1G GSX starts everytime - but dies in 5-10 seconds...

[JohnnyPC1]

Hey fellas!

I'm somewhat new to this DSM scene. I have a 91 GSX that will start everytime but then dies after about 5 - 10 seconds. So far I have replaced the fuel pump, ISC motor, fixed a idle control switch to close when the throttle is shut and replaced the capacitors in the the ECM. Still the same trouble...drats!

The car will start everytime and seem like everything is OK but then dies. If I step on the throttle it has little or no effect on keeping the engine running...even if I rev it to 3000 rpms it will still die. I bought the car with the same symtoms so whatever I have done to date has had no positive effect...

What should I verify next since I'm kinda running out of things I know to check...I've scanned the FAQs and forums for similar troubles but none really jumped out at me

Thanks!

 

[Slipstream]

okay, I'll take a stab at this one.

first the questions.
Does it require a resting period between starts or can you restart it as soon as it dies?
When was the last time you changed the fuel filter?
when was the last time you canged plugs/wires?

Here are a few first thoughts. common sense says air+fuel+spark=running. The fact that it will start but won't continue to run makes me think that you're losing 1 of the 3 once the engine starts turning. Check your plugs after it dies...smell gas? lots of it? are the plugs wet? in short, maybe the engine is flooding itself after it gets going. maybe it isn't getting enough air/isn't exhausting enough dead air to keep the ignition cycle moving, check for obstructions in the intake and the exhaust. just some idle thoughts to check on. how's it sound after it is started or when you rev it? does it die more quickly if you rev it or is it still in the 5-10 second window? and lastly, how does your voltage look both off and on?
-Slipstream

 

[realhp=turbo]

Check the MPI relay.

 

[JohnnyPC1]

Thanks for writing guys!

-The car will restart right after it dies...1-2 seconds on the starter and it lights right up...
-The fuel filter was also just changed...it's getting pressure evident by the smelly mess I made when the banjo fitting wasn't tight all the way...
-Plugs looked black, I cleaned them up but had no positive effect on the issue...

Losing one of the three, air fuel or spark, was what I was thinking about last night after I got fed up with it. While running, it will allow maybe a couple quick revs and then opening the throttle results in a "bogging out" effect....bleh blehh blehhh type of thing...

My question back to y'all is what happens "after" the car is running in it's first few seconds? If it starts then it must be doing something right - it just doesn't continue. How does the starting cycle differ from say when it's running?

If I were to guess, I'd say it was loosing spark because it just seems dead. What would be a test I could run to illustrate if this is what's happening? Where does the ECU find the signal that says things are running and therefore fires a spark? Cam sensor? If this were bad would it allow the motor to start? BTW the ECU reveals no codes in it (just blink blink blink real steady) but then again I did just re-do the caps inside it...do codes get set straight away when a sensor has bitten the dust?

I have a serial cable coming for my palm so I can hook up a datalogger to it. I'm going to make that pressure tester cap to see if I have a big leak someplace. And I'm taking the throttle body apart to make sure things aren't all carbonned up...

 

[crankbender]

Try this it only takes a few seconds. Get under the car and drop the downpipe from the o2 housing by a half inch or so and then restart the car. If it keeps running your cat is probably clogged up.

you can also drop the turbo a half inch off the manifold if you prefer to be on top of the car...you do not have to remove the parts just give the air another way to get out.

 

[JohnnyPC1]

Awesome - I'll give it a shot on Monday...Thanks!

 

[iamtall77]

it might be something more simple. like maybe the MAF plug is off. i forgot this once and that's how it runs. it could also be a very large vacuum leak. try starting it and pushing on the gas a little. if it runs, but really rough, do an intake leak test and find it. those are problems i have found with my car and other cars. they are simple to check and easily overlooked.

 

[92awddsm]

It sounds like the ecu is dead. The leaking caps probably damaged the circuit board. Do you have a friend that can swap ecu's with you to see if it is the problem. The steady blink from the ecu is usually a good sign its the ecu. My old ecu did the same thing. After it was repaired, no more flashing from the diagnostics and the car ran fine.

 

[JohnnyPC1]

It sounds like the ecu is dead. The leaking caps probably damaged the circuit board. Do you have a friend that can swap ecu's with you to see if it is the problem. The steady blink from the ecu is usually a good sign its the ecu. My old ecu did the same thing. After it was repaired, no more flashing from the diagnostics and the car ran fine.

That would be the ideal scenario - but I have no friends No friends with DSMs anyhow. I got my eye on a couple ECUs from ebay and have a '90 ECU on the way already. But I found misc info that a 90 might not work in a 91. Maybe you can tell me the story on using a '90 ECU in a '91 chassis. I looked in the 1000 Qs Faq and saw that going backwards you have to do all sorts of wiring but when it mentioned the 90 to 91 direction all it said was "this works - sort of" and then points to a dead link on dsm.org. Anybody know what the situation is in using a 90 ECU in a 91 chassis?

I checked the MAF connector and found it's connected...thanks brother! I've been through the engine bay with a magnifying glass hoping to find something simple like this but so far no dice. I'm doing the pressure check tonight. Took my throttle body off last night and got it sparkling...At least when all this is done I'll have put my hands on virtually everything under the hood and got it clean again...

Keep up the suggestion doooods! Big Thanks for all your consideration!

 

[boostedinaz]

But I found misc info that a 90 might not work in a 91. Maybe you can tell me the story on using a '90 ECU in a '91 chassis. I looked in the 1000 Qs Faq and saw that going backwards you have to do all sorts of wiring but when it mentioned the 90 to 91 direction all it said was "this works - sort of" and then points to a dead link on dsm.org. Anybody know what the situation is in using a 90 ECU in a 91 chassis?


Keep up the suggestion doooods! Big Thanks for all your consideration!

You can use it but you need to change two wires to make it work. I would just find a 91-94 ECU and use that. There are other things that need to be done for it to work correctly. The 90' use slightly different ignition parts so the car may run but your tach wont work. Liek I said find a 91-94 and use that. One less thing to go wrong or worry about.

I would say its a huge vacuum leak or the ECU took a dump on you.

Michael

 

[92tsifwd]

My 91 awd is doing the exact same thing as yours exept now it won't start at all. I just replaced my ecu caps also. I have gone through just about everything trying to figure out what the problem is. Let me know if you figure out anything and I will do the same for you.

 

[steve]

The steady blink from the ecu is usually a good sign its the ecu. My old ecu did the same thing. After it was repaired, no more flashing from the diagnostics and the car ran fine.

Actually a repeating long on off on off from a 1G ECU is its heartbeat when it doesn't have any stored error codes. It's a sign that the ECU is actually running. A dead ECU won't put anything out on diag pin 1. It's not proof that the ECU is good, just that it boots and runs.

Steve

 

[JohnnyPC1]

Thanks for all the input so far! I haven't had a chance to work with it this week...BUT...I just happened to buy another AWD Eclipse. What was I thinking? That's what my wife said anyway...

I want to use the extra car to swap parts in and out to minimize my guessing. I have a strong notion that it will turn out to the be ECU but I just couldn't wait for the guy to send me my '90 ECU i bought off ebay...that had the wiring and tach issues with a 91 anyway.

The situation with the "new" eclipse is that it's an automatic; the one I been struggling with is a manual. What do I have to consider when using the auto ECU in the 5 speed? All I want to do is test it to see that my issue clears up, then wait for the 90ECU, do the rewire and y'all again about how I might fix the tach issue.

Why didn't I just look for a 91 5 speed ECU? Well it's been a matter of impatience combined with the fact that '91 up ECUs seem to go for top dollar. Found my exact PN on eBay and I was in for $181 before it went on to close at like $187...Damn!!! Hard to be a penny pincher

 

[steve]

The automatic ECU is going to run rich because of the difference in injector sizes but it should do to help you decide if the ECU is causing your current problem.

Steve

 

[mitsutuner]

Mail or drive your ECU over to ZDR electronics in Orlando & have them test it & rebuild it if needed. They fixed mine, similar symptoms & put a 2 yr/24k warranty on it. Here is the link,
http://www.mitsubishiecu.com/

 

[92tsifwd]

Just curious if you have checked your timing belt because I put a brand new timing belt on and within an hour it slipped taking a good # of valves with it and my car was doing the same thing yours was? Have you done a compression test? That will tell you if the valves are bent. I got 0psi in 1,3,4 and 60psi in 2 after corecting the timing belt. I would just check if you have not already.

 

[beinstein]

My 90 TSi was running beautifully untill i began to fiddle with the cruise control motor. For some reason, after I unplugged and replugged the motor my car would start, idle for about 3 seconds, sputter violently, and die. I spent a day attempting to get it to run properly, and nothing would help; i could give the car gas to a point, but it felt as if the car was controlling the throttle on its own, even at WOT. I figured it had to be an ECU problem, which somehow lead to me eliminating sensors one by one and starting the car to see if it would run. Nothing worked until I unplugged the MAF (MAS, mass airflow sensor, whatever you call it); which makes the car run slightly rough, but well enough to get around. When its plugged in, the BCS clicks loudly, the car sputters and dies. Im trying to get a new MAF. Hope this could help someone. Also, an explanation as to how the cruise control affects the MAF would be greatly appreciated, if anyone knows.

 

[mitsutuner]

very interesting--what do you mean by "fiddled with the cruise control motor"-were you adjusting the linkage from the cruise servo to the throttle or getting into the electrical stuff?

 

[beinstein]

Just unplugged the electrical connection that sits on top of the motor. Weird huh.

 

[nelapse]

Ive seen this before... i think its a map sensor of some sort before i bought my DSM there was one identicle with the same problem i was gonna get and it had a map sensor problem when i started it it would die just like yours

 

[Kookykrispy]

I have a 2g, but last week I was in pheonix and my car was doing the same thing. It would start every time, run for a few seconds and then die immediately. The check engine light was on. I had to bring it to a shop, they put the scanner on it: Cam position sensor. Replaced that and it runs perfect again.

I'm new to these cars, but I thought I might be able to help here...

 

[JohnnyPC1]

Welp! I think it's mission accomplished I tested the known good ECU from my other GSX (auto) in the stalling car and it appears to run fine. At least the issue I was having has abruptly disappeared. I'm going to put the old ECU back in to see if I can re-create the problem I was having but I'll do that this weekend.

I guess the capacitors I replaced in the stalling ECU weren't all that was wrong with it. Looking closer, I found that there's a capacitor (c106) that just fell right off when I nudged it a bit. Some traces are grungy too. I'm going to replace that cap and see if I can resurrect this ECU - anybody know the value of the c106 capacitor?

Also, I haven't run the thing all day long or really beat on it with the auto ECU in it, but it does seem to run properly. Anybody have more input on using an auto ECU in a manual equipped car? I have my stuff for a datalogger cable coming later this week so I'll get to see some actual numbers of what's going on. I do get a check engine light that comes on occasionally when I coast to a stop but other than that, car runs well...

Cheers!

 

[steve]

C106 is a 47uf 50v 105c electrolytic. Usually you replace it along with C107 and C108.

Sounds like you had some electrolyte leakage. Get out your magnifing glass and look for corrosion around C14, C38, C44, D1 and the pins of IC110. I just had a ECU in that behaved the same way, starts and runs for 1-2 seconds then dies after I fixed all the major damage.
Turned out the trace feeding 5v to D1 and from there to IC104-109 had microscopic opens.

You have to strip the conformal coating off to clean the leftover distributed electrolyte to avoid future problems and to see any damage. Make sure to recoat after your done repairing.

Steve

 

[JohnnyPC1]

You're the MAN! But I told you some bad info. Yes, c106 is a 47uF - but I was speaking of c37...the surface mount cap just next to c106. I read the silkscreen wrong. Do you have a vaule for c37? I was just going to replace it with a regular ceramic cap...

 

[steve]

I would guess it's another .1uf 50v ceramic decoupling capacitor like C38.

Did it take the trace from the via to R44 out when it came off?

Steve