[keithsketchley]

Does anyone know how to retract the IACM on a 1994 3.3L engine?

It is a stepper motor with two windings which IIRC is typical of stepper motors, but Chrysler manuals do not detail other than four "driver" connections to the ECM.

Mine protrudes 1 1/8" whereas the factory service manual says should not be more than 1", their fix is to use their scan tool to command retraction. Cannot just turn the pintle to retract.

The longer length fits in to my throttle body but I am having intermittent run-rich loss of power, stalling, and failure to start. (I cleaned it and put all back together, started fine a few times though code 52 set (O2 sensing mixture too rich, previously had code 13, I reset Adaptive Learn values), drove around the block OK, then after sitting for a couple of hours would not start - engine ignites but won't run when starter released. Experience before some cleaning was operation troubled then suddenly OK.

(The IACM ports air past the throttle plate to control mixture and idle speed in some bizarre fashion.)

I will take the entire throttle body out and clean more thoroughly.

 

[rmoore0949]

The IAC motor controls idle speed only by bypassing air around he closed throttle plate. If you are setting a mixture code, you have another issue. You can push the pintle in to retract- but dont screw it. If the engine will start and run while holding the throttle, the computer may have lost control of the motor or the motor is bad. If you have access to a scan tool that can set idle speed, set speed to 1500RPM and see how long it takes to respond and close when deactivated

 

[KOG]

As long as the thing goes together without jamming (and it will) protrusion is a moot point. It constantly adjusts to hold idle speed. If you have variable idle speed problems due to a sticking AIS motor cleaning can help. Running rich is unrelated to AIS and cleaning it won't help.

 

[keithsketchley]

OK, the wiggling while pushing approach suggested in Allpar does work to retract the pintle.

From various threads and articles I understand this is probably a simple stepper motor as such go, one-quarter turn at a time, bi-directional. Should be able to turn it by energizing one winding, whichever retracts it. Unsure if 12v is too much, ECM outputs often 5v. I have all the relevant Chrysler manuals except training ones, which are scarce.

Re-iterating, the issue is retracting the pintle per Chrysler's instructions. (Not that their manuals are perfect - they seem to lack feedback from people who are not familiar with the designs but have tried to use the manual - even a writer who has experience with the design misses information that would be helpful to users.)

I don't have a high degree of confidence that the ECM always gets things right in unusual circumstances. There is of course the ability to clear the Adaptive Learn data to start afresh.

BTW, on my vehicle the passage from the IACM to the air inlet throat is quite large, so plugging of it is unlikely.

 

[ImperialCrown]

I believe that by rolling the key to 'run' will retract the pintle for an idle speed 'flare' at start up. Have a helper roll the key to on and then unplug the IAC. This should leave it in the retracted position.

 

[keithsketchley]

Thanks "Imperial Crown".

Otherwise, I say Hold on here people, if the IACM system is not allowing air in when it should won’t the engine will run too rich at idle, since the TPS is at lowest value as throttle is not being opened? (From its name: “idle”. I don’t know if it functions off-idle at all.)

The O2 sensor is key to much operation of the engine, information available to me indicates it is essential to operation of the IACM system. (The IACM is used to control idle speed as well, so engine does not stall when idling.) However, since the O2 sensor must be hot to work it I doubt it would contribute to a fail to start problem (this vehicle has electrically heated sensor so will get hot sooner). Note however that the ECM controls open loop until the engine is warm, only then does it use the O2 sensor (I haven’t yet understood how the – more reading of tests in the diagnostic manual, temp sensors are usually low resistance when cold thus if the coolant temp sensors goes open the engine will never go into closed loop).

Keep in mind this design calculates fuel based on throttle opening and other parameters such as air pressure and air temperature to determine air density, then adjusts it with O2 feedback (and stores some of the adjustments it found necessary).

The basic design is calculation, different from designs that use a mass airflow sensor which measures the actual amount of air directly.

Many variations even at that, for example this vehicle does not have a manifold air temp sensor, nor an air inlet temp in the filter box – presumably uses the OAT sensor in the grill area as this vehicle has the optional overhead console with OAT, compass, and trip/mileage computer. I don’t know if the ECM compensates for the air becoming warmer as it passes through the computer into the filter box, nor how the manifold air temperature is used in vehicles so equipped.

 

[keithsketchley]

Further complicating troubleshooting, without detail information on sensor design and without full scan tool, is that modern sensors may not be simple potentiometers, but rather tiny chips with an amplifier circuit. Theoretically more reliable, as potentiometers wear out (first at positions most often used, such as 55mph in the flatlands, so can be intermittent), but not testable with simple equipment).​

I will create threads for troubleshooting a couple of aspects, and for knowledge of functions, with appropriate titles and linking to this thread for details. I started a thread http://www.allpar.com/forums/topic/146614-egr-valve-could-case-startidle-problems/ to provide information on function and configurations of EGR valve control and back-pressure sensing.​

 

[AC TC]

The IAC isnt the problem.
The computer calculates fuel based on temperatures, vac (map) and speed (rpms)
and later when in cloosed lopp takes the lambda value into the calculation.
The computer doesent have a clue wheter its rich or lean until it goes into cloosed loop.
Id check the fuel pressure, lambda and map sensor first.
Have you read the plugs? ( so you dont have a leaky injector)

 

[keithsketchley]

Thanks AC TC.

Would a single leaking injector cause no-start? Front three plugs are the same, now significant fouling as expected due misfires.

Keep in mind O2 is not used as you say, until sensor warmed up.

Further re code 52 occurring during troubleshooting, it was a transient detection. One thing to keep in mind is how the ECM keeps codes for if the condition is not recurring. Code 12 for direct battery input lost is by fault definition retained for 50 start cycles. Good practice to clear codes during troubleshooting, advice in this forum is to clear the adaptive learn after mucking with sensors, which seems to clear codes as well.

Note that with misfiring the exhaust will be rich, but the ECM should not use the O2 sensor during starting and initial warmup as it will not be accurate until hot (it does have electric heating on this vehicle so may not take long to get hot once the engine is running).

Have to clean spark plugs after all the misfires, a PITA on this vehicle but I may have found a different way than raising the vehicle as I and others have done in the past (take off pieces, reroute wires, etc).

 

[ImperialCrown]

If an injector leaks down after shutting off the engine, it can cause a hard, rich restart. On a warm engine the fuel can vaporize and affect all 6 cylinders on a restart (multiple cylinder misfire). By holding the gas pedal down helps to clear a rich, flooded condition with added air and ease starting.
The IAC is basically for idle and lower engine speed regulation. It also opens on deceleration for emissions control and stall prevention, like an electronic dashpot.
The correct spark plug should be RN14PMP5. Other brands may not match the correct heat range and carbon foul easily. Aftermarket O2 sensors may also cause issues. It is always best to use OEM parts.
A fault code check by counting the 'ck eng' light flashes will always start with 12 and end with 55.

 

[keithsketchley]

Thanks Imperial Crown.

I have normal fuel pressure soon after turning key and it does not sag when engine starts to run then dies.
Fuel pressure does drop over several hours, I don't know if it leaks past a check valve I presume is in the pump or (will check my notes on the pump, don't think I still have the old that failed a few years ago) or an injector.

I'll try WOT again as a test, did not fire at all when I tried it..

Odd that code readout would start with 12 as that is a fault, 55 is obvious.

OK, plugs replaced with new ones, fouling even except aft ones perhaps a bit more

Throttle held open overnight with air cleaner cover off to let some fuel vapour escape.

Only codes 12 and 55. (After re-re TB, TPS, and EGR pieces I cleared adaptive learn which cleared various codes. Code 12 by definition must persist for 50 start cycles, 55 indicates end of code readout.) Check Engine light not illuminated. (The ECM does not set codes for everything, and cannot detect everything – not fuel pressure for example. The service manual provides some guidance. Some codes are only set when running – MAP for example only when an approximate idle speed is reached, though it uses data much earlier as it needs to know altitude (ambient px) then the lower px from cranking causing vacuum. Some codes are not set at high rpm.)

The engine will fire and accelerate while cranking but not run on its own. (The problem progressed from intermittent to always. Along the way I had cases where it would idle but not take load, others where it ran better at significant RPM.) No firing with WOT during start (on a mid-80s GM system fuel is cut off during WOT start attempt, to facilitate clearing of excessive fuel).

Thus the fault is logically with something that is not used during cranking but is once the engine reaches a speed that it should be able to run from.
Or several problems/marginal conditions. I suspect the ECM's coolant temp sensor is bad, TPS may be marginal.

I am slogging through checking sensors and wiring including long-shots.

 

[keithsketchley]

I started a thread on ECM sequencing of use of cam and crank position sensor information: http://www.allpar.com/forums/topic/146849-question-re-use-of-crank-sensor-by-computer-during-start-1994-grand-caravan-33l/

I intend to start a general thread summarizing what I learned in general and these two threads, and threads on specific things as appropriate (such as my thread on EGR valve control).