Engine & Gearbox intermittent warm starting problem

Discussion in '7th Generation (2003-2008) [Acura TSX]' started by pauloblade, Monday 28th Mar, 2016.

  1. pauloblade Junior Member ☆ ☆ ☆ ☆ ☆

    Hi I've got an 07 accord ICTDI, had an annoying starting problem for roughly 8 months now, car runs perfectly apart from the starting, when cold it starts fine, when engines warm, if I leave it for anything more than 5 mins, when I go to start it it stutters, it always eventually fires up, but theres definitely a problem, sometimes when it does it I get a warning light on, I've got a cheap code reader which reads p0339 crankshaft position sensor circuit. so I've had the crankshaft sensor replaced(genuine Honda) but still the same, I've also changed the fuel filter for genuine Honda one. Also tried changing the fuel pressure relief valve on fuel rail as heard this is a common issue with warm start problems. Tried a different MAF sensor aswell, just to rule it out. Could it be an electrical problem with the crankshaft sensor going into the ECU from the connector? Id have thought if it was an electrical issue then surely the problem would happen occasionally from cold not just from warm. Also are these codes usually correct or could it be a completely different problem to the crank sensor? Cars done 84k with FSH,. Hope someone can help, cheers, paul
     
  2. legend-ary Moderator Staff Team

    United Kingdom Legend The Big Smoke
    3,850
    2,282
    6
    @Zebster our resident diesel specialist might have the answer ;D
     
    Loading...
    SpeedyGee and Zebster like this.
  3. Zebster Expert Advisor ★ ★ ★ ★ ★

    United Kingdom .
    2,559
    1,247
    1
    I'd normally suggest that this sort of starting problem (especially when warm) is due to excessive leak-back to the fuel tank making it difficult for the fuel pump to achieve sufficient rail pressure under cranking, either because of the rail pressure relief valve leaking (quite common, but you've already replaced it!) or because one or more injectors have become worn and passing too much fuel through the lubrication path. The generic test for this is shown in this clip...
    ...all tubing from the containers goes back to the injector return connections. You're looking for the ones that have excess flow compared to the others (it is to be assumed that the best/least leaky one is within spec).

    However, you can't ignore the apparently recurring DTC suggesting that your crank position sensor is faulty! Assuming the replacement sensor is perfect, then your suggestion of a wiring fault could be possible. Using a multimeter it should be easy to unplug the sensor and check if the circuit back to the ECM shows some degree of resistance, rather than a simple open-circuit.

    Or maybe you have two separate problems?

    Oh no, when did that happen?!?! I'm going to need a pay rise...
     
    Loading...
    legend-ary and SpeedyGee like this.
  4. legend-ary Moderator Staff Team

    United Kingdom Legend The Big Smoke
    3,850
    2,282
    6
    We have been watching you ;D
     
    Loading...
  5. pauloblade Junior Member ☆ ☆ ☆ ☆ ☆

    thanks for the suggestions zebster, il get the wiring checked first before moving onto the injectors.
     
  6. Zebster Expert Advisor ★ ★ ★ ★ ★

    United Kingdom .
    2,559
    1,247
    1
    Please do let us know how this goes.

    Was I right to assume that the same P0339 code always returns after being deleted, following a warm start problem?
     
    Loading...
  7. pauloblade Junior Member ☆ ☆ ☆ ☆ ☆

    Yes correct, that same code keeps coming up when the starting problem arises
     
  8. hedleyf Senior Member ★ ★ ☆ ☆ ☆

    The code reference to the crank sensor is normally good, I would be looking at this circuit, the sensor normally is wired as a direct component so it's message is very accurate, if it's a induction type sensor its wired direct to the ECM so the code relates to its circuit,
    If it's a Hall effect sensor then it's also wired direct so code specific,
    Hall effect will be supplied by a voltage from the ECM, will have a ECU Earth and a signal return to the ECM
    Inductive will generate it's own voltage. Either system has a circuit to test ,
    Some engines do not need the crank sensor input once the engine is running but always need it to start so this could explain why it's ok until you switch off, heat could be causing a resistance somewhere ,bad connections?
    First step would be when this happens is to have a scope wired in and check the signal output.
    If you drove it to a dealer and they connected the diagnosis reader and run snap shot of the start say for 20 seconds this would give you and them a clue as to what's the problem,ask for the printout, hopefully it will show the code induce as you see the signal fall off,
    I know this cost money but guessing and throwing parts at it will end up costing you more. 45 mins should be ample time for this and will show all the sensors inputs as the problem is happening.
    Hope this is of some use.
     
    Loading...
    Zebster and jimjams like this.
  9. pauloblade Junior Member ☆ ☆ ☆ ☆ ☆

    Thats great thanks for that info. Will post back when I've got the circuit tested.
     
  10. Nels Moderator Staff Team

    Now, now. Play nicely everyone.

    There are pros and cons for both fuel types.
    To each his own...As long as it's a Honda! :sparta:
     
    Loading...
    jimjams and Zebster like this.
  11. Zebster Expert Advisor ★ ★ ★ ★ ★

    United Kingdom .
    2,559
    1,247
    1
    I had to use the Toyota to visit the pub tonight. First time in about 5 months!
     
    Loading...
    jimjams and Ichiban like this.
  12. nauest Premium Member Club Supporter

    Netherlands Ned Texel
    460
    246
    6
    I have the exact same issue with my car, but without the fault code you describe. Hard starting but only when hot (occasional). She will almost start and take a few more seconds than usual to get going.

    I have found the problem has not got worse over the last year. I have also done a bleed off test and fuel pressure test from the injectors and found nothing major.

    I seem to find the problem can be reduced by allowing the car to idle for a little while before shutdown, which is good procedure for Turbo engines anyway.
     
    Loading...
  13. pauloblade Junior Member ☆ ☆ ☆ ☆ ☆

    Yeah that sounds identical to my problem, would have thought the problem would get worse so I can diagnose what's wrong but its stayed the same for roughly 8-10 months. Hopefully a circuit test will show up a problem.
     
  14. nauest Premium Member Club Supporter

    Netherlands Ned Texel
    460
    246
    6
    I will monitor how you get on and restart my enquiries as well, see if either of us can figure this out! Just for your reference I have checked:

    - Fuel Rail pressure, running and starting, hot and cold (All within limits)
    - Fuel Filter
    - Fuel tank
    - Injector bleed off hot and cold
    - Running temperatures

    One interesting thing that I did find is that if you disconnect the engine temperature sensor, the ECU goes into cold start mode and never fails to start.
    This did make me think that the temperature sensor over reading or cooling too slowly could lead to wrong starting parameters (Fuelling and glow plug times)
     
    Loading...
  15. pauloblade Junior Member ☆ ☆ ☆ ☆ ☆

    Add your list to mine and we've covered quite a lot, I did think could it be an engine temp sensor but thought that would throw up an error code if it failed? Gonna get my crank sensor circuit checked on Friday, after that I'm not sure what else there is to check. Already thrown quite a bit of money at it and don't want to spend much more.
     
    nauest likes this.
  16. Ichiban Founder Staff Team

    England CJ Leeds
    30,178
    6,408
    516
    Loading...
  17. jimjams Expert Advisor ★ ★ ★ ★ ★

    United Kingdom nobody zoned
    2,166
    1,242
    if you want to check the engine temperature, there is a built-in diagnostic that displays real-time values of several sensors


    SEA3E50K71100000000BBAT00_4.JPG
     
    Loading...
    Simes, Zebster, nauest and 1 other person like this.
  18. nauest Premium Member Club Supporter

    Netherlands Ned Texel
    460
    246
    6
    I wonder if the temperature sensor could just be running out of spec with age or fault. If it failed completely you would get all kinds of issues.

    Be interested to know how your circuit check comes out.
     
    Loading...
  19. BB Baboonface Valued Contributor ★ ★ ★ ☆ ☆

    United Kingdom James Chesterfield
    402
    337
    1
    Not perhaps relevant, but oddly my old Civic Type-S occasionally had an almost identical issue (the VSA model is electronically very similar to the Accord)
    It was so rare that i never followed it up, but minor research at the time brought up mention of the coolant temp sensor being a possible cause.
     
    Loading...
  20. pauloblade Junior Member ☆ ☆ ☆ ☆ ☆

    Update. so had the crankshaft sensor circuit tested and all fine, the mechanic that did it says he think he knows what the problem is, he reckons its the timing chain, says he could hear it rattle when starting the engine. Wasn't sure what to think of this so rung Holdcroft for a 2nd opinion. Spoke to tony one of the mechanics and he said it does sound like it could be the timing chain, and that would throw up a crankshaft circuit error. But without taking the chain off to inspect he can't say for definite. Well its been roughly the same for 8 months so I think I'll see if it gets any worse. If it does then I think I better get saving.