Joined
·
1,288 Posts
WOW! Really good description, usually takes 3 or 4 posts to get all that info. I don't believe that it would be the egr because that would merely cause the engine to start firing less fuel and performance would go. If it was stuck open it would be really "stuck" like jammed open and would cause performance issues at all speeds and RPMs. It would mostly just cause a loss of performance, not a shaking issue like you have. I would say test the car in different gears, but you seem like you tried that.
My suggestion to finding the problem would be to accelerate to just over 35 (about 37) and get the car shaking. As soon as it starts to shake, get it into neutral AND TURN THE ENGINE COMPLETELY OFF. WARNING!!!!</span> You will lose ALL power steering. If you do this make sure you do it on a long straight away. You will be able to make minor course corrections, but even a lane change will be unbelievable difficult, but no problem to keep it straight ahead. If the car is still shaking at 35mph, then you can totally rule out the engine (it's not running, just sitting there, literally). However, if the car stops shaking when you turn off the engine, then you know without a doubt if it's the engine or not. This will instantly let you rule out the engine or whether it's the problem.
Now because you said that it's a front to back motion and not a side-to-side or a generalized shake, it definitely stands to reason that it is a problem of rapid acceleration and deceleration, which of course implies a drive train problem. Here's my 2 Theories:
1) Harmonic Balancer is shot: It's a LOOOOOONG shot, but it's possible that at 35mph, wheel rotational force is somehow imparting additional force upon the rest of the drivetrain, somewhat making the drivetrain fight itself and additional vibration in the drivetrain that the harmonic balancer cannot compensate for. This is a big *I think* as I'm not positive what the harmonic balancer does.
2) This is my better idea, and I bet 1 no one's though of it yet.
<span style="font-family:Optima">TORQUE CONVERTER CLUTCH LOCK-UP FAILURE
Oh damn I'm good!!! What I believe is happening is that the PCM has gone somewhat berzerk in the sense that it cannot determine when the best time to lock-up the TC is. What I think is happening is that it will lock-up, and then detect a slowdown of the turbine (tranny side of the TC) and so unlocks the TC to provide added torque response. It basically isn't making up it's mind about whether to lock-up or not. NOW, through the computer in the mix and it's 15 million cycles (15MHz processor) of calculations per second, and you get it pulsing the TC lock-up solenoid (what actually locks up the TC clutches) This rapid pulsing can cause a front-to-back shape. It's exactly as though in a manual car, you push the clutch in a little bit so you have slippage (exactly as though the TC unlocked) and then let go and let it snap back together and they clamp together (like the TC has just locked up). NOW, do this even 20 times per second and you get a front-to-back shake. And for sure, the computer can do it ALOT faster than 20 times per second.
Now, WHY? At somewhere between 30mph and 45 mph, the engine RPM and the car's speed through the transmission get close together, and at one point the engine side of the TC and the tranny side of the TC will be spinning at the same speed. At this time the TC *should* lock up to stop slipage and increase gas mileage while cruising. And so there is a sensor on the turbine that reads turbine speed and then there is the engine RPM. When they get close to each other, the PCM locks up the TC if it thinks you are cruising.
AND SO: I believe that MAYBE the turbine speed sensor has failed, so the PCM cannot detect the turbine speed, and the programing default is to lock up the PCM at 35 and above and to not lock it up below 35. Yet, when you accelerate past 35 it makes up it's mind and keeps the TC locked up. So then when you stay at 35, it doesn't know if your going to accelerate past or slow down. And, when it locks up, you lose torque and will slow down a tiny bit, but enough to drop below 35, so it unlocks, and you have more torque, so you accelerate past 35 and it locks, and everything starts all over again. The only way to test it though would be to wire in a relay on the TC lock up solenoid and connect it to an LED. The idea being when the TC locks up the light should be on, when unlocked its off. And when it does it's crazy bucking lock/unlock crap, it should flash really fast. Be sure to use a relay though, because you don't want to burn up the PCM or the solenoid. Using a relay will keep the LED and it's power source out of the engine's wiring harness and using a small relay on the solenoid line will not be any difficulty for the PCM to set of it and the solenoid. Have fun. I think I've answered my theory on the problem. Let me know and have fun!!
-mobiuslogic
Note: I just realized just how f****** long this is. Sorry to steal the thread, just think this is like 4 or 5 posts all put together.
EDIT: Also try the neutral test as by putting it in neutral with the engine on at 35mph, should tell the PCM to not lock up the TC. Oh and have the codes read, it should through a code if it thinks the TC sensor has failed.
So here's your test procedure:
1) Get to 35mph
2) Shift to Neutral
a) If shake continues, problem is likely engine. Shut off engine to confirm. If shake stops with engine, engine is the problem.
B ) If shake stops with neutral, then problem lies in the tranny. However you've ruled out the tranny itself so the problem should be the TC lockup I stated earlier.
3) Have codes read and confirm the problem.
4) REPORT BACK AND LET US KNOW!
Do that and you should nail the problem down and know what to fix. Also, I don't think that the TC itself as a whole is FUBAR (as in it got eaten up) as it would not be working at any speed, thus suggesting the lockup clutch problem.
My suggestion to finding the problem would be to accelerate to just over 35 (about 37) and get the car shaking. As soon as it starts to shake, get it into neutral AND TURN THE ENGINE COMPLETELY OFF. WARNING!!!!</span> You will lose ALL power steering. If you do this make sure you do it on a long straight away. You will be able to make minor course corrections, but even a lane change will be unbelievable difficult, but no problem to keep it straight ahead. If the car is still shaking at 35mph, then you can totally rule out the engine (it's not running, just sitting there, literally). However, if the car stops shaking when you turn off the engine, then you know without a doubt if it's the engine or not. This will instantly let you rule out the engine or whether it's the problem.
Now because you said that it's a front to back motion and not a side-to-side or a generalized shake, it definitely stands to reason that it is a problem of rapid acceleration and deceleration, which of course implies a drive train problem. Here's my 2 Theories:
1) Harmonic Balancer is shot: It's a LOOOOOONG shot, but it's possible that at 35mph, wheel rotational force is somehow imparting additional force upon the rest of the drivetrain, somewhat making the drivetrain fight itself and additional vibration in the drivetrain that the harmonic balancer cannot compensate for. This is a big *I think* as I'm not positive what the harmonic balancer does.
2) This is my better idea, and I bet 1 no one's though of it yet.
<span style="font-family:Optima">TORQUE CONVERTER CLUTCH LOCK-UP FAILURE
Oh damn I'm good!!! What I believe is happening is that the PCM has gone somewhat berzerk in the sense that it cannot determine when the best time to lock-up the TC is. What I think is happening is that it will lock-up, and then detect a slowdown of the turbine (tranny side of the TC) and so unlocks the TC to provide added torque response. It basically isn't making up it's mind about whether to lock-up or not. NOW, through the computer in the mix and it's 15 million cycles (15MHz processor) of calculations per second, and you get it pulsing the TC lock-up solenoid (what actually locks up the TC clutches) This rapid pulsing can cause a front-to-back shape. It's exactly as though in a manual car, you push the clutch in a little bit so you have slippage (exactly as though the TC unlocked) and then let go and let it snap back together and they clamp together (like the TC has just locked up). NOW, do this even 20 times per second and you get a front-to-back shake. And for sure, the computer can do it ALOT faster than 20 times per second.
Now, WHY? At somewhere between 30mph and 45 mph, the engine RPM and the car's speed through the transmission get close together, and at one point the engine side of the TC and the tranny side of the TC will be spinning at the same speed. At this time the TC *should* lock up to stop slipage and increase gas mileage while cruising. And so there is a sensor on the turbine that reads turbine speed and then there is the engine RPM. When they get close to each other, the PCM locks up the TC if it thinks you are cruising.
AND SO: I believe that MAYBE the turbine speed sensor has failed, so the PCM cannot detect the turbine speed, and the programing default is to lock up the PCM at 35 and above and to not lock it up below 35. Yet, when you accelerate past 35 it makes up it's mind and keeps the TC locked up. So then when you stay at 35, it doesn't know if your going to accelerate past or slow down. And, when it locks up, you lose torque and will slow down a tiny bit, but enough to drop below 35, so it unlocks, and you have more torque, so you accelerate past 35 and it locks, and everything starts all over again. The only way to test it though would be to wire in a relay on the TC lock up solenoid and connect it to an LED. The idea being when the TC locks up the light should be on, when unlocked its off. And when it does it's crazy bucking lock/unlock crap, it should flash really fast. Be sure to use a relay though, because you don't want to burn up the PCM or the solenoid. Using a relay will keep the LED and it's power source out of the engine's wiring harness and using a small relay on the solenoid line will not be any difficulty for the PCM to set of it and the solenoid. Have fun. I think I've answered my theory on the problem. Let me know and have fun!!
-mobiuslogic
Note: I just realized just how f****** long this is. Sorry to steal the thread, just think this is like 4 or 5 posts all put together.
EDIT: Also try the neutral test as by putting it in neutral with the engine on at 35mph, should tell the PCM to not lock up the TC. Oh and have the codes read, it should through a code if it thinks the TC sensor has failed.
So here's your test procedure:
1) Get to 35mph
2) Shift to Neutral
a) If shake continues, problem is likely engine. Shut off engine to confirm. If shake stops with engine, engine is the problem.
B ) If shake stops with neutral, then problem lies in the tranny. However you've ruled out the tranny itself so the problem should be the TC lockup I stated earlier.
3) Have codes read and confirm the problem.
4) REPORT BACK AND LET US KNOW!
Do that and you should nail the problem down and know what to fix. Also, I don't think that the TC itself as a whole is FUBAR (as in it got eaten up) as it would not be working at any speed, thus suggesting the lockup clutch problem.
