91 3fe rough idle no codes mystery (1 Viewer)

[Tinkercrincle]

Hi all, I have been trying to diagnose a rough idle on my 91 3fe and am running out of ideas. The truck has 185k and is desmoged. Symptoms: the truck starts right up and runs mildly rough at 1200rpm then settles normally down and stays at 650 where it has a very pronounced irregular miss/rumble. Somtimes it is enough to shake the truck but the idle never really gets down below 600. No hesitation, no stalling and revs and accelerates with full power. After it is warm the miss is imperceptible Above 800rpm. Felt like a vacum leak so thats where i went first, methodicly propaned the engine bay with no change in idle. Next i hooked up the vacuum gauge and its steady at 21in at startup and 20.5 with slight, slow .5 wander at 650rpm. That ruled out a major intake leak but i did a smoke test just to make sure. No vacuum leaks. From there i have tried to work through all the systems without being able to fix the problem. Here is what i have done so far.
Electrical:charging system tests good. Coil and Ignitor swaped with known good ones. New plug wires distributor cap rotor and pickup coils with air gap set per fsm. Plugs oem gaped .043. Power balance test drops cylinders evenly, and inline spark tester shows good spark at each plug. I also opened the wiring harness from the fire wall all the way to the injectors without finding any burnt wires. Also new o2 sensors. New water temp sensor.
Fuel: new sockfilter, new fuel pump, new inline filter, new pulsation damper, injectors replaced With a used set that were cleaned and balanced, new fpr. Pressure tests at 37psi and 45 return line clamped.
Air: new filter, afm pulled and bench tested per fsm. Srews never taken out But pulled the black cover and checked the solders anyway. New marine fuel line to replace the accordion section Intake hose. Throttle body cleaned tps tested and swaped with known good one. Iac bench tested per fsm and cleaned.
Mechanical: valves adjusted And checked hot three times durring all this. New valve cover gasket pcv valve and bolt grommets. Compression test number are: 160 155 154 158 155 154. I did a leak down test at 100psi, and i dont know If i trust it because i got between 5 and 10% on all cylinders Which seems way to good for that many miles. And i did Think i could hear some leaking at the tail pipe and a small amout From the dipstick but nothing rushing out. Also wouldnt a bad valve be real evident on the vacuum gauge? And poor compression? I havent done anything to test the ecu but how often Does that go bad? The truck haD always been garaged and never In extreme conditions Before i got it. Hope some one can point out somthing ive missed, i hoped to not take it to a shop ever but I'll have to fire myself pretty soon if i cant lick this. Thanks for any input.

 

[toyteg08]

Was gonna guess burnt valve, but looks like you've covered that.

I had a situation like that once with my little Honda. Couldn't figure it out for the life of me. Turned out the intake cam gear was off by one tooth.

I know that's an OHV motor with a timing chain, but it's worth checking out.

All of your grounds are in good shape and correctly connected?

 

[jonheld]

Idle issues are almost always attributed to unmetered air entering the system. The fact that when RPM increases the problem goes away is telling me that.
The fact that the engine runs well under power tells me that all other systems are functional.
This may be attributed to a vacuum leak or a vacuum blockage or a vacuum line reversal. Pay close attention to the vacuum diagram under the hood.
Unmetered air can enter the top of the engine through the oil filler cap and dipstick tube, both of which should have gaskets. The side cover gasket can also be a place for leaks.

You mentioned replacing the PCV valve but not the PCV valve gasket. They become dry and brittle over time and will cause vacuum leaks.

How extensive was the desmog and was it running well before this was done?

 

[LS1FJ40]

I'm with Jon on this one. (Probably because Jon was the one to help me with my same issue).

I had a similar issue on my 92 a couple of years ago. I had a tiny hairline crack in my PCV hose on top of the valve cover. It wasn't noticeable. I replaced every vacuum hose under the hood, one by one, and would check to see if that solved it. When I pulled the PCV hose I didn't even notice it because the hose was rigid. Only after it stopped after changing that hose did I inspect it further.

Vacuum hose is cheap. I'd start there next.

 

[Dave 2000]

I have seen a sticking valve cause what you describe. At idle any cylinder sealing problem is most pronounced, when valve is slow to close it causes the misfire. As soon as you increase rpm the valve duration/overlap 'masks' the lag and the problem is less noticeable. This is why it appears to be a burnt valve and gives pretty much all the symptoms of it being as such.

This sticking is normally caused by years of carbon getting trapped on the valve stem, when you first start the engine the carbon prevents the valve closing fully. As the engine warms the carbon is slowly knocked back down the valve towards the valve head allowing the cylinder to seal up better and returning a tolerable idle but the valve is essentially wedging itself close, the action of the cam forces it to open but there is only the action of the valve spring to close it, ultimately when the build up gets really bad the valve will be so slow to close the piston come up and will give it some assistance, often with disastrous consequences.

You will find a leak down test rarely shows up the problem as the valve may or may not close in exactly the same place i.e. the valve rotates as it opens and closes so the carbon 'ring' is rarely in the same place with each test.

Now I am not saying that is your problem but, you do seem to have covered all the other possible causes, a vacuum leak would almost certainly shown up with the propane test.

Interesting this one my friend, keep us informed oh and BTW welcome (official MUD welcome)

regards

Dave

 

[Tinkercrincle]

Thanks for the response, sorry i knew there were some things i forgot to clarify. The pcv grommet and vac lines are new. I did not replace the side cover gasket, although i did find most of the bolts were pretty loose, and i though I'd found my problem.Tightened them all up, no change. As much as i'd like to find that unmetered air is my problem, I have been able to find it before with propane if it's a major leak or even a very small leak with my home made smoke tester. Also when i simulate a vacuum leak it shows up immediately on the vacuum gauge as a decrease. Can there be a vacuum leak that these three tests dont pinpoint? Also when i simulate a leak my idle goes down substantially Which hasnt been happening.
I desmoged the engine bay as part of my trouble shooting because i planned on doing it latter to make for a compressor and battery. It did not affect the problem one way or the other. I do know that there very well could be substantial carbon build up as i have used mopar combustion chamber cleaner as a piston top soak to solve a pinging problem. If i do have a sticking valve what can be done short of pulling the head for inspection? Thanks again!

 

[SUMMIT CRUISERS]

Just a quick thought. If there is a valve in question, I would try adjusting the valves with the engine running.

 

[Tinkercrincle]

I can understand that logic, i guess i did not think that could be done with the 3fe. When you take the oil cap off or even loosen it it it runs super rough, so how would it work without the valve cover? I have seen the video off a guy doing it on a 3f 62 so its the same exept you hook up the intake again after you pull the cover?

 

[SUMMIT CRUISERS]

Here's how it's done copied from one of my earlier posts.

Adjusting valves on a 3FE?

Park on a level surface (keeps the oil from running off the valve train/upper head area). Remove and then reinstall your air intake hoses after removing your valve cover. Plug the two vacuum hoses coming off the valve cover (1 is PCV and other is up front) . Crack loose all the adjustment lock nuts and retighten slightly. Start the engine and adjust away. I have noticed by doing it this way there is about 1/4 of a turn of adjustment on one end where the engine starts to miss and on the other end the valve starts to tap. I position it as loose as possible just before it taps. This is usually about in the middle of the adjustment. After doing all, I shut off the engine, re-torque the locknuts, and double check with the feeler gauge just to make sure. Adjustments are always spot on and the only ticking you will hear are your injectors.

A few tips: be extremely careful removing and reinstalling the intake hoses from the intake to the air filter housing. If original, they will crack and break. You will need to reinstall these to allow the engine to run without the valve cover. I also check for level (front to back) on the head surface. You don't want the oil running out the back lip. You can jack up the front or rear to level it out. After you start the engine, try to manually idle down the engine by pulling back the throttle lever to the driver's side. You want to do this at low idle that is why I mentioned to start this with a warm engine (AC off). Another tip, don't were a tie, loose clothing, and if you have a pony tail cut it off before doing the adjustments.

 

[Tinkercrincle]

That makes sense thankyou, i look forward to trying that out. I have read in another valve post that between wear and gunk buildup that tuning the valves by ear may be the only way to get it just right on a higher milage engine. I am going to explore the valves as a problem area until i am satisfied. Mostly because i could hear air escaping the tailpipe on at least two valves during the leakdown test, and it is my understanding that although some leakage past the rings is expected, exaust valves should not be leaking at all. I am borrowing a borscope in the next couple days, just a rigid one with a mirror, So i hope that may give me some feedback as to what is going on one way Or the other.

 

[SUMMIT CRUISERS]

I have actually seen where adjusting valves properly can be as close as an 1/8th of a turn on older motors. I experienced this on an older FJ40 with the 2F and the only way to get the valve operating quietly/smoothly was to adjust it with the engine running.

 

[Tinkercrincle]

So, i was able to get a look at the valves and seats from inside the cylinder with an inspection camera and saw no signs of burning ect. Couldnt get a real look at the stems, but there was remarkably little carbon on the piston tops. Then i did as suggested and adjusted the valves with the engine running to see if i could fine tune. Just ended up a little looser than spec with no improvment following the 1/4 turn from when it stops ticking guideline. They weren't making undue noise to behin with really. But i will say i will be doing future adjustments with the intake hooked up because if nothing else you can check your work and get instant feedback by starting it up before you button it up. Then i did a compressiom and leakdown again and got the same good numbers so i think i can rule out valve issues? I dont know, i'm stumped. Could my "rebuilt" injectors be bogus? I would think even if they werent they would have produced a different poor idle, but it stayed the same after the change. Could a tranny issue cause rough idle? Motor mounts seem ok. Afm is about the only thing i didnt swap out but it tests ok. Ecu? Also i pulled the plugs and they seem very white after 5k miles. Here is one, they look pretty much identical. Lean miss at idle? Propane wanded the engine bay again for good measure. Does anyone know an independent toyota mechanic that i could take this to in portland? Landcruiser northwest is probably out of my price range. Thanks!

 

[Dave 2000]

That is waaay to lean IMO. Glad the valves turned out good. Have you tried a power balance test, scoping the six cylinders and then shorting them one at a time?

Something else I would try would be to leave the engine idling for as long as is practical perhaps 30 minutes? Do not touch the throttle shut it down and pull the plugs keeping them all in order, see if there is any colour difference? If they are all as white as the one in your picture there is something a bit awry?

regards

Dave

 

[Tinkercrincle]

I have done a power balance test, pulling the plug wires one at a time and each one drops the idle about 100rpm. I can disconnect the coil and crank and all the plugs get wet also. Fuel pressure is good at the rail, and a spark tester shows strong spark on each plug. I have run the truck with the o2 sensors disconected to simulate a rich condition and it idles the same with the miss and runs a little worse overall, but not as much as i would suspect. I know air leaks cause a lean condition but i am still stuck on the fact that i have 20" of vacuum while it is missing at idle. A stuck pcv valve can simulate at vac leak but i have tested it and even run The truck with the pcv hose plugged and it acts the same. Anybody know What are the other causes of a lean running condition? I will do as suggested and let it idle and pull the plugs to compare. Thanks

 

[Tinkercrincle]

I Havent had a chance to work on the truck in the past few days however, i did unplug the o2 sensors and it is actually running smoother without them. Seems to confirm a lean condition? So my question is how lean does the truck need to be running before it throws a code 25?

 

[jonheld]

How old are the O2 sensors and are they OEM? I wouldn't rely on the ECU to tell you when they're done. 1986 technology. There are tests in the FSM for the O2 sensors or simply replace them if they are suspect.

 

[Tinkercrincle]

The 02 sensors are 1 month old oem. Changing them out didn't have any effect on my rough idle. I have not tested them, but i did have a cel for the old ones that cleared when i installed the new ones. Do you think the ecu could see them but be misreading them?My reason for suspecting i May have a lean mixture Is that my vacuum is solidly in the green at 20.5" but wanders in correspondence with the miss about .5" up and down when the truck is idling at 650rpm. Also my plugs look almost new after 5k miles. No deposits or scorching, but too white i think. Vacuum leaks cause a lean situation i know but the engine passes any test for them. It was suggested that carboned valve stems could be the culprit also, and i think i can look at them with an inspection camera if i look up the exaust manifold. I listed all the things i have checked at the top of the post, anybody see something i missed that can cause a miss at idle with good vacuum and no drop in rpm? Thanks again.

 

[Tinkercrincle]

I have read that vacuum in the green but bumping up and down slightly is a sign of idle mixture being off, that is what I base that guess off of. That and putting the system in rich open loop by unplugging the o2 sensors seems to make the truck run slightly smoother rather than bad.

 

[Tinkercrincle]

Update: Hey i kinda fixed it. I got one of the area cruiser mechanics on the phone and ran down the diagnosis i had done so far. He said "you don't want to pay me $95 an hour to check everything you've already done." He said it had to be something on my list so start from the top and find what i missed. When o got to the valves i decided to do set lash with the engine running, but this time leave the intake at .08 and do the exhaust by ear since that was a suspect area and i wanted to experiment. Turned the out about a 1/4 turn till they started clacking then back till it just stopped then a touch more. After just doing the #6 and #5 cylinder the engine smoothed noticeably with almost no miss. Shut off the engine and checked lash. .16 on both. I not sure exacly what this means, but it will not run smooth with the spec lash on those cylinders. It still has a slight bump here and there but jeeze i can live with that. I suspect Dave 2000 may be correct about the valve stem buildup and a head job would probably eliminate the rest of the miss. Anyway it is idling about as smooth as it was when i got it and better Off idle due to all the things i did along the way. When i have some vacation time i will pck up A rebuilt head and see what other trouble i can get into.

 

[NC LX]

Back from the dead I know.... But how’s the 3FE. I’m having the same issues you described on a 62 3FE and I’ve done everything you have done. I’m gonna pull the valve cover off and adjust valves running next.