No announcement yet.

MS2: Idle extremely rich when engine hot/heat soaked

  • Filter
  • Time
  • Show
Clear All
new posts

    MS2: Idle extremely rich when engine hot/heat soaked

    We recently "finished" the tune on my new MS2 install, but are having an issue with the idle when the car is really hot (i.e. coming off the track). MAT and CLT temps are pretty normal and average, around 90-100*F MAT and 185*F CLT, which is pretty much the temps I see on the street and what the car was tuned at. However, oil temps are a good 20-40*F hotter than on the street (about 225*F on the track), and the engine bay in general is just "heat soaked" when coming off the track.

    What happens is that even though the MAP reading and RPM range is similar to when the car is cooler (although RPM varies because it's so rich), the car is so rich at idle coming off the track that it just wants to die. Target idle AFR is 14.5 and it hits that with only a couple of % +/- of EGO correction when the car is at a "normal" temperature. Coming off the track, AFR dips into the 10's and 11's and it just wants to die unless I nurse the throttle. If I shut the car off and let it cool off for 5-10 minutes, the idle is completely back to normal.

    The car is running open loop (warmup) idle. The VE table at idle and in the surrounding areas seems pretty solid. Logging shows that the Pulse Width is actually lower when this occurs, which is the opposite of what you'd expect when it is so rich. EGO correction is already trying to pull as much fuel as it can (7%) when this occurs. WOT AFR's are normal, cruising seems normal, if not slightly on the rich side at lower RPM's. Mainly, the issue is just at idle.

    Things we have tried already:
    - Replaced vacuum lines on PCV and put hose clamps on all of them
    - Put zip ties on the FPR and MAP vacuum lines
    - Replaced the hoses going from TB to brake booster (w/ hose clamps)
    - All gaskets are new (TB, IM vacuum port, IM gaskets, bitch tube o-ring, valve cover gasket and plugs)
    - The 3 TB ports where the booster lines and the ICV elbow connect to are JB welded so they won't be a leak source
    - The injectors (Bosch EV6 24#) and FPR (Bosch OEM 3.0bar) and fuel pump (OEM) are all brand new
    - Replaced MAP sensor with a new Delphi 3 bar (external)
    - Relocated the MAP sensor from under the IM to on the fender by the fuse box to rule out a heat soaked MAP sensor
    - Cleaned out the ICV with carb cleaner
    - Verified proper resistance on the ICV coils (at both the ICV and at the 55pin connector)
    - Turned off the MAT correction table (which only had minimal authority anyways)

    Things that we have not tried yet:
    - Smoke test the engine to see if any other vacuum leaks
    - Verified fuel pressure with a gauge
    - Replacing the ICV
    - Replacing the brake booster check valve
    - Replacing the PCV
    - Changing the ICV frequency in TS from 124hz (current) to 92hz. I don't know how much effect this would have, but I have read that the frequency of the stock ICV is 98-100hz. The closest available settings in MS2/Extra are 92hz or 124hz. My base tune came set at 124 so thus far we had just left it at that.

    It seems logical in my mind that the ICV could be sticking (more towards closed) when it's very hot. The DME then thinks it's more open, and is still feeding it the normal amount of fuel, but since there is less air, it shoots to extremely rich and wants to stall. My car has a 284/272 cam and it idles in the 75-80kPa range. It does not seem likely to me that this is a vacuum leak, with all that has been done, and that the idle is otherwise rock solid when the car is "cooler" (I say that because MAT and CLT temps are similar regardless of oil temps). I will still do a smoke test anyways.

    I'm at a loss. Any ideas on what to look for next would be appreciated!

    EDIT: On a whim I just probed pins 4 and 22 on the 55 pin connector again and found 196ohms resistance, until I re-seated the connector at the ICV. It went back to ~41ohms, which is within spec. I pulled back the rubber sheath over the connector and it was damp/oily inside. This is a new lead to chase. There may be a break in the wiring, or the connector and ICV contacts might just need a really good cleaning. Maybe it's time to finally buy a brand new engine wiring harness from BMW.
    Last edited by Digitalwave; 02-15-2018, 07:47 AM.

    Did some more testing. Put a fuel pressure gauge in. Fuel pressure is spot on at all temperatures, and operates as expected. Cleaning the ICV (again) and cleaning the connectors did not change much. I put hose clamps on the ICV and ICV elbow as a precaution again vacuum leakage, did not change anything though.

    Have a few things in the tune to try next. Will keep the thread updated in case anybody ever has a similar issue.


      Updating the thread, but not much to share. I've spent countless hours pouring through the tune, and I don't think the tune is the cause of the issue, nor is has it been easy to use Tuner Studio and data logs to sort out what is causing the issue. I can say that enriching the idle (from 14.5 down to 14.0 target) has made it much more stable and happy overall. But the car still idles incredibly rich immediately after a track session (about 12.0, even with -12% fuel pulled from EGO control).

      The pulsewidth is lower when this happens, begins the ECU is trying to pull fuel. So it seems to be an air density/volume issue.

      Some thoughts/musings:

      1) nando brought up that the stock intake manifold could be getting heat soaked, and that the air inside the intake manifold might be significantly higher than the air measured in my intake tube (before tb elbow), where the IAT sensor is. Moving the sensor into the intake manifold somehow (drill and tap?) would prove or disprove that theory.

      2) It could be some sort of heat soak or change affecting the injectors.

      3) My original thought was it was a faulty ICV, but after reading how it operates with MS, I doubt that is the case. It seems that in MS, the ICV is always open the same amount (after tapering down from the warmup PWM). It doesn't open or close while you are driving around, like I thought it did. I thought when it was hot, it was failing to open the correct amount.

      4) I do intend to make sure the throttle stop is set properly on the throttle body. I have never checked that.

      5) I notice that each time this happens, the idle drops from the normal ~1,000 RPM, down to about 800 RPM. I don't know if this is simply due to the fact that is it so pig rich or not. Nothing else is changing. The MAP reading changes by a small amount, but that is commiserate with the change in RPM. I'm using open loop idle.