Our Community is 940,000 Strong. Join Us.


'92 PAU with EGR valve issues


doorunrun
09-01-2008, 03:57 PM
Hello all,
I've been having this unsteady idle trouble for a while. I've posted the situation a couple of times. I had to take a short break from it, but now I'm back on the case.

I'm focusing my attention on the EGR valve as the cause of the rough idle; I think I've eliminated just about everything else. But really the point of this post is to carp about the digital EGR valve design and why do I not want to shell out $200+ to replace it.

There seems to be several styles of this valve. On my PAU the EGR valve's tension springs are exposed. I pulled one from the junkyard off a NA PA and the tension springs are somewhat encased in an aluminum sleeve. Here's just a quick review of how this valve is set up.

The device consists of three electrically operated solenoids that drive little "pintles." These are little piston like things that have an mushroom shaped end similar to that of a C-clamp screw. There is a chamber with 3 different sized holes so the computer can vary the amount of exhaust gas returned to the intake. So, the pintles open and close. From what I've read about these in the non-engerized state they should cut of the flow of air through the hole.

When I removed my EGR valve, I found the mid-size hole was leaking, the pintle is not sealing. On the salvaged EGR valve, the small hole does not seal. Well I had high hopes of swapping pintles but alas, on my version of the valve, the pintles are captive at least the salvaged valve can be moved around to other positions.

I don't know which design is better, but I've got to go back to the junkyard and find one that really seals or one similar to what I've got so I can swap pintles around.

How frustrating! I spent a long time cleaning up each unit only to find there something about the pintle design that keeps it from fully seating. The best I can come up with is there are two springs on each pintle. The real problems seems to be the small inner spring on the pintle that must fatigue and not push the pintle fully against the plate. On my stock EGR valve I found the defective channel's pintle sat about 1mm shallower than the other two.

If I can't come up with another salvage EGR that seals properly, then I guess it's the Ebay route. Sigh...............

dreager
09-06-2008, 02:20 AM
I got one on Ebay for $35, but after reading posts here I found the EGR valve creates more problems that it's worth, (And may actually reduce gas mileage) so I did what one post suggested and sealed it off with piece of tin-plate. Some say it runs hotter as a result but I've had no trouble in 30,000KM and I'm still getting 30-35 on long trips.

Bearing in mind that most of the cam sensors on these 3.8's are shot by about 100k and will ALSO cause rough idle I just chose to leave well enough alone.The final decision is yours.

Good luck with it.

doorunrun
09-06-2008, 07:34 PM
I should have mentioned I have just under 250k miles on my Ultra. Thanks for the advice; I'll give it a try. I was able to piece together a working EGR valve by taking the best parts from my original and two salvaged valves. It probably would be better to just seal it up though, as you suggest rather than have the thing wear out again.
After looking through the manual a little closer, I was a little surprised to see that only two of the solenoids are used/driven from the ECM. I haven't looked into closely enough to discover which of the three holes are not used.
It's time to get out the tin snips. BTW, I have a scan tool coming, I hope I can discover the root of the rough idle. Which, I might add, seems to worsen when I'm at a dead stop; it's as if the speed sensor has something to do with it. The car can be rolling ever so slightly and the idle sits right around 1000 rpm. At dead stop, the rpm drops and that annoying loping idle starts. Ugh!

HotZ28
09-07-2008, 08:06 PM
It's time to get out the tin snips. The EGR was designed as a `metered intake leak' and was developed to reduce combustion temperatures to below 2,500 degrees, (the threshold where NOx is created). The EGR valve meters a readily available inert gas (actually exhaust gas, which contains a lot of very inert Carbon Dioxide) into the combustion chamber to effectively reduce the volume. Less displacement means less fire, and less heat and thus lower temperatures, thereby controlling NOx emissions.

I am not sure if your state/city/county requires emission testing, however if so, the car will not pass the NOx test with the EGR blocked. If the engine stumbles or stalls at idle or lacks power under load, the EGR may be staying on and not shutting off. On the other hand, if the engine is pinging at cruise, the EGR may not be turning on to cool down combustion chamber temperatures. If the EGR is blocked, or otherwise malfunctioning, the intake charge will be much hotter and will eventually damage the valves in the cylinder heads. Much more costly to repair than the EGR! :crying:

doorunrun
09-07-2008, 09:01 PM
I took another stab at things and I'll leave the EGR valve alone for the moment and assume it's doing it's job.
I just came across some posts about "idle learn" and there seems to be several procedures floating around.
After messing with the car today....reinstall the original ICM and clean up the body ground near the battery. I tried my initial stab at doing the idle learn procedure as outlined by Mr. HotZ28.
Here's the drill if you haven't read it:
1. Disconnect the negative battery terminal for 5 minutes or so to clear memory. Make sure to not have climate control active when you disconnect.
2. Reconnect negative battery terminal, turn key to ON position. Turn off A/C or blower fan.
3. Start car without stepping on accelerator. Put foot on brake and put car into drive.
4. Don't touch the throttle and let the car run for 5 minutes...no more than 7.
5. Turn off the engine, the idle has been set.

Now, it seems like it might help if the engine was up to operating temperature before disconnecting the battery. I didn't do this. There are posts that suggest you should start the car a second time and let it run again for a while.

We'll see how things are tomorrow during the daily drive to work. But, what's the current thinking on the "idle learn" thing?

Add your comment to this topic!