IAC tuning. Am I crazy? (Solved!)

[pfabricy]

IAC is killing me too. I so wish I had a carburetor again. I overcome this though. I've got mine figured out. Here's what I did.

I had to increase the number to that IAC goes to when I start the engine. This can be found at settings, idle control, Start Value field. I increased mine so it would close completely. Now I know I have a good reference point and know it's closed. The 150 setting in that field gave unreliable RPM at startup. I went to 175 and that made my idle the same every time I started the engine. This force the valve against the seat in a visual sense.

Then I set my Cranking Position number to 100. At this point I knew that it's going the right direction because the engine started. If the engine didn't start it was wired backwards. If backwards it would have tried to close the valve more therefore not starting.

Once running I watch the temperature and RPM. I went to the "Idle Steps" under table menu. This is called the Idle Step Bins. I can immediately adjust the engines RPM at this point. Set the numbers on the right side fore the temperature that your engine is at that point in time. In a second or two you will notice an RPM change. It's pretty cool!

Once I got it warmed up and the idle set right I noticed black smoke and that I was running rich. I leaned it out by changing my injector size(not the way I should do it). Then the engine idled higher so I had to adjust the "bins" again to get the correct RPM. That's where I'm at now. I think my next step is to play with the PWM settings to get it running correctly.

[bluetrepidation]

1) Is everyone with issues using a GM IAC, either the inline connector or the square one? Or are these other stepper types? If so which ones?

2) jumpering out the big resistors, does this have an effect?

1) I am running an inline o ring style GM IAC.

2) Jumpering the big resistors has no effect.

I also tried changing out my RC components like JSC originally tried. I completely replaced my stepper chip as well. No change. Phase one seems to be half stepping on me. I don't think my file attachments still exist but you can look at some scope readings off my server here:

http://70.226.38.10/~admin/IAC_Pics/Free_Spining.gif
http://70.226.38.10/~admin/IAC_Pics/Jammed.gif
http://70.226.38.10/~admin/IAC_Pics/No_Motor.gif
http://70.226.38.10/~admin/IAC_Pics/NoJ ... hanges.gif
http://70.226.38.10/~admin/IAC_Pics/Phase_Inputs.gif
http://70.226.38.10/~admin/IAC_Pics/Stop_250setp.gif
http://70.226.38.10/~admin/IAC_Pics/Stop_252step.gif

Now I lifted phase one and connected it to the phase two output on the processor. This caused both phases to strobe correctly but of course they where in phase with one another. This is as far as I have gotten. My last effort wrecked my MS II card so a new one is on the way. I'll leave it up to the pros for now.....

A.J.

[Ballistic]

I've been watching this thread closely, and I'll just chime in again for Bruce and Al.

1. I have brand new inline GM IAC motor.

2. Resistors jumpered, no real change.

After warmup I need to cycle the ignition to bring the idle down.

[Bruce Bowling]

I've been watching this thread closely, and I'll just chime in again for Bruce and Al.

1. I have brand new inline GM IAC motor.

2. Resistors jumpered, no real change.

After warmup I need to cycle the ignition to bring the idle down.

From the description, it appears that the SM is losing steps somewhere. This can be a driver issue, a issue where the code does not issue all of the steps during warmup, or other like stepper motor acceleration/deceleration profile or step frequency.

One thing is the stepper motor itself, if unscrewed, the pintle should be strong and resistant to your finger while moving. One can see th difference to a winpy drive vs. a strong drive by simply unhooling one of the leads so there is only one coil active. The motor will not move anywhere and it will lack power, easily masked by a finger. Full power and it takes more effort to stop the pintle movement.

If someone could screw in a spare IAC into the port jus to maintain constant idle, and then plug in a second IAC and run the car thru warmup while holding their finger over the pintle, each step should give a good movement pulse.

Also note that the code has a 4 second pause between commanded movements, Al put this in to keep the pulse count caught up with fast-moving stim coolant movements. So, the coolant will be changed, and the motor can take up to 4 seconds before actually moving.

- Bruce

[Joethemechanic]

One thing is the stepper motor itself, if unscrewed, the pintle should be strong and resistant to your finger while moving. One can see th difference to a winpy drive vs. a strong drive by simply unhooling one of the leads so there is only one coil active. The motor will not move anywhere and it will lack power, easily masked by a finger. Full power and it takes more effort to stop the pintle movement

Bruce, while faulty stepper motors can be hard to diagnose, several of the MSer's having these problems went out and bought brand new units, including me.

As far as I can decern just about everyone is having the same issue, the stepper motor does not fuly extend during warm-up, an on/off cycle is required to seat the IAC.

On a side note, I have been fighting an intermittant reset, most of the time shortly after start-up, where the engine will die, if MT is hooked uonit losses coms, and usuall the Accel and tach LEDs remain lit, and the fuel pump remains on.
I bring it up cause on the few datalogs I have of it, it seems to happen just as MS commands a stepper motor movement.
Don't know for sure as its hard to tell exactly what is going on, but is it possible that the IAC driver is overdrawning the HC12?
http://www.msefi.com/viewtopic.php?t=18355&highlight=
Joe, MSII V3 2.35 ( working in 2.66 ) GM HEI TBI MSD V8

[bluetrepidation]

Bruce,

I got a chance to play with the stepper as you suggested. Keep in mind I tore a solder pad on my MS II card and had to rig in a jumper to get it back into working order. I will not have definitive results till my new processor arrives.

My test motor has the valve removed to I can see the motor movements or I can stick a screwdriver in to feel resistance.

Things I noticed where:

1) The resistance the motor had to physical movement against a stopped state varied between heavy resistance to light depending on what step the motor landed on.

2) You could sometimes on and off feel "noise." Like when you place you finger on a speaker that is playing loud music.

3) You can observe the motor "click" sometimes resulting in a forward movement but sometimes it fell back from where it started.

I'll rerun these steps to see if anything changes with a fresh card.

One thing I want to do is scope the new card before connecting it to my motor just to see if I'm getting a good output though.

I'll let you know,
A.J.

[Bruce Bowling]

My test motor has the valve removed to I can see the motor movements or I can stick a screwdriver in to feel resistance.

Seeing the motor movements while the vehicle operates is very helpful. This gives a visual to the operation under the slow warmup curve.

1) The resistance the motor had to physical movement against a stopped state varied between heavy resistance to light depending on what step the motor landed on.

Depending on where the pintle is in rotation you may be able to push it back a bit. There is a portion of the thread where the pintle can be pushed backwards, I think this is to prevent damage to the IAC if it hits a solid stop, it just slips.

2) You could sometimes on and off feel "noise." Like when you place you finger on a speaker that is playing loud music.

Do you have the power resistors jumpered out on the board?

3) You can observe the motor "click" sometimes resulting in a forward movement but sometimes it fell back from where it started.

This is the slip-back I was talking about above.

If you can rerun this test try the mode of "IAC Moving Only" mode. This turns off the holding current between movements. Compare to the "Always On" mode to see if there is any difference.

Thanks for the testing on this!

- Bruce

[grippo]

I have looked at the code and tested and tested and can not find any missing steps in the software nor in my (GM square) stepper motor. But I will make a new software version for testing in the next day or two and ask people to try that. It will have algorithm changes and/or diagnostics to help us track down this problem. However I would like us all to be on the same page right now as far as inputs are concerned.

1. Everyone should have the 1 ohm resistors jumpered

2. Everyone should use 15 Min after start IAC rather than moving only.

3. Everyone should have their motors wired so that when cold the valve is all the way open, and it closes as the engine warms.

4. Everyone should basically use the default IAC values except as needed to improve things. For example, as pfabricy said, you should try increasing the Start step (and correspondingly the last step position in the the step position vs coolant table). This allows some slop if missed steps do occur because you can use more than the required steps so it will always close tight. You can also try changing the time step to see if this helps. If it doesn't set it back to default.

One other thing that may be causing confusion is the time based after start. What this means is that you pick a point on your coolant vs step position curve where you want to go into this mode. Say your last two points are 140 deg, 120 steps and 160 deg, 160 steps. If you set a cold psoition of 130 then a little after120 deg the steps will no longer track with the coolant - they will go up automatically with time until they reach the last step position of 160 at a time of Cold Taper secs after reaching 120 deg. Very few people need this - I believe only those using a heavy weight oil which takes much longer to warm up than coolant. It only comes into play if the starting temperature when you first turn on the ECU is < the Cold Temp you specify (default is 0 so it is unlikely to be activated at this time of year, at least not in N hemisphere.

If we can't solve the problem in software, Bruce has made a circuit that people can try in place of the Allegro stepper chip just to see if the problem lies there.

[midnightblue]

Hi Guys,

I have done some tests on this.

They were conducted using MS-II (with resistors jumpered) on a V3.0 board using a GM 4-wire (square plug) IAC. All the testing was done using the on the bench using a stim and freshly charged 12V battery. The pintle on the two motors tested could not be turned as it runs in a grooved channel.

The CLT was increased slowly to cause single step increases in the IAC value, measuring the total length of the valve and pintle. Each algorithm was tested using a 5.0ms step size. The stepper became very hot during the ”Always On” tests



In all cases the pintle seemed to move, even where no movement was measured it twitched and clicked. It seemed to be moving half a step forward and the fall back to the original position, although I had no way to measure this. Possibly with multiple steps, the motor armature has enough momentum to complete each step?

I hope this helps, Matt.

[grippo]

Matt,

This is really useful data and it points to some solutions for the problem. I am curious about two things:

The 15 min IAC should be identical to Always On for the first 15 min. So I would like to know if you cycled power on the processor each time you did a test. If not, could you repeat the 15 min IAC test cycling power before you test. If there is a difference it would mean there is a software error in the 15 min IAC mode.

Could you repeat the Moving Only test with a step time of 200 (20 ms) and see if it makes a difference. (The idea here would be to allow more time for a stopped shaft to accelerate to the next position. Once it was moving then the time could be shortened - that is I could implement the acceleration inputs which currently do nothing and shouldn't really be in MT.)

Thanks,

-Al