My diypnp was running the Mario code and iirc it had a coolant based idle duty table. My MS3 doesn't seem to have this. When the car is cold it wants to idle very low, then as the car heats up the idle increases beyond the target rpm. I adjusted the idle set screw while the car was warm, with closed loop active to the desired rpm. However I think this might be the cause of the poor cold idle. I adjusted the idle set screw last night with pwm warmup and the idle valve overriden to a value of 19.1.

I then switched to closed loop and powercycled the MS. It held the idle that was set with the idle screw, but it doesn't seem to try to maintain the target rpm in the idle control settings. As a matter of fact, adjusting the PID values don't seem to do anything. I've always had my values set extremely low, and it seems to idle without oscillation. P=2.2, I=1.9, D=.1

My warmup idle target is 1500 rpm when the temp is 50 degrees. I performed the idle valve test and my rpms will increase and decrease with adjustment of the valve. IS this test just to test if it works. iirc I took the max and min duty values of where I saw change and transplanted them into my max and min duty% int he closed loop table. min=25 and max=63.3

 

warum-up only mode is the coolant based idle duty mode.

Keep in mind there's a cranking duty table that will hold the PWM% at a static value until the crank to run timer threshold has been passed. sounds like that needs tuning.

But the simple fact your PID settings are that low means the code isn't doing anything. In the least you need more I to drive the RPM to the target. IIRC I run 50 30 80.

 

MS3 does have coolant based idle speed targets. it's the big graph on the right side of the closed loop idle settings page. perhaps you have a monitor that is only 600 pixels wide?

also, as brain said, your PID settings will NEED to be tuned. There have been many PID code changes along the way for MS3 to make the code work better.
I run: P=7, I=33, D=12 on my 01

 

Well with the MS2 I did tune my PID, and those are the values that kept the idle stable without oscillating. The table on the right is the rpm target vs coolant temp. Maybe I'm trying to tune too close to the min rpm set by my idle set screw. However, rasing my target rpm on the curve didn't seem to have any affect on the actual idle rpm.

I'll raise the target rpm to something like 1200 when the car is warm, and then start the PID tuning again.

This sounds like the ticket. I have not played with this table. I also didn't really know what it did. I may have two issues going on.

EDIT: And here I thought my current PID values just meant I was awesome

 

The P value is the authority value. No P = no authority to modify idle valve duty cycle to correct error.

 

my mistake on idle DUTY target versus RPM target.

but with closed loop PID idle, you don't get to pick idle duties anymore (for now). the ECU does that.

incidentally, if you find you aren't reaching a set target, your Integral gain is too low.

 

Ok, then I really think what is going on is that I was trying to tune the PID too close to the idle rpm I set with the idle screw.

 

The mariob code has a temp to idle valve duty table added over the normal firmware. I think it was used as an initial value when you came into closed loop.

 

after the crank to run period in MS3, the current idle duty value is saved in memory and the MS always reverts back to the last value when coming back to closed loop + the dashpot.

 

Any tips for what a reasonable? I changed it to 65, 55, 45, then 19 when fully warm

 

As soon as the "crank to run taper" times out, (3s) closed loop idle kicks in and the car dies.

 

Appears that ,y idle open duty value set to 63.3 wasn't high enough

 

Victory is MINE! Thanks everyone. I ended up with 25 20.1 and 150. I had multiple combinations that worked. This is where I got tired of fiddling.