Mitsubishi 3000GT & Dodge Stealth Forum banner

1 - 3 of 3 Posts

·
Registered
Joined
·
516 Posts
Discussion Starter #1
My '93 base model just started acting goofy. First, the tach jumps up 500-1000 rpm or more. (it's lying - the engine doesn't change speed) A few seconds later, 'check engine' comes on and it goes into limp mode. It's intermittent and will clear for a minute or so then do it all over again.

It returns errors 44, 52 & 53, which as I interpret the manual, means it sensed failure on all 3 coils. Thinking this wasn't very likely, I swapped the power transistor module with a known good one, but nothing changed.

Before I start making test harnesses and hooking up an o-scope to start working my way through the service manual troubleshooting procedure, has anyone else had these symptoms? Is there a place in particular I should look at first?
 

·
20 yr owner
Joined
·
1,867 Posts
I had code 53 before.
The symtom was coil for 3 and 6 cly. would
burn up.

To solve it was to replace power transister.
 

·
Registered
Joined
·
516 Posts
Discussion Starter #3 (Edited)
Infrequenty asked question with frequent answer

OK, so that didn't turn out to be a question that a dozen of you guys could answer with, "Yeah that would be....... "





But I did find the answer on 3si -





The dreaded <a href="http://www.3si.org/vbb/showthread.php?s=&threadid=47505"?> suicidal ECM capacitors </a> strike again.





Besides the symptoms I posted, I also had the ECM failure symptom of hard cranking, which started around 2 months ago, but until I read that thread, I never connected it with the ECM and had been looking for loose crank position sensors, etc. The problem I've been chasing this week didn't show up until almost two months after that.





The troubleshooting trail of following the simultaneous fault codes of 44, 52 & 53 led me to put a scope to the inputs of the transistor module. Just before the check engine light came on, a periodic noise was starting to appear at the input to the driver transistor for the #1 coil, and the noise would quickly increase in amplitude. As it approached 10 volts, it was enough to fire the transistor. Since the tach module is also in the transistor module, and takes its input from internal connections in the transistor module, it counted the errant pulses and sent a higher RPM pulse train to the ECM. I'm guessing that when the ECM saw a hgh RPM compared to other inputs, that caused it to declare that all 3 coils had died.





When I pulled the ECM, I found a little charred SMT component; too far gone to tell if it used to be a capacitor or a resistor, but most likely a resistor since it was in parallel with an SMT cap.





The service manual was a good start, but the real info was in posts here on 3si; the leaky cap problem, how to pull the ECM, what to look for, and links to more info.





The moral of the story is, if you have a car up to '94 and you haven't replaced those failure-prone caps in your ECM, give yourself a break and go do it. The problem also occurs in some other Mitsubishis, Hondas and Suzukis from 89-94 with Mitsubishi-built ECMs, so if you have any of those cars you should read the thread too.



It's worth cold hard cash to upgrade the caps. I can probably get mine swapped or repaired for a couple of hundred bucks, because the damage is limited to replacable parts. If you let the problem go on until it blows out a proprietary Mitsubishi IC, chances are pretty good that a ECM shop won't accept your old one as a rebuildable trade-in and you'll be stuck for a much higher price.
 
1 - 3 of 3 Posts
Top