Ok TEAM we have a new one here. My daughters 99 Cherokee (I6 Cylinder, 180000) has a very specific problem. I have read all the post related to similar and made several changes but haven't resolved the problem. My background is Mech Engr so I have analyzed this until my brain hurts. Two shops haven't figured it out so I thought I would come to you guys. At nearly exactly 90 secs after a smooth start the RPMs start revving up and slowing down like it's almost a stall and then do this back and forth for 60 seconds. IF I let it idle for 3 min it will run fine! ( Which is how I let her drive it now). However if I drive it while the RPMSs are up and down it will buck like crazy. A better description is it's as if the throttle body is closing completely off and then and opening wide open back and forth. STILL at the 3 min mark, runs smooth and great from then on. I know it's not the time. It's the temp of either air or coolant. SO I have done the following;
Replace IACV ( and cleaned port on throttle body)
Replace Coolant Sensor
Replace Intake Air Temp sensor
Replace MAP
Replace TPS
Full tuneup
Everything else on the car works great so it's a Keeper

 

were any codes thrown on it?

 

Pcm?

 

MY 2000 does the same thing, I cant figure it out either.

 

this is my thought also

 

It's hunting... After cat o2?

 

Could be the pcm. Sounds like you have a problem while in open loop. Once the engine warms up and goes into closed loop it gets its air fuel ratios from the upstream o2 sensor.

 

After cat 02 sensor doesnt have anything to do with running the motor. Its only there to monitor the emissions comming out of the cat. before cat o2 sensor sends the signals to control your fuel ratios after warm up.

 

It's an output...the ecu compares the two voltages...is it there for just a cel?

 

check pcv valve and all vaccum lines had a couple cars at the shop do that and it has always been a vaccum problem but have heard of pcms doing this good luck

 

it was put there to make it easier to moniter emissions. Instead of having to use the tail pipe sniffer to read emissions the car is doing it it's self. It will tell the dmv inspectors if the emissions after the cat are up to government standards. the 02 sensor after the cat will not effect how the engine runs but will set off your check engine light to make you fail inspection.

 

Hmm... Based on your description, some initial thoughts:

Since it starts smooth, goes rough, then smooths out again, the problem seems to be happening at the transition from open loop to closed loop. If I'm not mistaken, AIC, crank and cam position sensors are applicable in both modes, so it likely wont be one of those. And you've replaced the AIC, so I doubt the issue lies there. My guess would be a conflicting reading from a sensor that causes the cycling, and when it can't get things right it may be going back into open loop mode after a few minutes. O2 sensor issue seems most likely, outside of a PCM that is going south. The only other thing I can think of (but not likely) would be some kind of issue with the fuel/injector operation once it goes into closed loop mode, and the PCM is taking a moment to adjust pulse width to compensate. But again, I can't think of a plausible scenario to support that.

Notice any heavy fuel smell from the exhaust? Has fuel consumption changed notably? Does this situation happen after a full warm up followed by a shutdown lasting no more than 20 minutes? No codes thrown?

 

I have no gas smell, and no codes. A restart in under ten minutes runs fine, not sure about twenty, but thirty or more and the problem returns.

 

TEAM! YOU"VE DONE GREAT with your responses. Thanks so much
I am going to respond in order of your suggestions
Jeep Junky- No codes thrown
BAMBAM017- THis makes the most sense. It has such a defined beginning and end which would explain why it fits very specific times. The word transition makes perfect sense.

I don't believe it's O2. I have had these go bad in other cars and they just run crappy. This morning I got out and measured the exact time ( and temp) things changed. It idled smooth for 88 secs. About the time the temp gauge is starting to move. The oscillating of RPMS started. When the slows down it almost stalls and then 'catches itself' and revs up, did this for exactly 90 sec and then smoothed out.

Not vaccuum, too specific

MacGyver35 Your comments along with BamBam make the most sense. It is a transition problem. The reason I hesitate on the PCM is because of how well it runs the rest of the time. No excess fuel smell, great gas mileage, runs strong after the 3 min mark, and never does it again if warm.

Hope my additional info helps us solve this.

 

Reposting