High NOx smog fail- EGR troubleshooting help

1991 Teg GS, Automatic
Failed SoCal smog dyno test with too much NOx gasses.
at 15 MPH, NOx 1033 PPM, limit is 565
at 25 MPH, NOx 877 PPM, limit is 746

Timing- passed at 17° BTDC, could not retard further than that.

Every car I’ve ever had that failed NOx failed because of the EGR valve.
I put a vacuum pump on the valve, and the engine stumbled at idle before recovering. The valve held the vacuum for several seconds. Even with 2 to 3 PSI of vacuum, it was a noticeable stumble.

Teed into the vacuum line and drove the car around. It is working. It comes on quite often at smaller throttle inputs, even at low speed. I am only seeing 2 to 3 PSI of vacuum though, which just seems low to me.

Looking in the manual, they talk about 3 to 5 PSI in that section, so I wonder if the valve is opening up enough. There is also some kind of control loop monitoring the EGR valve position. The CVC valve looks to be a vacuum regulator. There are probably different ways I could fool the system to open the valve more, or bypass the CVC and see what that does.

Has anyone else encountered this problem, failing smog with high NOx? Am I on the right track?

take it off the manifold and clean all the ports and make sure the vacuum diaphragm move freely and opens fully all the carbon build up can cause problems i recently did this and noticed while cleaning that the vacuum diaphragm pin was not fully opening i just used carb cleaner and small engine brushes to clean the ports and un stick the little pin.

Thanks. Can I take it off without removing any major components?

They do get crudded up. Here’s a picture of EGR “Plaque” from a different car:

L-R_difference.JPG

what were the hc readings?

all you gotta do is take of a vacuum hose and unbolt it from the manifold and you golden.

Welfare asked about HC numbers…
at 15 MPH- My car = 40 PPM, max is 85
at 25 MPH- My car = 34 PPM, max is 96

hmmm. yea, i was gonna say maybe you’re running lean, but those numbers don’t really support that

Hi Hondattic,
I took it off. It’s pretty clean and the motion is not impeded at all. So I am stumped here. For some reason I am running too hot even though it works, or else the EGR is not opening at the right times even though I can see the vacuum needle moving as I drive it. :shrug:
It’s a 1991 GS, body’s in bad shape and it fails smog but the AC works. Everything except the power door locks work. Should I just take the $1000 California smog program money?

Wankel Boy

what do you mean by “running hot” and it could also be a number of things do you have any mods.

Thanks for your interest. No mods on this 'Teg.

I don’t mean the coolant temperature. I assume from the high NOx that combustion temperatures are high.

New pieces of the puzzle today (first of all, I’ve been mis-reading the units on my vac gauge; how embarrassing). Vacuum between the CVC valve and the EGR solenoid is at spec- 10 inches. But the output of the EGR solenoid is weak, 2 to 3 inches when it ought to be 6 to 8.

My “The engine stumbles so the EGR must be working OK” theory could be the whole flaw here. With the EGR removed I could put the above observed levels of vacuum into the EGR and look at how much it opens. At 2 inches, nothing. At 3 inches (the highest I ever see driving) it moves maybe 1/8", perhaps 40% of all fully open. It takes 5 inches of mercury to get it to open all the way. Maybe it’s just not flowing enough exhaust gas to lower the combustion temps and NOx level.

Figuring out why the EGR vac signal is weak might be the whole key here. Since the valve’s opening has a servo controlled analog feedback loop, that’s a likely cause. Maybe the signal has deteriorated and it thinks it’s all the way open when it’s not. There’s some voltage I can check for this. Shortening the shaft of that sensor on top of the EGR valve might trick it into wanting to open more. Stay tuned…

Thanks again Hondattic. Comments are welcome by any and all.

most definetly will

rather than cheating the problem, i’d try solving it instead. it sounds like a small vacuum leak in the system. did this problem set an egr code? maybe try some starter fluid to check for leaks

No codes came on till I started messing with the vacuum tubes and monitoring what was going on.

I don’t think I will cheat the problem either now because I fully checked out the EGR valve and the feedback sensor on it and they’re fine.

I’ve looked for leaky tubes with the vac gage and haven’t found any. We decided it’s time to get rid of the car.

hmmm. well if the lines, sensor, and valve checked out fine, only thing left in the system is the solenoid. but it sounds like your mind is made up with this old peach

Agree

Yeah, that or the ECU. I’d love to stick a known, working EGR solenoid in there out of curiousity, just too see if I get better vacuum. Now it just blips the vacuum once when you start driving, and the Check Engine light comes on and then it doesn’t even try.

Even if I fixed all that, I’m just educatedly guessing it would fix the NOx levels. The half axles are starting to make noises again, too.

My wife’s family has gotten 19 years out of this car so we can’t really complain.

Can’t leave it alone

Well this thing still bothers me, even though the car is bound for retirement. I checked out the EGR solenoid valve. It opens with 12 volts as expected. The shop manual is not clear here about whether it’s a simple on/off, a pulse width modulation, or a current control.

well, the fact that the hc readings are in order, and that the egr is not reading the specified vac, would leave anyone to suspect that that is the issue here. at this point, i’d just shotgun it with a new solenoid. if it doesn’t solve it, return it. i dunno. that’s what i’d do. though typically i wouldn’t condone that

Thanks for your help. “Shotgun it”?
Agree, because I’m not about to drive around town with an oscilliscope on the hood trying to figure out the way the ECU modulates that solenoid valve.

The 'noid is $190 from the dealer but I might be able to get one at a local place that would probably let me return it even though it’s electrical. They’re pretty cool about that kind of stuff.
Odd, but this solenoid shares a connector with some other one that leads to a black box.

The other theory is that it’s some power transistor in the ECU that died a slow death.

BTW I traced the feedback sensor off the thing on top of the EGR valve all the way back to the pin on the ECU. That’s good.

Do you have an automatic, welfare?

haha, nope. in fact i’ve never dealt with the egr system on our cars before. maybe because i’ve never had a 92+ and i’m in canada