Idle surge

Any vacuum leak at the head or before (IM, TB, IACV, PCV) can cause a surging idle. You mentioned a cracked "exhaust" manifold (header)... that would not cause a vacuum leak.

On cold start the motor uses a fuel table to idle.. coolant temps, air intake temps, (and a few other parameters) are involved. This is open loop. Once the temps reach a certain level, or you start driving, the system goes into closed loop and uses the primary o2 sensor and MAF (plus others) to adjust the fuel trim. But from what you've replaced and tested, I'd say its not the o2 or MAF sensors.

From Hondata's website (Honda tuning site)

Have you tried to bleed the coolant system of air bubbles? That can also cause a surging idle.

An old school way to check for vacuum leaks is to spray carb fluid (or something safer around sensors/gaskets) all around the areas you think might be the culprit. Not sure if you already tried that one.

 

man I have sprayed the carb cleaner everywhere, replaced vacuum hoses and cant find any vacuum leaks?

You say air bubbles in the coolant system can cause idle to lope?? How the heck does that happen. Can you please explain?

Thanks

 

Your 01+ exhaust manifold shouldn't be cracked.
That issue was on the 95.5-00 2.4 and 2.7.

 

^^^yes i can explain
the motor uses coolant temps to help control idle. if you have air bubbles in the system, it will fuck with your idle

 

Now that you have the ultra-gauge, is there anything listed under pending codes while it misbehaves? Perhaps something that's not there long enough to set a DTC and goes away after you shut off the truck.

Interesting that it's going open loop. Hmmmm. Ignoring sensors and hunting for idle in essentially limp mode without a check engine light is very weird to say the least.

 

No pending codes, but I think I have narrowed down that I got a bad replacement IACV from rock auto. Not sure if those are rebuilt or not. After hours and hours of research and dropping $300 bucks for a OBD2 reader and scope and I have checked everything. I downloaded a FSM and found 2 bench type test for the IACV and when I tried them both of the test failed. The IACV is partially open. When I start the truck it warms up and should send a signal to close the IACV. But this does not happen. It stays partially open causing the idle to increase. Then the ECM goes into a lean condition and cuts the injectors back to try to get the idle under control. The open loop I am getting is the short term fuel trim. It goes open when the injectors cut back the fuel.

I am ordering a new TOYOTA IACV today.

I hope I am correct!

 

Ok I have eliminated air in the coolant, IACV, coolant sensor, vacuum leaks, compression check/head gasket and cleaned the MAF but still no luck. I am still stuck on oxygen sensors. I replaced both from rock auto with denso units so I cant see those being bad. I have read so much crap from the fsm and haynes manuals. I am going crazy and about ready to sell this thing.

Does anyone else have anything else they can point me to?
Thanks in advance.

 

Forgot to add spark plugs are white, lean as I have ever seen.

 

Those little propane tanks are good for finding vacuum leaks, i would try the old o2 sensors back in the truck, maybe the cam postions sensor if it has one might be worth looking at, but i cant see it making the truck go into limp mode. But maybe if its going out giving it a rough idle.

 

Been there done that with Propane, I am 200% positive no vacuum leaks. I know the maf measures the mass of air in grams per second. I read in haynes manual at idle you should see 4-7 g/sec on your gauge. I am getting 10-14 g/s on my MAF. I then unbolted the MAF housing from the air box and set it up so I can see the sensor. I slowly covered the intake with my hand and the idle does come down and stops loping but the g/s go up to 60 on my ultra gauge. And then the truck runs rich and starts to smoke. Shouldnt the g/s come down as I manually reduce the air coming in? I am ordering a new MAF on Monday. I think I might be onto something.

 

I read that your coolant temp sensor tested "okay", have you replaced your T-stat? I've had issues with erratic idle on my '98 Taco with the V-6, no CEL, every sensor checkout okay just like yours. Replaced the T-stat, idle went back to normal.
Found that the T-stat was not fully closing allowing the engine to warm up properly, sending signals to the ECU to increase idle, and decrease idle within seconds. Good luck!

 

^^ the t-stat "not fully closed" will just take longer to warm up. But once it gets to operating temps that should not affect idle control.

TJF, plugs are white (lean).. not enough fuel. so what are the things that would limit fuel and at the same time cause a surging idle? hmmmm. give me a few secs to think. we should make a list and start with the "most likely" item and eliminate one by one

 

just thinking out loud.....
something is telling the ECU to run lean, but a lean condition by itself should not cause the idle to surge

 

Before we continue trouble shooting this pita shit can you clarify something or double check it?

You said this..

which is really weird.
On cold startup, the motor is in "open" loop mode which you have said the idle is perfect. Then once it warms up, the idle surges you said. Once warm, you should be in "closed" loop, also the same while driving normal - under low load. So something is not kosher here?

 

At this point I'd find another ECU and see if it helps any.

 

I will post a video. Let me go shoot it.

 

Thats kinda what i was thinking to.

 

Maybe try unplugging each sensor one at a time, if the idle stops surging on one its most likely the problem.

 

definitely could be the ECU or the wiring to/from the ECU or the main ground to the ECU. Good point btw.

 

Correct, however my old truck would idle fine at red lights. If the truck was idling for more then 5 minutes even after driving on the freeway, the erratic idle would start back up -not as bad, but it will present itself again.

I have done this before on older vehicles (non-toyota). Pack ice around the ECU, when the problem is present. If the problem goes away, ECU is most likely the culprit. To be sure, you can remove the ECU itself, take it down and get it tested.