No announcement yet.

Random rough start

  • Filter
  • Time
  • Show
Clear All
new posts

  • Random rough start

    Hi all,
    my m3 has got an issue that is driving me mad for years now. It randlomly starts very rough, like it is missing a cilinder. This Can be at Cold start (mostly) or at a hot start (sometimes), it occurs about 25% of time and disappers after about 20 sec.

    I changed plugs, wires, cap, rotor, brandnew AFM, tested the injectors for leaking, but the problem is still there. Because it is randomly, it is hard to diagnose, was better when it always were present. My last tnought is the ecu not firing an injector?
    maybe one of you Guys recognize this,
    Thanks all.



  • #2

    sounds like an injector not spraying correctly. Have you just checked for leaks but also if they are opening/spraying properly? Cleaning them with a ASNU device might help. Another possibility could be, that the wiring to the injectors is faulty.




    • #3
      Hi Nils,
      all 4 injectors are professionally cleaned and tested, so they shouldn’t be the problem.


      • #4
        I think it's unlikely to be the ecu. Seems like it would either fire all injectors or none. I can't imagine why it might only fire 3 of 4. If one injector is not firing it's more likely to be the injector itself or as Nils suggested bad ground.

        Have you considered fuel pressure issues? Maybe you have a failing pump and on initial start it has trouble getting the feed line from 0 to 40psi but once it's up to pressure it's making just enough to maintain.
        Last edited by Mick; 07-08-2021, 10:44 AM.


        • #5
          Hi mick, Thank you, that sounds very logical indeed. The fuelpump might be the problem. Will check it, Thanks!!!


          • #6
            Fuel pump should achieve 3.0 bar/43.5 psi whilst the return line flows 0.75 liter in 30 sec. Injectors are batch fired in pairs so if only one is missing check for a good 12V supply to each as the ecu switches the pairs to ground (not to 12V as one might think). ETM is mandatory to check this.


            • #7
              I had a similar problem. It turns out that the o2 sensor heats itself during a cold start. This wasn’t happening on it’s own but would run smoothly when the exhaust/sensor reached a higher temp.


              • #8
                Originally posted by Mick View Post
                I think it's unlikely to be the ecu. Seems like it would either fire all injectors or none. I can't imagine why it might only fire 3 of 4. If one injector is not firing it's more likely to be the injector itself or as Nils suggested bad ground.
                I don't know the internal schematics of the e30 m3 ECU, but if it's not too different from the 80s norm, it uses individual transistors to trigger each injector. So, conceivably if one transistor went bad, it could just affect one injector signal and not the others...


                • #9
                  Dont think that was the norm in the 80s tbh but as I wrote (implicitly) in my earlier post the motronic 1.1 generation of ECUs on our M3 has 1 power transistor for two injectors. All 4 injectors share +12V (live when ignition on and during start). Injector 1 & 2 share the wire to power transistor 1 into the ECU that pulls it to ground to fire and injector 3 and 4 share power transistor 2.

                  Since only one injector works intermittently the transistor is fine as else two injectors would show the symptoms.
                  My take on it is that it is a bad contacting inside connectors or damaged wiring from the splice in the loom to the one injector.
                  The wires to the splice of either +12V or to earth might be bad(ly corroded).
                  Checking this unfortunately requires cutting the sheeth around the wires but a wiggle test or continuity test is non-intrusive.
                  Continuity to the same coloured wire on the injector next to it for both wires.

                  As a temporary fix (or proof that the wiring is fine) you could bridge the wires of the same colour at the injecor connectors thus circumventing the splice one colour at the time.
                  If the symptoms remain the circumvented splice is fine. If the symptoms disappear you found the bad splice/bad connection.
                  Trouble is there is.little room in the injector harness housing to do so...

                  Here are schematics
                  #955 are the injectors
                  4 red/white (rt ws) joined/spliced
                  2 brown yellow (br ge) spliced to terminal 14
                  2 brown white to terminal 15 of the ECU
                  Click image for larger version  Name:	11.jpg Views:	0 Size:	334.6 KB ID:	1302838
                  Click image for larger version  Name:	12.jpg Views:	0 Size:	442.2 KB ID:	1302837
                  Report back what you find please! Would like to know if the 20min typing this were worth the effort :-)
                  Attached Files


                  • #10
                    I would also check the wiring first. I once had the problem with my car esp during damp weather conditions that the car was sometimes not firing on all cylinders. This was due to bad wiring to the injectors. My car used to have an immobilizer that had been removed later again by the previous owner. This immobilizer was integrated in the injector wiring. When it was built back they closed the open wires very badly. After repairing that everything was fine again.


                    • #11
                      Hi all,
                      I finally found the problem of my random rough starts. It appeared that my head had a very small crack in it, right above the valveseat. So a little bit of coolant dripped into the cilinder. After start the coolant caused the spark plug to fail.


                      • #12
                        Wow. Thanks for the update and glad you found it. Expensive, but hopefully you're done chasing that gremlin.


                        • #13
                          Thanks Mick!
                          I am really really happy I found the Gremlin indeed. I was at a point that money didn’t count (not unlimited of course :-) but I had to find the problem. Indeed it is an expansive case, but So happy!