3.2 IAC seems fritzy after rebuild, high idle, tapping helps, what next?

[r-mm]

OOPS I posted this as a reply in another thread by accident...

I rebuilt my 3.2 @ 106k (not long ago) but didn't touch the IAC as the car idled like a dream. After the rebuild everything felt the same except the idle is now slow to come down from about 1900 and sometimes settles around 1250 instead of ~880. Behaves the same if the car is hot/cold. Tapping the IAC a few times, decently hard, with the back of a screwdriver does tend to settle the idle back down below 900 where the lord intended it to be.

So I just yanked the IAC valve. This thing is basically a big stepper motor with a little throttle at the end, right? So the only thing to clean is the rotation of the little flapper that controls airflow?

After cleaning, I started the car. Idle bounced up and down until I reved it up some after which it settled to a very nice ~900rpm idle. Let it warm up, turned it off and got the same. Thinking I'll just drive it for a while and see if the bouncing gets worse or better.

[ischmitz]

Could you have a vacuum leak in the intake tract (runners to heads)?
Is the idle micro switch engaging properly (throttle link adjustment)?

Ingo

[Mr Project]

I've got a similar situation - RPMs bouncing at idle intermittently after a rebuild. Doesn't always do it, but has persisted after a cleaning of the IACV.

There are a number of good reasons that it could be a vacuum/intake leak, though in my case (and maybe yours) I'd like to think that it's not - I just had all that stuff apart and it's been sealed up with new gaskets. Mine is also intermittent and not engine-temperature-related, and I'd like to think that a vacuum leak would be fairly consistent.

Next on my list is setting the idle mixture and speed, since I haven't done that yet - have you already done that step since the rebuild?

[r-mm]

Frustrating because it flys in the face of my instinct to not be afraid of sleeping dogs and to try and improve everything I have access to when the engine is out. I too would like to believe its not a vacuum leak, but as was the case with the "pipeline" that many of us see cracked (I repaired mine with self fusing silicone tape) there may be other rubber components that are frail and escaped notice.

However since the symptoms I'm getting changed after I cleaned the IAC, I'm pretty sure the solution lies with the IAC.

Does anyone have a description of how it works? Maybe this is in Bentley? Mine's in the sub zero garage, should pull it out and check.

[Hunt3R]

Does it keep bouncing when you're driving around and push in the clutch or does it only bounce at startup? The idle should never get up to 1900, even 1250 should only be seen for a few seconds on a cold start on a cold day. After doing a rebuild you've got to reset the base A/F mixture and then reset the base idle to 880 after the A/F mixture is good. Assuming your ICV is clean and functioning, it can still only adjust up to a certain amount around the base settings.

[techman1]

Ok, you asked for a description! But for the more technical "Why?", Ingo is the man!

My observations were: 3 devices work together to control your idle in a 3.2 motronic car. The throttle closed switch, idle valve (also called ISV, IAC..), and the DME brain.
The throttle closed switch tells dme that the throttle is not open, so the DME is expecting a idle within its internal chart, and it will adjust the signal it is sending to the idle valve to regulate the idle.

If the idle gets too high the DME will send a signal to the idle valve to close. For example, if the idle valve was stuck open. If the idle gets above a certain point, about 1200, the DME cuts signal (don't know what it actually cuts, maybe fuel injector pulse?) , idle drops, dme resumes, idle goes too high...repeat. My bet is this is usually a bad ICV, bad ICV driver on the dme, or idle screws played with/damaged by dpo.

Another scenario, big intake leak, idle gets too high. The DME sends a close signal to the idle valve. Valve closed, idle still over 1200. DME will go through its on/off cycle, resulting in a fluctuating idle. The idle valve needs to be in the middle of its rangeat normal operation.

The idle needs to be adjusted, with the idle valve disabled, so the idle is at the specs for your year. Then, when the idle valve is in use, it will have an open and closed range to correctly adjust your idle.

Is the closed throttle switch operable? With a lot of light in the area, (careful of loose clothing in the fan!) raise your throttle above 2000. Then reach in and press the throttle closed switch(Thats why you need a lot of light, hard to see that little bugger!). Your idle should begin fluctuating. Make sure your throttle is adjusted so that when you are off the throttle, the switch is engaged, otherwise the idle circuit is never used!

[r-mm]

Thanks this is very useful.

The switch audibly "clicks" every time I get off throttle. I've checked this many times. I will try the test you mention to make sure it is actually operating the circuit.

Since after cleaning the IAC I can get the car to idle very nicely at 880 just like it used to, but I sometimes get hunting, I'm really tempted to see this as an IAC issue, but the vacuum leak theory is appreciated as well. Maybe in removing and replacing the IAC I stressed some unseen but creaky/brittle part of the intake tract? I replaced all the intake runner gaskets during my rebuild and gave everything a visual inspection, but did not replace the elbow.

[ischmitz]

Here is how I think about the idle on a DME engine:

The throttle plate gap sets how much air is able to enter into the engine at idle. There is a hard stop adjustable with a set screw. The ICV is a variable bypass over the throttle plate controlled by the DME. The DME adjusts the size of that bypass via PWM. Changing the ICV value is equivalent to ever so slightly stepping on the accelerator (without opening the idle micro switch). It increases metered air into the engine.

The AFM flapper box has another adjustment. It is a little bypass channel. The total amount of air going through the AFM is the sum of air deflecting the barn door (metered air) and the amount of air going through the bypass (unmetered air). As you increase the bypass the barn door deflection at idle reduces and vice versa. When you open the bypass it causes the mixture at idle to lean out. Setting the mixture at idle is important so the O2 sensor regulation is close to its center.

The hard stop adjustment on the throttle body sets the center position of the ICV. In other words if the throttle body is too far shut due to wrong set screw setting the ICV is maybe at 90% open maintaining 880 RPM and has little margin to open further to catch an idle-drop. If the hard stop is misadjusted either way the ICV will have a harder time to influence the idle properly. The feedback loop gains programmed into the DME assume a certain effect that is lessened and thus idle hunting is the result.

The DME has several mechanisms to impact idle:

- ICV position: Controlled by a feedback loop looking at the RPM, invoked when the micro switch is closed.
- ignition advance: The ignition map has more advance at sub-idle RPMs to help push the RPM back to higher numbers. This happens at all times regardless of micro-switch position.
- Mixture: is controlled by the O2 sensor and tries to keep it stoichiometric.

If you did not touch the AFM adjustment I would start by checking the throttle hard stop. And make sure the micro-switch engages solidly when the pedal is released. The throttle plate needs to come to rest against the hard stop (not being constrained by the throttle linkage). When you open the throttle first the idle switch disengaged and then the throttle plate moves off its hard stop.


Good luck,
Ingo