Idle problem kinda unique ??

[Rich A]

I've read other posts here about idle problems but can't find anything exactly like my problem. (and I'm not familiar with the forum rules here and did not want to hijack a thread)

Over a year and a half ago this started. Engine starts fine. After warm up idles around 800 to 900 RPM. Very rough. When idling it will pulse up and down, each time going higher and lower until it goes below 300 or so and stalls. Like up to 1000 then down to 800, then up to 1200 then down to 600 then up to 1300 and down to 500 and so on.

Brought it to a mechanic. He worked on it for about 30 days. (I told him to take his time and do it right as it wasn't my primary transportation)

Got it back, with some sensors changed and other adjustments to the tune of 500 or so $$. Still same problem. Had an argument with them and went looking for another mechanic.

2nd mechanic: Told him to change the timing belt, do a brake job and replace the clutch AND fix the stupid idle.

He had it for NINE months. Everything worked fine (he was highly recommended) EXCEPT for the idle problem. Brought it back and another 3 months passed and after picking up the car a second time I still had the problem. He had set the idle to around 1500 rpm to keep it running. He apologised and gave me credit for some of the work and said that he frankly couldn't fix it.

Brought it to a 3rd mechanic. This guy had it 2 months. Just picked it up and guess what? Although "better" it is still doing the same thing. Now it only idles at around 1000 RPM after warm up but still runs rough and changes RPM up and down as I mentioned earlier until it stalls.

So far they have replaced, fixed or adjusted everything that I've read in these forums that could be the problem. Including timing belt, all sensors, carb. parts etc.

This is a 92 Eclipse with the dual over-head cam engine .. manual 5 speed. So far it's been in various garages for 18 months and all of the 3 different mechanics who worked on it have just admitted that they don't know what's wrong and said I should bring it to bring it to a Mitts dealer.

So it looks like that's what I'll have to do, but thought I'd ask here first as I've already spent nearly 2 grand on this beast. Or maybe someone knows a mechanic in Connecticut who will not charge me if he can't fix it ??

Yeah I know I should just junk it, but the engine is strong, the car handles great, it burns NO oil, gives me great gas mileage (when I get to drive it) and is a fun car to drive.

I wonder if it could be the actual on-board "computer" that has a problem ?? How would I test that?

 

[yokotabrat]

Sounds like it's time to spend some quality time with your car and a multimeter.

If we break down the idle control system, that might help find your problem. Here are the things that control the idle on a 1g dsm. ISC solenoid, throttle plate, ecu, FIAV, BISS, and all the associated wiring.

First things, first, though. Check for boost leaks. Any additional air that bypasses the throttle body will cause the idle to be high causing the car to idle surge. Pressurize the inlet of the turbo with 20-30 psi and check to see if there are any leaks, especially around any clamps, the throttle body, injector seals, or gaskets.

The isc is easy to test. http://dmtalon.v8eaters.com/ISC.html

BISS - base idle speed screw procedure is here: http://www.vfaq.com/mods/BISS-1G.html

The throttle plate could potentially be sticking or held partially open. I'd try to remove the cable from the throttle body and verify the idle doesn't go down or stop surging. If there's no change, then we move on to the next item.

FIAV - fast idle air valve. The bottom section of the throttle body is the FIAV. It's a small passageway that opens and closes depending on the coolant temp. When the coolant is warm, it melts a wax pellet that controls the opening/closing of a valve. When they go bad, they allow additional air to bypass when it shouldn't, causing the idle to surge. I don't know of a good way to test it other than blocking it off and seeing if that fixes the problem. JM Fab makes a nice blockoff plate that would be easy to install: http://www.jmfabrications.com/catalog/product_info.php?cPath=21_23&products_id=57

The ecu has internal chips that control the isc. If they go bad, they can no longer control the position of the isc. You can remove the ecu, open it, and see if there are any burnt chips. There's a picture here: http://www.dsmtuners.com/forums/showthread.php?t=116585 IC105 and IC107 are for the isc coils. If they're burnt, they're definitely bad. If not, they could still be dead, especially if your ecu has had any problems with capacitors leaking. Easiest would be to get a hold of a known good ecu and swap. If not, pm me and I can talk you through testing them with a multimeter.

More than likely it's one of these things. If all these don't solve the problem, them you might have to start tracking down wires, or getting a logger to see if there aren't other problems.

Hopefully I'm not leaving anything out. Good luck, and keep us updated.

 

[yokotabrat]

Oops, just realized I'm assuming your car is a turbo model. If not, all of this will probably still apply as the non turbo's still use the same idle control scheme I believe. You'll just have to do a pressure test at the throttle body instead of the turbo inlet. Good luck!

 

[Rich A]

Thanks for the detailed reply. I guess I'll have to sleuth this thing out for myself following your notes. This is not a turbo by the way. I'll let you know how I make out.

 

[yokotabrat]

No problem. Let me know if you need more details on anything I listed above. Good luck.

 

[Rich A]

At long last the car is running fine. Problem turned out to be a bad ECU.

Thanks for all the help ..