I've been experiencing an issue for a few months now. Sometimes the 4G will run fine, sometimes it will feel sluggish. Occasionally, while I'm driving and the car feels sluggish it will suddenly lurch forward and drive normally again. Sometimes it will do the opposite. I had decided it was likely either a flakey TPS or PPS.
I've been to the dealership. They're no help. They told me that the problem is that I'm running premium fuel. They're certain that filling it up with 85 octane will cure the problem, and that Mitsu doesn't know what they're talking about when the manual and filler cap both say premium only. How, I love those discussions...
Anyway, I decide that the only way I'm going to get it fixed right is to troubleshoot it myself. So I buy a DashHawk and start datalogging.
From what I can tell, there are two TPS and two PPS involved. All four are appparently poorly calibrated at the factory, so the ECU averages and normalizes their readings. Easy enough.
What I expected to find was a single PPS or TPS giving an inconsistent reading. I figured when it spiked one direction or the other it would throw the average off and cause the throttle to partially open or close.
Looking at the datalog, I can see the throttle sporadically open/close about 10-15%. These correspond perfectly with my throttle surge issues. Great news, as this confirms that the issue is throttle, not timing or MIVEC engagement. However, when these events occur, the two PPS track perfectly with one another. The two TPS also track perfectly with each other, but both with the big spike. Both TPS also perfectly track the ECU Command Throttle signal. It looks like the ECU suddenly just decides that the throttle needs to open (or close) an instantaneous 10-15%, even though both PPS readings are constant.
I have a hard time believing that the PPS are at fault. Like I said, they track perfectly with one another (though offset in calibration by about 10%) and match my pedal inputs perfectly.
I also have a hard time believing that both TPS sensors are identically flawed. They also track perfectly, though one is really poorly calibrated and reads ~53% when the throttle is really at 2%. However, correct for the offset and it tracks perfectly with the other TPS and both track the Commanded Throttle signal.
So I think, maybe its a bad wiring harness...but the PPS is stable (matches my pedal input) and both TPS track the ECU command signal. So ECU???
So I wonder, perhaps the ECU is seeing another spurious signal which is causing it to occasionally dial back or add throttle. I've tried turning off the TCS, but the problem persists. I've logged timing and IAT, but both act normal when the problem occurs.
So what do you guys think?
1. Two bad PPS that have identical flaws? Too unlikely.
2. Two bad TPS or bad actuator? I don't think so, since they track the ECU commands perfectly.
3. Bad ECU? This would be an awfully specific ECU problem.
4. Bad wiring harness? While the issue is sporadic, it isn't random. The spike always occurs when I'm accelerating and the pull-back always occurs when I'm decelerating.
5. Some other sensor I haven't thought of that would cause the ECU to occasionally pull throttle? I don't see a way to directly track knock with the DashHawk, but I don't see timing getting pulled which I would expect in a knock condition.
What am I missing? There has to be a simple explanation that I'm missing. Otherwise, the ECU seems to be the least unlikely of the options I have.
Any feedback is appreciated.
I've been to the dealership. They're no help. They told me that the problem is that I'm running premium fuel. They're certain that filling it up with 85 octane will cure the problem, and that Mitsu doesn't know what they're talking about when the manual and filler cap both say premium only. How, I love those discussions...
Anyway, I decide that the only way I'm going to get it fixed right is to troubleshoot it myself. So I buy a DashHawk and start datalogging.
From what I can tell, there are two TPS and two PPS involved. All four are appparently poorly calibrated at the factory, so the ECU averages and normalizes their readings. Easy enough.
What I expected to find was a single PPS or TPS giving an inconsistent reading. I figured when it spiked one direction or the other it would throw the average off and cause the throttle to partially open or close.
Looking at the datalog, I can see the throttle sporadically open/close about 10-15%. These correspond perfectly with my throttle surge issues. Great news, as this confirms that the issue is throttle, not timing or MIVEC engagement. However, when these events occur, the two PPS track perfectly with one another. The two TPS also track perfectly with each other, but both with the big spike. Both TPS also perfectly track the ECU Command Throttle signal. It looks like the ECU suddenly just decides that the throttle needs to open (or close) an instantaneous 10-15%, even though both PPS readings are constant.
I have a hard time believing that the PPS are at fault. Like I said, they track perfectly with one another (though offset in calibration by about 10%) and match my pedal inputs perfectly.
I also have a hard time believing that both TPS sensors are identically flawed. They also track perfectly, though one is really poorly calibrated and reads ~53% when the throttle is really at 2%. However, correct for the offset and it tracks perfectly with the other TPS and both track the Commanded Throttle signal.
So I think, maybe its a bad wiring harness...but the PPS is stable (matches my pedal input) and both TPS track the ECU command signal. So ECU???
So I wonder, perhaps the ECU is seeing another spurious signal which is causing it to occasionally dial back or add throttle. I've tried turning off the TCS, but the problem persists. I've logged timing and IAT, but both act normal when the problem occurs.
So what do you guys think?
1. Two bad PPS that have identical flaws? Too unlikely.
2. Two bad TPS or bad actuator? I don't think so, since they track the ECU commands perfectly.
3. Bad ECU? This would be an awfully specific ECU problem.
4. Bad wiring harness? While the issue is sporadic, it isn't random. The spike always occurs when I'm accelerating and the pull-back always occurs when I'm decelerating.
5. Some other sensor I haven't thought of that would cause the ECU to occasionally pull throttle? I don't see a way to directly track knock with the DashHawk, but I don't see timing getting pulled which I would expect in a knock condition.
What am I missing? There has to be a simple explanation that I'm missing. Otherwise, the ECU seems to be the least unlikely of the options I have.
Any feedback is appreciated.