Idle All Over The Place

[mluder]

OK... This is probably going to be hard to diagnose on a message board but let's see if anyone can 't point me to some things to try.

I have had this problem with the idle not sitting steady for quite some time. When the is just started up its seems fine. Once it warms up I start to get this fluctuating idle hunt. It also shows up in the dwell reading (when I push the wide-open throttle button it stops fluctuating in the dwell... not sure about idle).
See video.

https://www.youtube.com/watch?v=D-BTvpLs9gc

What have I done to try to fix it??
Most recently I completely replaced all the spark plugs, wires (except coil wire), distributor cap , and rotor.
About a year and a half ago I had completely torn apart my engine to clean, and replace gaskets, vacuum lines, injectors, water pump, etc.

As far as I know I've got good spark and I think fuel delivery is good. I checked the spray pattern about a year ago via. Have not done a fuel pressure test.

Anyone have any suggested diagnosis process?

Thanks
Steve

[DMC5180]

You have classic lambda induced idle hunt. Your idle also seems to be a little high.
I'll let the Daves jump in and walk you through the adjustment process.

[sdg3205]

Have you tried swapping out the idle ecu?


Sent from my iPhone using Tapatalk

[mluder]

Have you tried swapping out the idle ecu?


Sent from my iPhone using Tapatalk

I have not... Bought a spare last year as a "just in case" when it popped up here but it turns out it's not the correct one. Different pin configuration on the connectors.
Ugh $345?!?! And what if that doesn't fix it?

Steve

[mluder]

You have classic lambda induced idle hunt. Your idle also seems to be a little high.
I'll let the Daves jump in and walk you through the adjustment process.

Looking forward to it...

Steve

[Bitsyncmaster]

From your video, your idle hunt is not as bad as others that have the problem.

1) First thing to do is replace the O2 sensor.
2) A rich mixture can also add to hunt because it cools off the O2 sensor. Adjust mixture with dwell.
3) A sticky idle motor also can cause the problem, clean it.
4) Curb idle setting (lower screw) can run the idle motor more or less closed. The idle motors have very little torque near the full closed position.

[mluder]

From your video, your idle hunt is not as bad as others that have the problem.

1) First thing to do is replace the O2 sensor.
OK... Easy enough...



2) A rich mixture can also add to hunt because it cools off the O2 sensor. Adjust mixture with dwell.
I have done that and it is oscillating with the idle hunt but hovering around 25 on the 8 cyl scale which should be around 50.



3) A sticky idle motor also can cause the problem, clean it.
How do I do this? Carb cleaner?



4) Curb idle setting (lower screw) can run the idle motor more or less closed. The idle motors have very little torque near the full closed position.
Not sure what the direction is here? Or is this just a fact? I believe both idle screws are set correctly though it was pointed out previously in this thread that my idle is a bit high.

By the way... It's always sort of done this to more or less degree. Here's a video of my dwell reading from 2 years ago after I refurbed the engine.

https://www.youtube.com/watch?v=pT59bB06XiU

Thanks for the help, Dave.
Steve

[Bitsyncmaster]

The fact your dwell is swinging tells us the lambda system is seeing varying mixture.

You can just "play" with the curb idle screw. Try 1/2 turn out and 1/2 turn in to see if it reduces the hunt.

Yes, use most any solvent to clean the idle motor. Pour solvent into the valve body and quickly twist the unit. Don't use any metal tool to rotate the valve or you may scratch it. A wood tool would be fine.

[David T]

The fact your dwell is swinging tells us the lambda system is seeing varying mixture.

You can just "play" with the curb idle screw. Try 1/2 turn out and 1/2 turn in to see if it reduces the hunt.

Yes, use most any solvent to clean the idle motor. Pour solvent into the valve body and quickly twist the unit. Don't use any metal tool to rotate the valve or you may scratch it. A wood tool would be fine.

Before trying to fix this by adjusting things, when I see this the first things you have to do is make sure the motor is mechanically good. A compression test will tell you if all of the cylinders are working equally. If they are not you will never have a smooth idle. A smoke test will find vacuum leaks. A vacuum leak can make 1 or more cylinders lean which, again, will make them run unbalanced causing a bad idle. Any adjustment you make is "global" meaning it will usually affect all cylinders equally. Can't make up for imbalances between cylinders. Next you need to make sure the injectors and ignition system are all working evenly on all of the cylinders. On one car the problem turned out to be the distributor. The impulse coil was replaced and whoever did it messed up the gaps of the reluctor causing the ignition system to fire erratically. On another car I found someone used too long a bolt to hold the wires on the front, right side of the intake manifold. It crushed the plug on the front of the runner and made a vacuum leak mainly affecting #4 cylinder, leaning it out. The lesson here is you have to go over the whole motor and make sure EVERYTHING is as it should be or you will never have a smooth idle.

[mluder]

Before trying to fix this by adjusting things, when I see this the first things you have to do is make sure the motor is mechanically good. A compression test will tell you if all of the cylinders are working equally. If they are not you will never have a smooth idle.
Have not done this but probably should. I think AutoZone rents the tester.


A smoke test will find vacuum leaks. A vacuum leak can make 1 or more cylinders lean which, again, will make them run unbalanced causing a bad idle.
Where's the best place to introduce smoke into the system to look for leaks?


Next you need to make sure the injectors and ignition system are all working evenly on all of the cylinders. On one car the problem turned out to be the distributor. The impulse coil was replaced and whoever did it messed up the gaps of the reluctor causing the ignition system to fire erratically.
I suppose a measured fuel test - i.e. using jars collect a sample from each injector simultaneously and make sure they are equally distributed?
As I said I replaced all the plugs, wires, etc. Did not replace the impulse coil as I was told by a good authority that they almost never go bad. I have not done a spark test on the individual cylinders.


On another car I found someone used too long a bolt to hold the wires on the front, right side of the intake manifold. It crushed the plug on the front of the runner and made a vacuum leak mainly affecting #4 cylinder, leaning it out. The lesson here is you have to go over the whole motor and make sure EVERYTHING is as it should be or you will never have a smooth idle.
When I tore down the engine 2 years ago I'm pretty sure all the bolts I removed were original. Anything replaced came in a kit from one of the vendors i.e. water pump replacement kit comes with most of the intake bolts, etc.

Based on what you're saying it sounds like I could chase this until the cows come home and never find it. Probably should just chuck it and start over... :smile:

[DMC5180]

On another car I found someone used too long a bolt to hold the wires on the front, right side of the intake manifold. It crushed the plug on the front of the runner and made a vacuum leak mainly affecting #4 cylinder, leaning it out.

I'm having a hard time seeing how too long a bolt causes a crushed plug.

I'd also like to see a video of you performing a smoke test sometime just so we all understand the process involved.

[David T]

I'm having a hard time seeing how too long a bolt causes a crushed plug.

On the front of each side of the intake manifold is a large freeze-out type of plug. A cupped disk pressed into a large hole. The bolt that holds a bunch of ground wires is tapped into the manifold on the top, right side. If you put too long a bolt in that hole it goes too far and hits the plug bending it and causing a leak.

[DMC5180]

Interesting, who would ever guess that even being possible.

Thanks David

[David T]

Have not done this but probably should. I think AutoZone rents the tester.


Where's the best place to introduce smoke into the system to look for leaks?


I suppose a measured fuel test - i.e. using jars collect a sample from each injector simultaneously and make sure they are equally distributed?
As I said I replaced all the plugs, wires, etc. Did not replace the impulse coil as I was told by a good authority that they almost never go bad. I have not done a spark test on the individual cylinders.


When I tore down the engine 2 years ago I'm pretty sure all the bolts I removed were original. Anything replaced came in a kit from one of the vendors i.e. water pump replacement kit comes with most of the intake bolts, etc.

Based on what you're saying it sounds like I could chase this until the cows come home and never find it. Probably should just chuck it and start over... :smile:

I didn't say you could have a bad impulse coil, I said if went bad and was changed and the gaps are messed up on the reluctor the motor will run badly. A bad impulse coil would cause the motor to not run at all. Testing the injectors is more than just fuel flow. A bad pattern can mess up a cylinder too. Did you carefully gap each spark plug the same? Did you put the secondary ignition wires back exactly as they were? Is the distributor cap on all the way? I plug the inlet to the mixture unit and pump smoke in through the hose from the crankcase inlet.

[sdg3205]

I have not... Bought a spare last year as a "just in case" when it popped up here but it turns out it's not the correct one. Different pin configuration on the connectors.
Ugh $345?!?! And what if that doesn't fix it?

Steve

No! Don't buy a new one, but if you've exhausted all other possibilities it only takes 1 minute to swap an idle ECU from a fellow owner.

I seriously doubt its the ECU, but hey who knows. Wouldn't be the first time a new problem came to light. Ask me about my PPR o-ring debacle.

[David T]

No! Don't buy a new one, but if you've exhausted all other possibilities it only takes 1 minute to swap an idle ECU from a fellow owner.

I seriously doubt its the ECU, but hey who knows. Wouldn't be the first time a new problem came to light. Ask me about my PPR o-ring debacle.

I brought up the problem with the bolt because you would never find such an unusual vacuum leak without doing a smoke test. On one car I did when I hooked up, smoke started pouring out of the left/rear pontoon. When we opened it up there was no carbon canister! The hoses were just hanging loose in there. On another car, EVERY injector seal was shooting smoke. Very common to find the air pipe for the idle motor to be leaky too. Another unusual one was no "O" ring on the cold start valve. You could take the whole motor all apart and find this stuff but using a smoke machine is a lot more efficient. Unless you can control ALL of the air going into the motor (meaning NO vacuum leaks!) you will have a hard time getting it to idle correctly and set the mixture. Bad idle ECU's are rare but we have seen cases on this forum where there was a problem with them. Before assuming a bad ECU you should rule out all other possibilities. Even if you do have a bad ECU and replace or fix it, it can't work correctly if you still have other problems like vacuum leaks or things not adjusted properly.

[PJ Grady Inc.]

It's definetly not a bad ECU so don't waste time or money swapping it. Rather than doing compression or leakdown testing (much better indicator of engine health BTW) at this stage of diagnosis I would simply adjust the CO plug allen screw back and forth a half turn or so to see if it stabilizes in either direction. Too rich or too lean will both cause hunting. Ideally the dwellmeter should oscillate back and forth between 40 & 50 degrees when the enrichment is spot on.
Rob Grady
No! Don't buy a new one, but if you've exhausted all other possibilities it only takes 1 minute to swap an idle ECU from a fellow owner.

I seriously doubt its the ECU, but hey who knows. Wouldn't be the first time a new problem came to light. Ask me about my PPR o-ring debacle.

[alexwolf1216]

It's definetly not a bad ECU so don't waste time or money swapping it. Rather than doing compression or leakdown testing (much better indicator of engine health BTW) at this stage of diagnosis I would simply adjust the CO plug allen screw back and forth a half turn or so to see if it stabilizes in either direction. Too rich or too lean will both cause hunting. Ideally the dwellmeter should oscillate back and forth between 40 & 50 degrees when the enrichment is spot on.
Rob Grady

Silly question I know, but where do I connect the leads for the dwell?

[PJ Grady Inc.]

Silly indeed! It's in the tech manual. Shame on you for not having one! You're not supposed to ask questions on info that's readily available ;-) I'm gonna leave that for you to look up.
Silly question I know, but where do I connect the leads for the dwell?

[alexwolf1216]

Silly indeed! It's in the tech manual. Shame on you for not having one! You're not supposed to ask questions on info that's readily available ;-) I'm gonna leave that for you to look up.

Ive got one, but not at work :hihi2:

[PJ Grady Inc.]

You can't work on your car at work can you....ha ha I can!
Ive got one, but not at work :hihi2:

[alexwolf1216]

You can't work on your car at work can you....ha ha I can!

Stop being ugly and rubbing it in.

So I pulled it up on my phone. I cant seem to locate connection instructions, but it says that it should be 35-45˚ with sensor in. I just got my dwell meter from eBay, and I plan to tinker tonight. I read to leave the air filter on, bring to operating temp. I am assuming that I pull the plug on the fuel distributor and adjust using the allen wrench.

Throw me a bone here Rob!

[PJ Grady Inc.]

I need to be in the shop right now. If no one comes to your rescue I'll toss you the bone later. Can you say "Arrf Arrf?
Stop being ugly and rubbing it in.

So I pulled it up on my phone. I cant seem to locate connection instructions, but it says that it should be 35-45˚ with sensor in. I just got my dwell meter from eBay, and I plan to tinker tonight. I read to leave the air filter on, bring to operating temp. I am assuming that I pull the plug on the fuel distributor and adjust using the allen wrench.

Throw me a bone here Rob!

[alexwolf1216]

I need to be in the shop right now. If no one comes to your rescue I'll toss you the bone later. Can you say "Arrf Arrf?

I found it, finally. Orange wire at the diagnostic plug, then ground to the block :elmo2:

[PJ Grady Inc.]

See I knew you could find it yourself.
I found it, finally. Orange wire at the diagnostic plug, then ground to the block :elmo2:

[DMCMW Dave]

It's handier at the bulkhead connector. Top left pin on the blue plug, back-probe it. Orange there too.

[mluder]

OK... So I should see oscillation at the dwell. Rob says 40-50 and Alex says 35-45... The shop manual agrees and says centered on 50. So, why the lower number Rob?
BTW - a half a turn seems excessive and would probably put my car way out... I found as little as an approx. 2 degree adjustment swings 5 or 10 on the dwell reading.

At this point... I will clean my idle speed motor as instructed by Dave M.

With apologies to Dave T... I'm not up on all the specific lingo for the motor... Can you explain to an idiot where to input the smoke?

Thanks everyone... I'm determined to get this figured out.

Cheers
Steven

[DMC5180]

Better do a short video showing the smoke process. Where do you buy a can of smoke anyway? I know beekeepers use smokers to calm the hives, but I have no clue where to find those.

[alexwolf1216]

Ha my dwell has alligator clips. Now what. 33038

[Bitsyncmaster]

Ha my dwell has alligator clips. Now what. 33038

Clip it onto a finishing nail and you have a "probe".

[David T]

Clip it onto a finishing nail and you have a "probe".

There is a smoke machine made and sold primarily for automotive testing purposes most commonly for testing evaporative emission systems. It is able to make smoke and pressurize a little. You remove the air cleaner, block the intake, and push the smoke in the hose that goes to the air cleaner from the oil filler.

[mluder]

Thanks, David T... Based on your first instruction I thought that was the way but wanted to be certain.

There are a ton of videos on YouTube for home made smokers to test for vacuum leaks. I'm gonna probably go that way if Auto Zone doesn't have a rental unit.
One way or another I'll try to do a video this weekend if I can get around to testing it.

Cheers
Steve

[mluder]

Ha my dwell has alligator clips. Now what. 33038

You can also just pull apart the connector at the diagnostic plug on the left side of the engine. Clip the green alligator clip onto the spade connected to the orange wire (not the diagnostic plug side) and the black (ground) to the engine block. I use the ground screw located on the left side of the manifold.

Cheers
Steve

[PJ Grady Inc.]

Hi Steven,

I like it when people use a real name when posting but I guess I'm old school!
To answer your question page 122 Step 6 of the tech manual says "dwell should be steady within 40 to 50" degrees with the O2 sensor disconnected. In Step 13 with the sensor connected "dwell reading should fluctuate between 35 to 45" degrees.
I was going by memory but in my experience most B28's fluctuate between 40 & 50 degrees when set up to 1% CO specs with the O2 connected (using MY 35 year old Sun dwell meter). "If it ain't broke don't fix it" I say. This information is from an old tech using old equipment. Your experience may vary but you may attempt this at home!
Yes a 1/2 turn is a lot but I suggested it because the engine should run better in one direction than the other and I wanted him to be able to notice the difference.
Rob


OK... So I should see oscillation at the dwell. Rob says 40-50 and Alex says 35-45... The shop manual agrees and says centered on 50. So, why the lower number Rob?
BTW - a half a turn seems excessive and would probably put my car way out... I found as little as an approx. 2 degree adjustment swings 5 or 10 on the dwell reading

At this point... I will clean my idle speed motor as instructed by Dave M.

With apologies to Dave T... I'm not up on all the specific lingo for the motor... Can you explain to an idiot where to input the smoke?

Thanks everyone... I'm determined to get this figured out.

Cheers
Steven

[alexwolf1216]

OK... So I should see oscillation at the dwell. Rob says 40-50 and Alex says 35-45... The shop manual agrees and says centered on 50. So, why the lower number Rob?
BTW - a half a turn seems excessive and would probably put my car way out... I found as little as an approx. 2 degree adjustment swings 5 or 10 on the dwell reading.

At this point... I will clean my idle speed motor as instructed by Dave M.

With apologies to Dave T... I'm not up on all the specific lingo for the motor... Can you explain to an idiot where to input the smoke?

Thanks everyone... I'm determined to get this figured out.

Cheers
Steven

I got 34-45 from D:04:15

Pulse Ratios:
A. Normal Operation (Sensor connected and
CO properly adjusted) —
35-45° (reading pulsates)

Very well may not have been the right spot. I am a complete green pea, I dont know this engine from a hole in the ground. So, dont take anything I say to be worth anything.

[mluder]

Hi Steven,

I like it when people use a real name when posting but I guess I'm old school!
To answer your question page 122 Step 6 of the tech manual says "dwell should be steady within 40 to 50" degrees with the O2 sensor disconnected. In Step 13 with the sensor connected "dwell reading should fluctuate between 35 to 45" degrees.
I was going by memory but in my experience most B28's fluctuate between 40 & 50 degrees when set up to 1% CO specs with the O2 connected (using MY 35 year old Sun dwell meter). "If it ain't broke don't fix it" I say. This information is from an old tech using old equipment. Your experience may vary but you may attempt this at home!
Yes a 1/2 turn is a lot but I suggested it because the engine should run better in one direction than the other and I wanted him to be able to notice the difference.
Rob

Thanks, Rob. It pays to read thoroughly. I read the 40 - 50 degrees but not the part about the O2 sensor being disconnected nor about the 35 - 45 with it connected. Do you have an opinion on if the readings are better taken with the O2 sensor disconnected? In other words, can a fouled O2 mess up an accurate read or do you want it connected as it runs to compensate?

Thanks
Steve

[DMCMW Dave]

Thanks, Rob. It pays to read thoroughly. I read the 40 - 50 degrees but not the part about the O2 sensor being disconnected nor about the 35 - 45 with it connected. Do you have an opinion on if the readings are better taken with the O2 sensor disconnected? In other words, can a fouled O2 mess up an accurate read or do you want it connected as it runs to compensate?

Thanks
Steve

If you disconnect the O2 sensor you are reading the open loop value of the computer (45). You can also ground the O2 sensor lead to the ECU and put a battery on it to test the extremes, this is actually covered in the manual. D:04:15

Don't adjust the CO with the O2 sensor disconnected, unless you have a CO Meter sniffer and are trying to set baseline.

Once in a while you'll get a bad O2 sensor that causes the ECU to read <20 no matter what the adjustment is doing. That's probably a fouled sensor. Replace it.

[David T]

If you disconnect the O2 sensor you are reading the open loop value of the computer (45). You can also ground the O2 sensor lead to the ECU and put a battery on it to test the extremes, this is actually covered in the manual. D:04:15

Don't adjust the CO with the O2 sensor disconnected, unless you have a CO Meter sniffer and are trying to set baseline.

Once in a while you'll get a bad O2 sensor that causes the ECU to read <20 no matter what the adjustment is doing. That's probably a fouled sensor. Replace it.

Typically the sign of an old/fouled, contaminated O2 sensor is much slower reaction times, ie, it fluctuates slower or not at all. They are supposed to be changed every 30,000 miles at the same time you adjust the valves, and reset the Lambda counter. You should be doing a tune-up also. This was meant to also be every 3-4 years based on about 10,000 miles a year. I would still go by the 30,000 miles but would add "or every 10 years".

[DMCMW Dave]

Typically the sign of an old/fouled, contaminated O2 sensor is much slower reaction times, ie, it fluctuates slower or not at all. They are supposed to be changed every 30,000 miles at the same time you adjust the valves, and reset the Lambda counter. You should be doing a tune-up also. This was meant to also be every 3-4 years based on about 10,000 miles a year. I would still go by the 30,000 miles but would add "or every 10 years".

This all goes out the window on a "back from the dead" car. I've seen cars with an indicated 10,000 miles and a dead O2 sensor. You don't know why someone parked the car 20 years ago, and in the mean time may have run it with the fuel distributor wide open and toasted the O2 sensor.

On a 30-year old car with a known history of a speedometer that fails all the time, mileage on the odometer is about the least accurate way to determine maintenance.

[PJ Grady Inc.]

The procedure the factory recommended was to disconnect the O2 sensor, set the base idle at 950 rpm and using a CO meter set CO to 1%. After that is done reconnect the sensor and check the dwell. The whole factory procedure has to be red through and followed from start to finich if your setting it up with a emmisions tester but is the most accurate method and helps with diagnosing problems.

As Dave already pointed out without a CO meter this can't be done so if you're using a dwell meter to set the CO up you have to adjust it with the sensor connected. My suggestion is try to get the reading close to 45 degrees as the median reading as it will normally pulse up and down slightly.
As to the second question yes a fouled sensor (or numerous other problems) will mess up your reading so in that instance the factory method is more accurate for adjustment purposes. If your engine circuits are functioning correctly than the dwell meter method for CO adjustments is OK.
Rob Grady


Thanks, Rob. It pays to read thoroughly. I read the 40 - 50 degrees but not the part about the O2 sensor being disconnected nor about the 35 - 45 with it connected. Do you have an opinion on if the readings are better taken with the O2 sensor disconnected? In other words, can a fouled O2 mess up an accurate read or do you want it connected as it runs to compensate?

Thanks
Steve

[alexwolf1216]

I adjusted the dwell. It needed like half a turn left. It didn't even read at the beginning

http://youtu.be/qyFpi39zgsQ

[mluder]

So I have constructed my smoker to test for vacuum leaks and posted it in the how-to section.
http://dmctalk.org/showthread.php?11479-Smoker-To-Find-Vacuum-Leaks

I have not yet figured out which smoke generating method I'm going to use and subsequently not been able to try it on my engine... Hopefully this weekend.

Cheers
Steven

[PJ Grady Inc.]

I previously suggested start with half turn increments to make it obvious which direction was better and you could fine tune it from there. So how does it run now as you didn't really say how things turned out after the adjustment?
Rob


I adjusted the dwell. It needed like half a turn left. It didn't even read at the beginning

http://youtu.be/qyFpi39zgsQ

[alexwolf1216]

After adjusting for a while it's running like a champ! I hear some rattle, almost like a grind, after I rev but I haven't had a chance to diagnose. Drove it today but it got too hot, it's 80 here in San Antonio.