Strange problems with engine control

[Xnke]

So we've progressed from won't idle under 1500RPM to will idle at 750RPM...but strangely.

New camshaft, got all the bugs worked out with that, engine still idles high and runs very well from 2000RPM and up. Same as before...so time to fix the idle.

Started working the idle down a little at a time, got it to go along at 1000RPM. That's acceptable to start working up again. Shut the fully warmed up car off, coolant temp 190F, IAT 100F, supercharger belt slipping a little so time to re-tension (new belt).

Re-tensioned belt, fire the car back up, coolant temp now 140F. Car fires right off, idles at 750RPM VERY rough. Wideband shows VERY lean. Start richening up the idle, to get idle AFR's to 13.8-14:1, and the table values skyrocket...idles very pretty at 750RPM and 90-93VE. Stable idle...pedaling the throttle revs poorly, going lean as RPM's climb, but returns to 750RPM reliably.

Turn car off, go to check for stupidity that I might have left under the hood.

Come back, fire up the car...have to kick the pedal and get the revs to 2000RPM and it'll idle at 50-53VE at 13.8:1 AFR, again, perfectly stable but revs decently now. Idle will NOT drop below 2000 without dying, unless I choke it by smothering it with my hand...then it drops and will idle at 750, again, stable, but up in the 90-93VE range, and won't rev over 2000RPM.

You know, it'd be really awesome if I was trying to build a possessed car, but it's getting old now.

Has anyone had an EDIS module do silly stuff like this? I'm starting to think it's ignition based, as this is the third different set of injectors, both injector drivers have been rebuilt because one of them acted funny after I used some very bad injector settings, and I've calibrated and checked the wideband sensor in two different cars to make sure it's not causing the problem.

MSQ and two datalogs, one is a very strange 400RPM idle that would NOT rev at all, the other is the strange 750RPM idle. Can anyone find anything really wrong here? Firmware is 3.3.1a, re-flashed it this morning before i started working on the car to eliminate bad firmware causing strangeness.

 

Drop the ".zip" extensions when you get the file, they're only on there to allow it to be added to the post. The files are not compressed or anything.

crazyidle.msq.zip

2014-04-26_15.54.41.msl.zip

2014-04-26_15.56.16.msl.zip

[Dan_Austin]

Before looking at the logs I would have guessed IAT heat soak.  After

looking I would say it is a possibility, as both logs have 140F+ MAT in

the beginning and taper down as the car runs.   Perhaps try adding

5~10% fuel to the MAT correction  for temps about 140F would help

the lean hot restart.

 

Your idle MAP and VE look high to me, but that may well be just how

a super charged setup should look.  I've been following your threads

from the beginning and suspect you know way more than I do about

the supercharger and MS platform.  That said I wonder if the superchargers

idle bypass valve/system is not functioning as desired and you happen

to be pushing air even at idle, not enough to be 'boost' but more than

the engine needs at that load.

[Xnke]

MAT heat soak could be a problem; the MAT is in the main plenum assembly. I've removed the supercharger charge piping and am doing tests using a 38mm throttle body (it's just a test!) stuffed into the plenum coupler this afternoon, trying to at least get the car to idle properly and return to idle. I've tried the dual-throttle setup before and it worked fine...but was such a pain that I removed it, trying to keep thing simpler.

 

If I have to run two throttles, that's fine...Miata guys do it regularly, with very similarly placed superchargers...That's actually where i got the idea that this might just work from.

 

The idle MAP and VE ARE sky-high. I don't have a clue why.

Edited April 27, 2014 by Xnke

[Xnke]

Idle MAP is still high, but not so bad as it was.

 

I removed the supercharger charge piping and fitted a 38mm throttle body backwards stuffed into the 2.25" coupler on the front of the intake.

 

The result?

 

VERY reasonable 800RPM idle...with a damned rowdy cam. A perfectly drivable (although completely illegal...I had a throttle cable coming up through the hole in the hood into the open window...) 950RPM idle...which was the initial target. Battery corrections need help, as when the electric fans kick on it nearly stalls, but now I know the root cause of the engine that wouldn't return to idle.

 

The MASSIVE plenum volume. If I reduce the plenum volume, the engine runs semi-normally. This means, unfortunately, the double-throttle setup will have to return...blagh.