No announcement yet.

Idle issues with Megasquirt MS3X on SR20DET

  • Filter
  • Time
  • Show
Clear All
new posts

  • Idle issues with Megasquirt MS3X on SR20DET

    Hi folks,

    Been swapping my engine management for my redtop over to a Megasquirt MS3X. We got everything wired up and the car started on just a base map. Tuning at speed has been fairly successful, but we are really struggling to get idle to a reasonable place. The idle is constantly surging, sometimes really badly (+/- 300 rpm) other times less severely (+/- 50 rpm), without making any changes (with and without idle control on).

    Even more troubling than that is returning to idle when approaching a stop. The engine dies nearly every time. The RPMs will dip way down and either stumble and save it itself, or usually just flat out die.

    The first thing we are trying to verify are calibration values that are very hard to find online for some reason.

    Does anyone have this info?:

    Tomei 555cc Injector Deadtime (Website says .71ms, but no voltage listed)
    AEM UEGO 30-4110 Calibration
    SR20DET IACV Frequency

    After that, is there something obvious we should try? This engine has Tomei cams and a light flywheel, so Idle is going to be a bit of a challenge, but we've tried a lot of different stuff to no avail.

    Thanks for your help guys!

  • #2
    The wideband should be linear and found on AEMs site. They have good instructions.

    On injector lag time, start with that, with some smaller values at higher voltage and larger at lower voltage, and if you notice low end Fueling is rich adjust it down or the opposite.

    I've heard some people say the stock IACV is on some crazy high frequency like 300 Hz or something, but I had a hard time controlling it on my infinity at that frequency. I use about 20-25 hz now, and it seems to react much better. I think the high frequency can hold it at intermediate positions quietly, but it has weird deadband chwracteristics that made tuning it a pain. I'd rather hear a little clicking and have it actually control well.

    As for stalling, there are a bunch of other parameters like idle Decel etc that can affect it. But I'm not familiar with MS's idle and Decel scheme to talk specifics.
    '18 Chevrolet Volt - Electric fun hatch for DD duty!

    DefSport Koni Sleeve and Spring Perch Buy!!!


    • #3
      The question we have about the wideband is: should we trust the value on the gauge? We used the AEM provided table for calibration, but it is pretty far off from what the gauge says. We ended up using a different, pre-programmed AEM calibration and it is much closer to agreeing with the gauge. Not sure if we should trust the laptop or the gauge itself.

      I can verify your method of IACV calibration. I took the valve out of the car yesterday and cleaned it really thoroughly. Before putting it back in the car we spent a bit of time testing various frequencies and pulse widths(PW). We tested 20Hz, 50Hz, 100Hz and 200Hz. At each frequency we tested PW from 0-100. At 20Hz the valve is actually actually opening and closing at different rates, unlike the higher frequencies which "hold" the valve at different positions. At 20Hz there is control from about 15%-90%PW. 50Hz was really narrow, but 100Hz has pretty good control as well. Usable values are about 50%-95%. 200Hz has a little less range, more like 70%-95%. We are now using 20Hz as it seems to have the best control.

      We've found that once we get idle really stable with idle control off, we need a really really low IACV PW to get the car to idle without massive fluctuation. This becomes a problem when we are trying to use closed-loop idle. The idle fluctuates wildly and wont enter closed-loop(CL) when the valve is too far open, but it won't close the valve until it enters CL. Similarly, when driving and suddenly pressing the clutch the RPMs will drop straight down and kill the engine. The control of entering closed loop is a real sticking point for us. I know this is MS specific, but if anyone has suggestions or experience with this, I would appreciate any advice!


      • #4
        Hey Colin, there's obviously a ground offset causing the wideband gauge and your logger to not match... The gauge would be accurate (although probably slow response) so try playing with the calibration to get it to line up. If you can log the AN volt channel when powering up the AEM, I think they output a specific voltage during warmup, check their site for the manual. your logger will probably be off, and the difference will be important.. if the AEM outputs say 2.5v and AEM cal shows 2.5v as 1.0 lambda during warmup, and your logger is showing ~2.4v you would adjust the cal at 2.4v to correlate to 1.0 lambda.. then you'll just need to figure out the slope. verifying a couple more fields in the cal should give you enough to extrapolate the whole slope. I've not had much problems with AEM's and ground offsets, Innovates were the worst. switching to serial or ideally CAN bus widebands would eliminate any issues of course, but dealing with offsets isn't a huge affair I'd think you'd be able to get it all sorted rather quickly!

        For your idle issues, you need to check if any other table is possibly interfering with idle control... make sure idle timing values and ignition timing values match for same rpm/load points, and disable all DFCO, transients, etc while dialing it in so you aren't chasing your tail... make sure enable/disable qualifiers aren't too close, and not within the thresholds of other adders. Like if DFCO is still enabled into the idle threshold, you'll wind up with oscillating issues where both controls are trying to cancel each other out.

        Also, like you've found, idle control is only needed to HELP the engine maintain a steady idle... and when completely warm should have minimal intrusion... with a warm motor, unplug the IACV, and obviously disable all idle functions and get a decent idle if you can with the throttle plate set screw and the IACV bypass screw first... then you can re-enable idle control, and it won't have to work as hard. PWM will be much lower for that temp, idle rpms should be stable, with hardly any timing modifiers.. then you can start working on cold start, where idle control is obviously more important.

        I'll check out the mega squirt software in a bit, but if you've been able to get open loop idle working with PW tables and idle thresholds, I'd assume you'd be able to use those as starting points for the closed loop...


        • #5
          dang, my first post??? jeez I feel creepy and embarrassed now


          • #6
            Do you guys think there is any benefit to removing the OEM IACV and instead converting to one of the remote IACV blocks that use the Jeep valve? I dont know much about the two but know that that seems to be a popular option for some.


            • #7
              While it continues to be true that there is a discrepancy between my AEM gauge and the output from the controller, it seems to be close enough to trust, as long as we stay on the safe side of AFR. I have the wideband and the MS box grounded together, so they shouldn't have any issues there. The calibration we are using must have the same slope, just a different offset as we have the same error at all values. We could probably figure out that offset and calibrate it better, but I'm not too worried about it.

              Today we had some success getting into idle better, rather than the engine dying every time I pushed in the clutch. It took a lot of fiddling with IACV settings and closed loop settings. There is an "out of closed loop" IACV setting which has a big influence on entering closed loop. We've found we need that setting quite low. The valve staying open to far makes the idle fluctuate wildly and it will never get into CL. Also made some of the CL settings a little more lenient so that it was easier to enter into control. The faster we can get it into CL the faster it will stabilize the idle.

              What we found made a huge difference is making the mixture VERY rich at idle. It isn't what we really wanted, but its been so much better that it seems like we will have to stick with it. We are at a point now where it will usually catch the engine, as it slows after pushing the clutch, somewhere around 1k rpm. It isn't doing so every single time, but on a much better percentage than before. I'm hoping the next time we fiddle with settings we will get that to be a 100% thing. I would like it to idle at around stoich, but to get it to idle at stops we are now sitting around 12 AFR.

              I don't think deleting the IACV is necessary. We have good control of it and it seems to work, just need to control it properly and make it work with the rest of the car's set up.