Hi guys,

I am trying to figure out why my car has a poor idle and want to test the EGR system.

What are the parts that need to be tested?
And instead of testing each part individually is it possible to somehow disable this system to save time/effort in case it is working as it should?


Background info:
KJ2
New plugs, filters, TB cleaned, sensors tested, idle speed tested, engine mounts checked, vacuum leaks checked, PCV system checked.
Car runs smooth, has good power and good economy, the poor idle is the only issue.

Remove the EGR valve and check to see that it has a good seal when closed, ie. you can't blow through the valve. If you can blow through it with no vacuum applied to the diaphragm then there's your problem, replace your EGR. If it seals ok, return it to the car and leave the vac hose disconnected and blocked. Under these circumstances the EGR valve will never open. Alternatively, remove the pipe rising up from the manifold that joins at the other end of the plenum chamber and replace the gasket with a blank, possibly cut from a Coke can, to seal off the exhaust gas.

I would only do this to help diagnose the issue. If there's no change, return it to normal operating condition.

Thanks Ensoniq, how about testing the solenoid though?
And are there no other parts to the EGR system?

That's a bit of a loaded question. The EGR system is basically the valve and the solenoid, but it is integral to the whole engine management system. If there's an issue with vacuum, eg. a blockage in the TB, the system won't work. Also, if any of a number of different sensors are a bit dodgy they'll give false info to the ECU which might open the EGR valve when it shouldn't. I had a stalling issue that turned out to be a MAF on the way out which fed crappy data to the ECU, which kept the EGR valve open at low revs. Nothing actually wrong with the EGR system, even though that was what was causing the stalling issue. The solenoid can be tested by disconnecting it and running 12V across the terminals, from memory it should allow air flow when unpowered, and closes when powered, though if you disable the EGR valve the solenoid won't make any difference working or not.

You can test the system by taking the vac hose off the EGR valve and putting your finger over the end so you can feel if there's any vacuum (or use a gauge if you have one). With the engine running and cold, opening the throttle quickly should generate no vacuum at the EGR. As the engine warms up, opening the throttle quickly should generate a brief bit of vacuum. If it's doing this, the EGR system is working correctly.

Thanks Ensoniq,
I am assuming/hoping that since the car runs well and economical that the MAF is fine. There were no error codes stored either; not sure if a faulty MAF will cause an error code?

Mine didn't light the engine warning light or throw a code until it died completely, at which time the ECU used default airflow settings and the fuel economy blew out to something like 35L per 100k! To be honest, I'm not totally convinced that a dodgy EGR system would cause a rough idle without also affecting running at other speeds, particularly since idle is regulated by the ISC. If your TB is really clean, tight and has an operational wax pellet, no vac or gasket leaks, clean and working ISC and correctly set TPS, tight and clear PCV circuit, clean filters, good plugs, leads and dizzy, clean injectors, no splits in the intake duct, etc. etc. and your idle is still unacceptably bad it might be worth running through the full TB reset procedure, in case it's been fiddled with at some time in the car's history and is outside factory settings. A slightly low, lumpy idle is a pretty common 3rd Gen thing apparently, mine's got a bit low again since cleaning the TB about a year ago but is in no danger of stalling and isn't really a big deal. Checking all the above is a lot of work for no guarantee of improvement, I guess this needs to be weighed up against how bad your idle is.

It never stalls, or feels like it will stall, but at times it gets pretty bad.
I've noticed that after a reset (battery disconnect) it idles a bit better for the first 200km or so. Not sure of this can point to something?

The reason I ask about the EGR is I read if it does not work properly it will act as if there is a vacuum leak and a symptom of that is fine running at throttle and poor idle.

A leaking EGR, or one that's open when it shouldn't be, will have a similar effect to an air leak in the inlet manifold, leaning out the air/fuel mix. At moderate to wide throttle openings this doesn't make much difference, but at very small throttle openings obviously the effect is much more pronounced as a small leak will affect the mix much more. The thing is, at idle the ISC is effectively regulating idle speed and will make allowances for any small leak, so it would be less noticeable. The more common symptom of a leaky EGR is stalling or staggering just as you take the revs off idle, eg. as you are taking off at the lights with a light to moderate throttle. If you can very slowly open the throttle from idle and the revs rise evenly and smoothly I'd be surprised if the problem is related to your EGR valve.

If I understand the EGR valve's purpose properly (which is questionable), it should only open on fairly decent acceleration to add a bit of exhaust gas to the mix. This has the effect of lowering the temperature in the cylinders which prevents the creation of the more noxious smog-producing fumes. I don't recall the names of the gasses involved but I'm sure a bit of Googling would shed some light. In functional terms, all that's important is that the valve shouldn't open at any time other than when you're caning it a bit. If the valve itself isn't leaky and the vac line is off at idle, or you've blocked the exhaust gas duct, and your idle speed doesn't improve, the problem's definitely not with EGR.

On resetting the ECU by disconnecting the battery the ISC returns to a neutral, default setting which usually gives a quite high idle speed (this can feel better but it is usually uncomfortably high, especially going from N to R or D in an auto with the accompanying lurch). Over a few hundred kilometres the system learns the ideal settings and the idle speed should settle down to a lower speed. Possibly something is sufficiently far out of whack in your TB that the system cannot fully allow for it, but really it's a bit of a needle in a haystack situation. You mention your TB is clean, which is kinda unfortunate because I would have said that a clogged vac port or ISC channel was the most likely culprit, or a maladjusted TPS. If it's intolerable all you can really do is systematically eliminate all possible causes for less than perfect running, and as there are so damn many that won't be much fun.

Systematically eliminating all possible causes is not financially feasible on a car that is worth so little.
The hope is to eliminate possible culprits via testing rather than replacing, and if all things are found to be okay then maybe replace some of the remaining possible culprits.

Do you think that an experienced mechanic could diagnose the poor idle?
I've had it to the local mechanic, and though it seems his testing was limited he was unable to diagnose the culprit.
Though I think all he did was check for codes, vacuum leaks and fuel leaks.

I suspect you'd need to bring it to someone who really knows Magnas, like Mal down here. Not sure if there's a Mal-equivalent in NSW, other members should be able to help there.