Discuss all things tuning in this section. News, products, problems and results.
Ive got a problem with my car.
It is a Toyota Celica Gt4 rally car with the Gen 4 3SGTE, it has had a freshly rebuilt motor and a brand new "milspec" wiring loom with all new sensors and electronics. ECU is a link Fury.
we first put it on the dyno and has a misfire at exactly 4800rpm under full boost (19psi wastegate). we narrowed down the problem to cylinder 2.
we have swapped around multiple set of coils, spark plugs, injectors and the problem persists on cyl 2. the compression is perfectly uniform across all cylinders.i also removed the valve springs and they were fine. i also separate wires for the injector and coil directly to the ecu bypassing the loom. we also tried another ecu and updated the firmware.
I then put a different wastegate spring (10psi) on and the missfire is still there, however not as aggressive.
The misfire occasionally goes away but is there 90% of the time
I hope you guys may have some ideas on why could be wrong (im now out of my pay grade)
PS love all your content, keep it up!
hi, not your most knowledgeable guy here but i have had a very similar issue with misfire and hot start issues on a fairly new build. my misfire was on cylinder 1 and took me a while to pin point. i changed an array of sensors, plugs and coils and nothing seemed to help. the problem boiled down to the fuel pressure regulator, there was a very small tear in the diaphragm letting fuel into the vacuum side and then tracking down to the inlet manifold where it was plumbed into above piston 1. i swapped out the fpr for a new one and problem was solved no more misfire.
Can you post up a log file taken on the laptop of the misfire happening? Do you get any trigger error counts at this point at all? Would be good to see if we can narrow it down to something configuration or sensor related, or something more mechanical :-).
thanks for your input guys
That certainly sounds like a unique problem Anthony, i did a quick check of the fpr and found nothing suspicious. its plumbed into section that would feed all cylinders equally.
Ive never had any trigger errors Zak, and the tuner seemed happy enough scoping the triggers on the dyno. I would suspect a trigger issue would effect all cylinders rather than just one?
Ive attached a log and a tune file
What injectors do you have in it?
I would try these things. Apologies if I'm repeating some things you've already done.
1. Use a timing light to watch that cylinder during a dyno run. This may help catch a trigger- or spark-related problem. Add a temporary spark plug wire between the coil and spark plug if necessary.
2. Look at all the spark plugs, make sure they are all showing similar color on the ceramic and electrode.
3. Add 20% fuel trim to that cylinder.
4. Try much lower fuel pressure, then try much higher fuel pressure.
5. Double-check there aren't any odd blockages or leaks in the intake or exhaust manifolds?
6. Swap injector 2 and injector 3, to see if the misfire follows the injector. Then swap coil2 and coil3, and the spark plug wires (if the engine has spark plug wires).
Turned the crank trigger was wired backwards
dont understand how this would effect one cylinder but it does!
Thanks very much for everyones input!
time to go racing
Glad you found and fixed it, wiring the crank sensor backwards will certainly cause problems.
There are at least two variants of 3SGTE crank sensors, but they all have 2-wire Variable Reluctance (VR) sensors if I remember correctly. The signal from VR sensors will look a little like a sinusoidal wave, but one of the edges will be close to a vertical slope while the other will be more 'diagonal'. The point where that vertical slope crosses zero volts does a good job of representing crankshaft position, in other words it will stay at the same angle compared to the actual teeth on the trigger wheel. The zero crossing of the VR sensor's 'diagonal' slope will migrate significantly compared to the actual crank teeth on the trigger wheel. This is especially easy to see if you focus on the signal near the missing teeth gap in the trigger pattern. Some ECUs can be calibrated to assume that vertical slope is on the rising edge of the signal (as voltage increases from low to high) or the falling edge (as voltage decreases from high to low), some ECUs don't allow this to be changed in software and dictate which edge must be the vertical/significant edge. Either way, wiring the sensors backwards compared to how the ECU is expecting will flip the vertical slope so it's on the wrong edge, and the ECU will now be receiving bad information. Even if the ECU doesn't detect a trigger pattern error, you can expect that the ignition timing (as measured with a timing light on the engine) will drift by a lot especially at high RPM.