Crazy Proportional Idle table WTF?

[Hib Halverson]

I've been working on the tune of my LS7 (MSD intake, ported heads, cam w 116 lobe sep). The WOT fuel and spark was already done by the in-house calibrator at the facility which built the motor. I've been working on idle quality. In the process of that, I've developed this freakin' burning question...

Why does the rpm low side of the proportional idle table seem back-asswards?

I've read the long (and really good) thread (http://https://forum.efilive.com/showthread.php?9740-E38-idle-tuning.) here on idle tuning on this forum I've even been on the HPTuners forum and read another great thread (https://forum.hptuners.com/showthread.php?58157-Gen4-idle-tuning-guide.) on the subject. In both these threads reference is made by various members of those forums to the rpm low side of the proportional idle table having negative values. When rpm is below desired, these tables command the OS to add a negative airflow amount, ie: when rpm goes low, it's subtracting airflow, rather than adding which seems more intuitive.

This is not just some GM error with LS7 cals for E38s. I have found this characteristic in all sorts of GM calibrations for gasoline engines in vehicle platforms going back as far as 1999. That tells me there is a reason for GM's calibrators doing that.

Anyone know WTF is up with GM calibrating proportional idle like that?

[joecar]

PID (proprtional, integral, derivative) control loops use feedback calculations using some weighted function of those 3...

the weighting of each component usually does not make sense by itself...

The derivative tries to predict future values, the integral accumulates previous values (possibly averaging them in some non uniform scale)... those two are added to the proportional component which then has to make the sum total "fit" (so you may see negative values).

[Hib Halverson]

Hey Joe-
First, thanks for helping me out with this. I have raised this question in a number of places in the last few months and you are the first answer I've gotten.

Is there a "derivative idle" table somewhere in the calibration to go along with the proportional idle and integral idle tables?

Also, while I got you, another question...

When I look at the cal with EFI Live, I cannot find the table for the time delay before adaptive idle control is enabled.

Thx again.

[joecar]

Hi Hib,

It comes under the topic of closed loop feedback control theory which is a very complex/exotic/esoteric subject (lol, which I suffered thru in college many decades ago).

For E38/E67 I don't see a derivative idle table... so I don't know if it does not exist, or if it was simply not discovered/published during reverse engineering.

[Hib Halverson]

Ok.
Thanks, Joe.