NB2 MS3 Improving Idle Quality (logs included)
 

Hey guys,

I'm looking to improve the idle quality a bit on my brothers MS3 NB2 (details below). The end-goal is to get the car to idle, in all conditions, the same or better as an OEM car. That leads us to the first issue that I'm stuck at:

Problem: During hot idle, when you blip the throttle, it creates a period of instability and it takes around ~5 seconds for the idle speed to come back in line.

Possible solutions?
1. Maybe I could continue to speed up the PIDs for closed loop idle
2. AFR seems to follow the RPM instability, going more lean as the RPM increases, and back inline when the RPM dips

Car Details: 2003 NB2 Miata. Full bolt ons, 93 octane, MSlabs MS3 enhanced w/ 3 bar MAP sensor

What do you guys think? Anything that I'm missing? Log is attached (it's like 15 seconds hot idle with 2 throttle blips). Tuner studio project is also attached.

Your VE table looks a little jumpy.

VE isn't jumping 40% from 40 to 55 KPA, not physically possible. Same on the low RPM stuff. Fuel and idle PWM will fight each other if you have spikes like that.

Are you actually running 240kpa of boost, 20+ PSI? If not I would rebin your tables closer to the max boost you are running. Then you can add more resolution down low near idle. You also aren't using overboost, sure hope you have no turbo. If so, then rebin with 100-105KPA as the max. If you are adding a turbo, rebin for your planned max boost.

Rev 2 with VE table adjusted + some accel enrichment changes. Looks like that improved the initial lean spike, but the RPMs are still a bit unstable. We're NA currently, but I've scaled the map for boost this summer. We'll certainly take advantage of the overboost protection when the time comes.

I have very similar issues, i'm on a ms3 mini on an nb2 as well. When I'm driving around and the car is cold the idle will creep up to almost 2500 rpm and i will have almost no engine braking.