Rough idle/no power/ whats wrong? - Son of a @#$%#

[Js92Tru]

Hey Guys-

Its been a while since I have been on this board. I sold my 92 Subaru Turbo Sedan to my girlfriend about a year and a half ago and bought an Turbo Audi... During that time I have had a wheel bearing, a dead instrument panel, and a bad ignition control module in the Audi, and not a single problem with my girlfriends turbo sedan! Until now that is. Here is the break down of events

---Sunday Aug, 15- car running fine, decide to be nice and wash girlfriends car after getting done with mine. While I'm at it I decide to clean the motor (probably a bad idea but who knows if this is a result of that or not) Made sure I bagged all the connections I could, once done can tell its missing shut it off and blow out the plug holes with some compressed air, seems to do the trick, drive off into the sunset.

---Monday Aug,16 through Wednesday Aug, 18 no problems that I hear of from my girlfriend car drives like it always had.

--- Thursday Aug, 19 at work and get a call from my girlfriend that the car lost all power while driving and is idling "really wierd and bumpy" (her terms). she doesn't want to drive it any more and I agree with her to leave it there walk to the mall (.1 miles away from where she was) and I will come by there in an hour once I get done with work. When I get there we start it up and it is still idling really rough like its missing but we manage to limp it the 1 mile home to get in the garage to my tools. the following is what I have done.

1. replaced sparkplugs - one fouled rest fine doesnt solve problem

2. examine plug wires even closer - sludge of some sort in one clean it but doesn't solve problem (all are firing and resistance within specs)

3. examine ignition amp- no problems detected all resistance within spec according to haynes)

4. pull codes get the following
21 - coolant temp sensor
35 - purge control solenoid
Doesn't really tell me much about the problem I am having

5. Try to clear codes but they dont seem to go away (even overnight with the batt disconnected)

6. Check tps and idle switch according to haynes manual
idle switch resistance 0 at closed throttle infinate at just a little bit off closed (no feeler gauge so I couldn't tell for sure what distance away, will probably pick one up tomorrow)
voltage at tps at closed 4.75 volts down to 1.3 volts at WOT (Haynes says should be almost opposite 0 to 1 at closed and up to 5 at WOT, could this be a problem or just haynes being stupid?)

7. recheck connections plug connections find a little more sludge of some type in a wire clean it out and decide to restart car, Idles perfectly!!! yay!! for 1 minute >=( "ARGH" back to rough idle.

8. Pull codes again get the following:
21 - coolant temp sensor
35 - purge control solenoid
31 - tps sensor!!!!!! (Could I have caused this? or is it an actual problem?)

Thats currently where I stand and I am running out of ideas my next step is to get a timing light and see if the timing is good and to make a scan tool and see if that tells me anything.

Please if you guys/girls have any ideas I would love to hear them.


Thanks in advance for any help you can provide

Jay

[Js92Tru]

Another note I'm told she filled up at Holiday Gas Station with 93 octane on Monday morning. And she did tighten the fuel cap until it clicked multiple times.

[dzx]

Did you check for a vacuum leak. It would explain the rough idle but i dont know about the loss of power. If it idles fine sometimes i doubt its the gas.

[Js92Tru]

(This is an x-post from BC-BF Legacyworks)

Just to fill in a little, Rough Idle, no power,
detonation under load, ect.

Checked: Replaced plugs, checked vac hoses, plug
wires, coil, fuel filter, fpr, fuel pump. Getting
spark on all 4 cyls, getting plenty of fuel, no vac
leaks.

I got fed up so I brought the car in and was informed
that cyl 3 wasn't firing it's injector and the noid
light wasn't even detecting a pulse signal from the
computer, therefore I was informed that a computer
replacement should take care of it. Before I just went
ahead and ordered a computer I checked continuity of
the wires all came out fine so I ordered the computer
from Canada.

Computer came in yesterday, I hooked it up cleared
codes and attempted to start. The car ran exactly the
same as it did before. Needless to say I am confused.
I guess my next best bet is going to be to swich the
injectors between 1 and 3 and see if the signal
changes or if it stays the same.

Does anyone Else have any other ideas? I will try
anything short of running to the moon and back, it's
getting to be snow time and I want to put my car away
for the winter but I can't do that until I get my
truck back from my girlfriend.

Thanks in advance

Jay

On a side note if anyone is interested in purchasing a
Turbo Legacy ECU I have an extra one on hand and will
entertain all offers that are at least 70% of what I
paid for it. It's located in Southern Minnesota and I
can ship pretty much anywhere.

[tris91ricer]

umm.. did you check the alt? I know you went through all of your electrical stuff.. but.. did you start simple? I don't see alternator anywhere in there.. rough idle and loss of all power and such to me, sounds like a bad alt --probably just the VR. Get it tested anyway.

[vrg3]

That's odd that injector 3 isn't firing but you're not getting a trouble code for it.

But still, I would suspect the wiring for the injector. When you checked the continuity, did you try both when the engine was warm and when it was cold? Did you try jiggling the connections?

Did you check that it gets a steady +12v on the other side of the connector?

The Haynes manual does have the information backwards for our TPSes.

Did you replace the coolant temperature sensor?

Did you do a compression check when you changed the plugs?

How much did you pay for the ECU?

[Js92Tru]

Sleeper- I replaced the alternator probably 2 or 3 months before that, I suppose that could cause a problem but my voltmeter(autometer wired to a ignition switched fuse) is at 12 when off and around 14 when on.

vrg- yeah no trouble code, I didn't check cold and warm and I didn't run 12v through it (whatever my radio shack voltmeter does to test continuity is what I used). Thanks for the update on the Haynes manual, that had me a bit confused.

I did replace the coolant temp sensor to no avail.

I didn't do a compression check, if I do what should the compression be? and would it all of a sudden cause a car to run great for 131000 miles then out of nowhere cause it to do this? I suppose if a ring failed but I imagine I would hear some noises that aren't too plesant if that were to happen.

Lastly I paid $104 for the ecu and would like to get as close to that as I can, I can get $70 back if I return it to the pick and pull that I got it from (they charge a 30% restock). So if your interested give me a price between $70 and $104 =)


Thanks

Jay

[BAC5.2]

I got 160psi on all cylinders when I did my compression test.

A gone ring wouldn't make much noise.

I'd suggest building a scantool and checking the Injector Pulse Width. It should be easy to spot a faulty injector that way.

It could just be a clogged injector, so you could run some seafoam through the gas and see if that'll do it.

[azn2nr]

if you havent replaced the wires do that. my car had a similar problem and i checked everything but as soon as the wires were swiched out the problem was gone

[Js92Tru]

bac - sounds like a plan, if you happen to have the link to the scantool buildup I would appreciate it otherwise I can search.

azn - are you talking about the injector harness that you replaced or a different set of wires(i.e. plug wires, ect) if it was the harness where did you get them from? Did it come as a whole unit or did you build it yourself and just buy the connectors? Any info is appreciated.


Thanks again for everyones help

Jay

[vrg3]

http://www.graphics.cornell.edu/~v/bcbfscan/

It'd be good to know at least what it's saying about the coolant temperature.

I think Jason is talking about the ignition wires. It is possible that there are cracks in the insulation that you couldn't see even upon close inspection.

[professor]

I have found through hard experience that you cannot rely upon resistance readings to determine that plug wires are good. The only way is to replace them or swap out a known, good set. A meter measures the resistance to a measly current under static conditions, a far cry from a high voltage intermittent jolt that comes from the coil. Typically you may run OK for a little while, but once the wire heats up in the bad spot, you stop firing on that cylinder

The fouled plug is trying to tell you something, sensor problems wouldn't cause just one cylinder to stop, but wires, ignitor, or injector problems would.

How often do you folks actually see injector problems ? I find them rare, and the last thing I worry about normally.