Weird idle problems and fail smog test...

[nashbearwa]

Hello, I'm new here but did read dozens of posts related to the idle and emission problems. The car is a 1992 Pontiac Grand Am, 3.3L, Automatic with 158k on it. This is what has been done to the car.

Six months ago when the check engine light came on, the car stalled at a light and wouldn't start back up. The car was taken in for service. They replaced the crank pick up sensor, 4 of the 6 injectors (they said they were going out) the computer under the coil pack, the throttle positioning sensor, oil change, new air filter and fixed an exhaust leak from a broken bolt under the hood from the cross over pipe. When I got the car back it would once in a while idle funny when I would park it. The idle would pulse from nearly dead up to 1200 rpm and back down to 500 or so. The car would only do this in park, no other gear. The shop that fixed the car had no clue why it was doing this but told me not to worry about it. Shortly after that it was time to renew the tabs on the car. The car failed the emissions test. Here are the results.


HC (PPM) CO% CO+CO2 O2
Cruise Limit.............150.............1............... .6............n/a
Cruise Emissions.......260...........7.65.........17.95.. ....... .39
Cruise Result...........FAIL............FAIL...........n/a..........n/a

Idle Limit................220............1.2........... ...6............n/a
Idle Emissions........1438...........8.74..........17.7 4........ .80
Idle Result.............FAIL............FAIL........... n/a..........n/a
RPM at Idle.....920


After reading the posts from other people I did the following.

1. The led lights on the shifter inside the car were all working except the one for the Park. I took it out of the car, cleaned it inside, checked for bad solder points and put it back in the car. The light now works but the car still runs the same.

2. Next I checked for vacuum leaks. I have not one leak. I checked the way my Grandpa taught me, with a spray can of carb cleaner. Not one leak, no change in idle anywhere.

3. The next thing I did was, as suggested, checked the MAF. I tapped on it, took it off and gently cleaned it, even borrowed one from a friend who works at a junk yard just to be sure. Car still runs the same, pulsing idle.

The car will not pulse in Drive or Reverse, but will now pluses in Neutral if left to idle for about 5 mins in Neural. It will now pulse idle in park all the time. Next thing I did was to "teach" the computer how to idle following the instructions by 04GASE1 from post " '93 3.3l pulsing idle, help please" http://www.gaownersclub.com/forum/showthread.php?threadid=47818 This post is the one that fits exactly what my car is doing, and nobody can fix. The idle is still the same.

4. The next thing I had done was to have the CAT checked out. They told me that it is not plugged (I didn't think it was as the power, speed and driving of the car are fast and no plugged exhaust hiss) but they didn't check to see if it was working just that it wasn't plugged, go figure.

5. I then took the car to AutoZone to check the computer....no codes.

I have not replaced the O2 sensor as if it was bad it the computer in the car would know it, I think...

I am out of ideas and will have to scrap the car if I can't get it past emissions test as I have no other choice. The car runs and drives too damn well to send to the junk yard. I know it has something to do with the damn pulsing idle as to why it is failing emissions tests, but every place I've taken it to have no idea why it is doing it, and neither do I.

Another board suggested replacing the electronic shift sensor on the top of the transmission. The one that tells the car where you have placed your shift lever....P R N D 2 1...but I don't know how that can effect idle so I have not done it yet and it is the only post I've found that said it and it was from 1997.

I forgot to add, the whole time I've owned the car the fan for the radiator is always on when the car is running, even when cold. I don't think the problem is related but in case it was I wanted to let you all know.

Please help.

[rixGAphx]

You have a 14 year old car with 158k miles, bad idle and bad emissions readings.

There are 3 major things on a modern car that control emissions, and they are all intended/expected to fail:
* Spark plugs.
* O2 sensor.
* Catalytic converter.

In reverse order:

1. Your cat is bad; not plugged, just worn-out.
Cars from '96+ have OBD-II computers that monitor cat performance.
Your OBD-I doesn't.
But the OBD-II computers look look at data very similar to what your emissions test shows, and they then declare: "This cat is dead!"
IMO, no working cat would EVER allow that much crap to escape.

2. Your O2 sensor is bad; dead; faulty.
Your computer is too stupid to know that, and it can't tell you anything.
When an O2 sensor is failing (but not completely dead), it still works a little bit. But it takes longer and longer (in milliseconds) for it to make decisions and send data to the computer, so the computer ends-up hunting all around and chasing the slow signals.
IMO, no working O2 sensor could allow that much crap to get to the cat to need cleaning.

3. Check your plugs, since they are prolly fouled with the rich-running you've been having.
Change them if they're more than 30-40k miles old IMO.

4. Pokesmot, our resident 3.3 expert, will be along soon.
He and many others have had horrendous problems with the Fuel Pressure Regulator causing symptoms similar to yours.
* The FPR has a vac connection, and it has an internal diaphragm that leaks.
* This especially causes a too-high fuel pressure at idle.
* When this diaphragm leaks, it canNOT be sensed by your 'gamps method' (which is also mine BTW :D ), since the 'leak' isn't sucking-in outside air with spray cab-clean.
IMO, this is why all that crap is leaving the cylinders to begin with.

* Look at your tranny: See that vac line?
It controls a modulator inside the tranny, which is used for sensing load on engine when in the running gears.
When in 'P', it's not 'sensing' anything, so it's kinda off-line.
That's why a vac-related problem can be present in 'P' or 'N' yet disappear in a drive gear.

Hope this helps,
-Rick

[nashbearwa]

Thanks for the reply Rick.

I checked the plugs. They are nice and clean, I remember they are not very old maybe 10k and 1.5 years. I put them back in. I am going to borrow the tools to check the fuel pressure to see if the FPR is bad or not. However, how can I tell the difference from the fuel pump and the FPR is the reading is off?

Since I have the older computers I'm going to assume that the O2 and the CAT being bad it is not going to affect the idle at this point. My first goal is to get the idle to behave then get the rest of it fixed.

Thanks for the help
Eric

[rixGAphx]

Originally posted by nashbearwa
Since I have the older computers I'm going to assume that the O2 and the CAT being bad it is not going to affect the idle at this point.Well, yer right about the cat:
A bad cat (not plugged) just causes dirty exhaust.

But yer waaaayy off base about how the O2 sensor works and what its purpose is.
A bad O2 sensor gives the computer bad data (or good data, but very slowly), and the computer can easily cause rough/pulsing idle.

Electronic fuel injection has been relying on O2 sensors from the get-go in the late '70's, with computers much older/dumber than yours.

[/B][/QUOTE]My first goal is to get the idle to behave then get the rest of it fixed. [/B][/QUOTE] I repeat: "When an O2 sensor is failing.... the computer ends-up hunting all around and chasing the slow signals."

By 'hunting', I mean it increases and decreases the pulse length of each squirt of fuel to find an 'ideal'.
This causes the idle to cycle or pulse just as you've complained about.

Replace the O2 sensor.

Good luck,
-Rick

[nashbearwa]

OK, new O2 going in tomorrow. However while the car was just running in the driveway, and I was showing a neighbor all the fun times me and the car have been having, I pulled off the vacuum line to the FPR just to see what the car would do. At first it idled a bit faster then slowed down a bit. After I put it back on the car that is when the idle started to mess up again. My neighbor then unplugged the O2 to see if it would make any difference. The car ran just fine with it unplugged, in fact it made no difference. No check eng light, nothing.

So it gets a new O2 in the AM and maybe a new FPR if that doesn't fix it's idle prob since it acted up when the line was removed and then placed back on the car.

Keep your fingers crossed for me.

Eric

[nashbearwa]

OK installed the new O2, let it run in the driveway. I thought it was ok, but it started, again, with the up and down idle. I can't afford to keep throwing money at this car. I am going to get a used FPR as I have no other choice. If that doesn't fix it...

Next!

Eric

[mtarggart]

I have a 96 GA SE 3.1L and it is doing the same thing. But, I had to change the Timing Chain Cover gasket because it was leaking coolant. While I had the cover off, I noticed the timing chain had about 1/4" of slop in it and the cam gear looked a little worn also, so I decided to replace the entire timing set. When I tried to get the gear off of the crankshaft, the crank turned (of course I didn't have it lined up for TDC on #1 & 4, that would have been too easy). It turned only a little bit so I turned it back to where I thought it was and then put the new chain and cam gear on without replacing the crank gear.

The Haynes manual I was using said that #'s 1 & 4 should both be a TDC with #4 at firing position for the timing marks to be in position. So I found this by removing the front valve cover and #4 spark plug and turned the bolt in the crank with a socket wrench and watched the intake and exhaust rocker arms and watched for TDC on #4.

When I reached this point, the crank was off a tooth according the the dot on the gear. There wasn't a dot on the cam gear, new or old, just a 10mm hole, but there was a mark (looked like some sort of paint) that was on the old cam gear from this hole straight out to a tooth, so I assumed this was the timing mark (yes, I know what happens when one assumes).

Put the engine back together, checked coolant level and added new oil and the engine ran very rough for a few seconds then smoothed out like normal. I shut the car off and finished putting everything else back together. Started it up and it ran fine for a few seconds and then started pulsing in park at idle. If I gave it gas, it would smooth out and run great. I thought that since the battery was disconnected for so long that it needed some time to "relearn" everything. Then the check engine light came on.

I let it idle for a good 15 minutes to let the thermostat open and circulate the coolant. With the check engine and low coolant lights still on, I drove about half a mile to the gas station and filled up with regular unleaded gas. While I was pulling out to the parking lot, the check engine light went out and shortly after so did the low coolant light. I thought everything was fine, but it still pulses when it is in park or neutral. It does a tiny bit when in gear, but hardly noticeable.

I looked and listened for vacuum leaks, but didn't find any. The one vacuum line that I took off, I put back on, but maybe it is not on right. I did notice that when I had the spark plugs out of 1 & 4 that they were pretty worn so I bought new plugs, but haven't put them in yet as it was 10:30pm when I got back to the house.

I did put in some fuel system cleaner as I thought it could be a plugged or bad injector. I had a 91 Grand Prix that three of the six injectors were bad and it acted sort of the same way, but it wouldn't run well at all no matter if it was in park or in gear or how much gas you gave it.

I will put new plugs in and double check for vacuum leaks. If that doesn't work I will try a new O2 sensor and maybe a new Crankshaft Position Sensor (had to replace that one time in the Grand Prix also).

If none of these work, I think I am just going to have to break down and take to the dealership.

Does anyone think that I might have the timing set off a tooth and it could be causing this too? I don't hear any knocking like a valve is hitting a piston, but I don't know if the valves get that close to begin with. I really hate to take into the shop as I don't want to throw a bunch of money at it since it's only worth $1500. I could put that money towards a new Vibe.

If anyone has any suggestions or ideas other than what is mentioned already, I'd love to hear them.

Also, I searched through a lot of posting on this site and tried every search combination I could think of, but no one has posted directions on how to tell if you have the timing marks set properly on a 3.1L or 3100 V6. If they did, I must have missed it but I think this may be helpful especially when one of the gears doesn't have any markings.

Sorry this is so long, I didn't think it was going to be when I first started, but wanted to make sure that I included every thing that I did.

Thanks,

Mike

[rixGAphx]

The 3100 (and it's 2.8 and 3.1 and 3400 sisters) 60* V6's are 'non-interference (or 'freewheeling'), meaning the piston crowns will never interfere with the vlves, even when the valvetrain beaks and the crank is 'freewheeling' relative to the static valves.
So don't worry about that.
* * *

On my Jeep's carb'd/distributor 2.8 60* boat anchor, the timing chain/gears control both the valve opening and the ignition timing.

Not so our FI'd crank-fired 3.1/3100/3400 engines.
The spark is controlled by the ICM, which gets its signal from the CKS, CranKshaft position Sensor.
So your ignition/sparking are still fine, since the crankshaft hasn't been affected by the timing chain/cam gear.

If the cam gear is in wrong, then each cylinder's valves are (both) opening either late or early relative to the position of the piston; but their 'overlap' (opening of the intake relative to closing of the exhaust) hasn't changed.
* If early, then the exhaust valve opens at the bottom of the powerstroke, robbing the engine of some of the power it would have otherwise produced.
AND, the intake valve is opening while the piston is still travelling upward on the exhaust stroke (trying to exhaust the spent gases); a *little* exhaust gas is forced up the intake valve/runner. You *might* get a backfire into the exhaust manifold, but this is a rare condition in a port-injected engine.
Then the piston tops-out and heads downward for the intake stroke, sucking-in the intake charge. The intake valve then closes earlier than it should, so the air-fuel mixture that's to be compressed is 'less powerful' than it would be with perfect valve timing.
* If late, then the piston is already heading downward on the intake stroke when the intake valve opens, and it stays open a little too long while the piston heads upward on the compression stroke. So the air-fuel mixture is again weaker than it would be with perfect valve timing.
The exhaust valve opens late so the exhaust is late at exiting, but this is no big deal.

In either case, early or late, the main effect is an engine that returns poor gas milage and less horsepower/torque than an engine with proper valve timing.
* * *

BUT, remember your Fuel Injection?
With Port Injection, each injector must fire for a specific length of time, and at a specific point in time relative its cylinder's TDC.
There is a CaMshaft position Sensor on the 60* that reports to the PCM.
It's no more precise than the CKS, but it tells the PCM which cylinder needs fuel (remember the dual-spark that the CKS/ICM produce? They don't know whether they're firing #1 or its 'paired' #4, 'cuz it doesn't matter).
So if you cam is out of position by a tooth, then so is the electronic signal that the PCM gets for determining fuel injection.
I doubt if the PCM is sophisticated enough to 'resolve' conflicting CKS/CMS signals.

Frankly, that is the extent of my knowledge; I don't know if one tooth would have any significant effect on the intake mixture.
Mebbe more effect at low rpm when each 'squirt' is distinct, as opposed to 5000 rpm, with 25 'squirts' per second essentially causing a blur of continuous fuel.
*Might* cause the 'pulsing' you're experiencing.
* * *

Your 'check engine' light went out 3 or 4 'drive cycles' after you had totally flumoxed the computer by diconnecting stuff; I *think* this is what it should have done.
The codes that set it are still stored; I would scan them just to know what they were.
* * *

As to checking the cam timing:
On every reasonably-normal gas engine, including the 3100/3.1, the exhaust valve be closing (but still open a tad) at the top of the exhaust stroke (TDC exhaust) and the intake valve will be opening (but not fully-opened). This is the 'valve overlap'.
So, remove the front valve cover ( :eek: ) so you can see the rocker arms.
Remove the #4 plug.
Carefully use a 3/8" wooden dowel thru the plug hole to 'feel' for TDC.
TDC-compression will have both valves closed; TDC-exhaust *should* have both valves about half-open.
If the overlap isn't nearly-equal before-after of TDC (sorry, I don't know exactly how many degrees there are supposed to be on this cam and engine), then the cam timing is off.

Hope this helps,
-Rick

[mtarggart]

Thanks for the extra advice Rick. It's a good thing that I bought new plugs as the four of the old ones were showing gaps of .070 and greater. No wonder it wanted to not run well.

But, I did find the problem. I didn't notice this before, but as I was reinstalling the alternator, I noticed a vacuum nipple sticking off the back of the plenum right in front of the alternator. I think it had a rubber cap on it at one time and the end of the cap broke off with just the rubber collar left around the nipple. So, I started the car and put my finger over the hole and it ran smooth as silk. As soon as I took my finger off, it knocked and started pulsing at idle. My biggest problem was finding a pice of host big enought to put over the hole and put a bolt in the end of the hose. I took the car out for a test drive and it has more power now that it did before I changed the timing chain and things.

Wow, what a difference. My brakes feel more solid now too. Does anyone know what this nipple is for? Is it to attach a vacuum tester to? I didn't pull any hoses off of it nor did I see or feel any loose hoses that would plug into it.

I'm sure glad that it is fixed. I hope this also helps others with thier rough idling too. I've never noticed this vacuum nipple before and I've put several alternators on the car.

Thanks agian for your help Rick.

Mike

[rixGAphx]

Damn.
Perfectly good essay on the effects of poor valve timing gone to waste :banghead:

:D :D

j/k, it was an entertaining exercise.
* * *
Glad you're rolling fine now.

That nipple does nothing for any GA that I know of.
Could easily be used for testing vac, but so could any other nipple.

Based on general design principles, and the fact that GM uses this engine in various configurations for different models, I'd *guess* that the nipple was declared 'a good idea' in case they needed a vac port in that location.

And the GA only uses vac for HVAC and Cruise Control.
Somebody mighta wanted 'spare vac' for some other operation.
Flip-up headlights, a la 'Vette and Fiero, come to mind.
Because of their critical nature, this system would have its own 'reservoir' and might even be monitored.

My '85 Jeep with part-time 4WD, and the 2.8 version of this same V6, has a vacuum-disconnect/connect on the front axle.
When rolling down the highway in 2wd, the front axle is disconnected to save gas and wear on the 4wd components.
When I get to a trail and shift the transfercase into 4low or 4hi, a vac switch connects the front (right side) axle.

When I started driving in HS, dad's old '55 Lincoln was the battlewagon we used.
Electrical systems were 6 volts at that time (the 'universal changeover to 12V was in '56), so wires and motors for the power windows and seats were huge (to prevent loss of power due to the greater ohms being carried).
If something could be done with vac instead of electrical, it was so.
The power radio antenna was vac, and so were the windshield wipers!
Climbing a hill on a rainy night was a white-knuckle experience, as the wipers would lose vac and slow to a crawl.

-Rick

[nashbearwa]

SUSCESS!!!

I got it running correctly, and PASSED the smog check.

The mystery has been solved. Even though it tested as OK, I started with the Fuel Pressure Regulator. I got a new one for 40 bucks at AutoZone. Put it in the car, started it up,

TADA it ran perfectly. Not a hiccup.

So the FPR and maybe the O2 sensor were to blame, but I not sure about the O2 as I replaced it last week and it didn't run any different.

I went to the smog check station. Instead of polluting 260 and 1438 of unburned HC it put out 2 and 6. How about that. The CAT was fine in the car, I did not replace it.

Thanks for the help. Maybe this is what is wrong with your car Mike. It is fairly easy to replace both the O2 (13.99) and the FPR (40.00) worth a shot.

Eric