MS3 return to idle issue. Help needed.
 

I can't tune this out, and whilst it looks pretty innocuous it's driving me potty.

Tune and a short datalog are in the links at the bottom of my post

https://www.miataturbo.net/attachmen...ine=1385507684

If you see the RPM line, this is post a blip of the throttle. What should happen is the last know good value of the idle valve DC gets a dashpot added to it (and I've tried from 3% to 15%, the same thing happens albeit at higher and higher RPM's), the revs then drop to just above idle ideally 1100rpm for me, then drop to target.

What is happening is my RPMs drop to where you'd expect for the idle valve DC, but then rise again briefly before the CL idle valve engages and drives the idle back down.

Why the heck are the revs momentarily raising before they can be driven down? it's not the idle valve causing it as it doesn't really increase at all during the rev raise.

I just can't figure it out at all. Messing with the PID values doesn't have any noticeable improvement either.

https://dl.dropboxusercontent.com/u/...urrentTune.msq

https://dl.dropboxusercontent.com/u/...6_22.01.00.msl

Subscribed

From your screen cap, it looks like it's in your spark advance.
However the idle response also looks pretty good.

Any ideas why idle is advancing when I'm requesting a flat 10deg across the board when idling?

Pretty good isn't perfect ;)

I have a similar issue. Can someone expand on what specifically about the idle advance is causing this ( ? )

Too much, not enough
is it disabled ?

Thanks

+1 on Ben's analysis. In this area, your spark advance is operating off your spark map. Make sure the bins in your spark map around this area are all about the same. What's happening is that spark is advancing which produces more torque and increases RPM. Your idle duty then responds to that (with lag). As the RPM decays, the reverse happens. Bottom line, you've got a negative feedback loop. Post your spark map. Should be obvious.

MS3 has separate Idle VE and Idle Advance, if desired. There is additionally an "Idle RPM Timing Correction" that can be turned on in the same place as Idle Advance. I have never used that, but I will try it tomorrow.

IIRC the idle advance settings are so it can respond to load, it won't help with too much idle advance in your spark table causing the negative feedback loop. I had the same problem, and lowering my timing below 2500 helped, but I need to pull a bit more still.