impossible since on the same spot in tables you have very different fuel needs.

[pmcquay]

I would certainly be interested in learning more about RAM logging, do you have any information about what RAM addresses I should be logging? Is there a document around that I should be looking at, or some other way to find parameters to log?

[kur4o]

Only way to figure ram addresses is by making a disassembly or have some gm docs that specify it per OS you use. Too bad a ram address don`t translate the same from OS to OS. It is OS specific.

[kur4o]

I started porting some RAM addresses, but not sure what you need for log.
It is also very likely, the injectors and cylinder order are mapped incorrectly for DynaFuel and DynaAIR logic.

[pmcquay]

I cant find any table like that in airflow dynamics, but here are the tables for fuel dynamics:

Screenshot 2024-05-21 070839.png

I would sort of expect them to be correct, as I havent changed them, and they came from a reliable source. Worth checking though, I suppose.

[pmcquay]

ohh I get what you're saying now. The manifold runner configuration is not the same in that it doesn't really "pair" the cylinders like the l31 lower manifold does, so those tables could be wrong. Ohh that's interesting.

I wonder if thats the source of the choppy AFR I see when under more load:

Screenshot 2024-05-21 073454.png

[pmcquay]

I also think it might be possible that those tables are configured wrongly in the XDF, as I don't have 9 cylinders...

Anyway, if I disregard the channel #8 there, it looks like this is not a firing order, it appears to be something like a delay, where the "constructing" table is used for determining the amount of dynamic fuel mass, and the "receive" table is the cylinder channel that gets the fuel mass. I don't see why tables are even needed here, much less two tables. It appears that the two tables are one cylinder channel apart for my tune, and for an LS1 tune they are 3 cylinder channels apart. I can't figure out why the offset from the start of the table though.

It seems to me like this could be a scalar. Something like "number of cylinder channels to delay dynamic fuel mass compensation". I am probably missing something though.

[pmcquay]

Yeeeahhh.. so. I've done some more testing and playing with things. Turns out my assumptions about MAF curve were quite wrong. I've done some very minimal tweaking on it and gotten quite good results in taming the unruly rich dips and lean spikes.

Lesson to me I guess.

[antus]

Bit late to the thread, and sorry if I am missing the point here, but it looks like rich on accel, and lean on decel. Its not just accel enrichment or decel enleanment or calibration to those? For the choppy, its an aftermarket intake? It looks like it could be air reverberation through the MAF so its getting double measured as it reverberates artificially inflating the high peaks beyond the actual airflow. Stock intakes sometimes have a plastic chamber of a measured size hanging off the side in a measured location to absorb pulses. It could be real reverberations and nothing you can do about it in tune as its a physical problem with the intake airflow.