IAC tuning. Am I crazy? (Solved!)

69 1/2 Six Pack Bee · Sun Apr 16, 2006 10:02 am

Hey, JSC
Thats encouraging news, bro!

I am ready to make the change as soon as you let us know what you did and what you used.
More than likely we can use plain'ol components as opposed to the super-small caps and resistors as I am sure they will fit.

Thanx again for the help. Maybe NOW I (we) can put this pesky stupid IAC issue behind us now and move on to the rest of the tuning!

JSC · Post by **JSC** » Mon Apr 17, 2006 6:21 am

Three cold starts now and the IAC has worked as advertised each time. As stated above I changed C14 & C16 to 820pf and R7 & R8 to 56.0K ohm on the MSII daughter board. NOTE: with this component change you do not want jumpers on R5 & R6. My IAC step size is 5.0 ms, haven't tried it with a lower value yet, but I believe it will work down to 2.5 ms.
The components I used from Digikey are, PCC821CGCT (CAP 820PF 50V CERM CHIP 0805 SMD) $0.076 ea. and RHM56.0KCCT (RES 56.0K OHM 1/8W 1% 0805 SMD) $0.038 ea. Bud suggested using standard components rather than SMT, I think they should work and should be easier than dealing with those tiny little SMT parts. If you try it let us know how it works.

Scott

Philip Lochner · Post by **Philip Lochner** » Tue Apr 18, 2006 12:41 am

JSC wrote:Conclusion, the 2916 is going in to some sort of fault mode.

Question for the Gurus: The component values for the oscillator RC network on the 2916 are out of the range specified in the Allegro 2916 datasheet. Could this be an issue?

Scott

Hey Guys

Perhaps there is something else we should consider.

Under other topics, it has recently emerged that the ONLY truly useable Ignition timing predictor algorithm is "Last interval". "1st Derivative" also seems to work but on my car even that goes bad when I activate the AC when idling. So what does this have to do with IAC? Read on...

I've been using "15-minute IAC" on my car which seems to work to fine except for the "time based afterstart" bit which I can't get to work. Last night I put it on "Always on" and the IAC did not do AT ALL what it was supposed to do. My point is that perhaps we should not ASSUME that all the IAC algorithms work as advertised. On the bench I had IAC algorithm on "moving only" but that also created behaviour that could not be explained ito the "Idle control" and "IAC steps" settings. (This was WITH tach pulse input to simulate start and running conditions)

My point is that perhaps some IAC algorithms work better than others.

Personally I am convinced this is the case.

bluetrepidation · Post by **bluetrepidation** » Wed Apr 19, 2006 2:37 pm

Now that I'm back from work I can tinker with my car again. I'd be willing to mod my MS II card to verify what you have discovered. One thing though. Seems Digi-Key is out of the 2916 stepper chips. Is this where you ordered your new chip from? Know of another supplier? Once I can get my hands on what I need I'll try this setup out. Let me know.

A.J.

JSC wrote:Three cold starts now and the IAC has worked as advertised each time. As stated above I changed C14 & C16 to 820pf and R7 & R8 to 56.0K ohm on the MSII daughter board. NOTE: with this component change you do not want jumpers on R5 & R6. My IAC step size is 5.0 ms, haven't tried it with a lower value yet, but I believe it will work down to 2.5 ms.
The components I used from Digikey are, PCC821CGCT (CAP 820PF 50V CERM CHIP 0805 SMD) $0.076 ea. and RHM56.0KCCT (RES 56.0K OHM 1/8W 1% 0805 SMD) $0.038 ea. Bud suggested using standard components rather than SMT, I think they should work and should be easier than dealing with those tiny little SMT parts. If you try it let us know how it works.

Scott

bluetrepidation · Post by **bluetrepidation** » Wed Apr 19, 2006 2:40 pm

I also messaged another Saturn owner with a turbo MS II setup and this was his reply to getting around the IAC issues he had too. Kinda interesting

S.Bretz:

"I couldn't get it to work for more then a few minutes before it "just stoppped working" for some reason. at 18-22psi, the boost messes with the postion and the MS doesn't know where the pintle is anyway, so I unplugged it and sealed the IAC passage off. I set the configurations so that is would idle warmed up, then I let it cool down and messed with the cold advance. The advance makes it idle faster, but wilth the engine cold, it runs just right. Then just pull the timming back to 0 at 160 degr."

midnightblue · Post by **midnightblue** » Thu Apr 20, 2006 3:23 am

I've also been having issues with unpredictable IAC responses.

I bench tested two GM square plug IAC steppers yesterday using the stim and a freshly charged 12V battery. I have the resistors jumpered on the MS-II board and v2.35 of the code.

I measured the total length from the back of the plug to the end of the pintle using vernier callipers. Large changes in the Idle Steps value were reflected by obvious movements in the pintle. When increasing the CLT level slowly, MS was increasing the Idle Steps one at a time.

The stepper seemed to react with a single step change, it clicked like it was moving. Holding the motor body, it felt like something moved. The vernier showed that there was no movement at all.

Further testing showed that even incrementing many steps produced no movement in the pintle if they were done one at a time.

I tried pulse lengths from 2.0ms to 5.0ms, all with the same results. I was using "Moving Only". I will test "Always On" and "15 minute IAC" when I have some more time.

I don't want to try changing SMT components on the MS-II unless I really have to. Lots of people seem to have similar issues, I guess they may all have the same root cause. Does anyone have a proven solution?

Many thanks, Matt.

JSC · Post by **JSC** » Thu Apr 20, 2006 8:15 am

A.J.

I didn't change the driver on my MSII, I did buy a 2916 DIP for my test setup, I got it from http://www.newark.com/. My IAC still seems to be working fine.

I think that there's an issue with the program and the way it pulses the 2916 motor driver. The indication of this is that on my test setup I can't recreate the issues we see with the MSII, but I'm driving the 2916 with a L297 Stepper Motor Controller. Lance indicated that his IAC works properly, I'm wondering if it has something to do with selected options (ie, dual MAP for realtime baro). Maybe Lance could give us the details on his installation for comparison. Maybe one of the software gurus could take a look at the program (I haven't done any C in 20+ years and don't have time for the relearning curve right now).

Scott

Bernard Fife · Post by **Bernard Fife** » Thu Apr 20, 2006 8:29 am

Lance indicated that his IAC works properly, I'm wondering if it has something to do with selected options (ie, dual MAP for realtime baro).

Scott,

For the most part, my settings are the defaults in MS-II (Al adopted these a while ago).

However, I have tried a number of code combinations and settings. Since I added jumpers to the resistors on the bottom of MS-II, I have had no issues at all.

It's hard for me to post settings because I have so many combinations that I test, many of which are for code that has not yet been released. Putting an MSQ out that isn't compatible with current code would cause more confusion than it solves, I think.

What we need someone with IAC problems to do is to 'scope each of the outputs, both with the stepper connected and with it disconnected. If they could remove the stepper and check it on the bench as well to verify the physical movements, that would be great too. This is what I do in my testing - on car 'real world' testing + bench testing with an oscilloscope and micrometers.

Lance.

69 1/2 Six Pack Bee · Post by **69 1/2 Six Pack Bee** » Thu Apr 20, 2006 8:37 am

In the "For What It's Worth" department, I did solder the standard components onto the SMT pads albeit a tight fit I was successful in being able to set the chip back into the socket without an issue. I took some closeup shots of the install that I will post this evening. I did fire up the truck and I really could not tell if the swap honestly made a difference but I think it did help a little. Unfortunately I was encountering some other issues at the time that I was trying to sort so I was jumping around in the program as the engine was warming up so i can not say anything conclusive. Suffice it to say it was yet another non-productive weekend as far as my tuning efforts went and after careful deliberation, fueled by the fact that I was getting nowhere fast, a decision was reached.

I am not able to be of further help because as it stands right now, I was just having too many tuning issues and I just ran out of time, patience and ambition to continue on. Over the Easter wekend I yanked the entire MS system and went back to the carb/HEI setup. Since I use this truck for the family vacation I needed to not have to continually fool with it to get it to be where I wanted it to be. I just am very short on patience and I guess the older I get the less tolerant I grow with certain things.

I'll let the younger and more ambitious folks sort out the issues at hand. Sorry I was not of more help but I did learn alot.

I will poke around from time to time to see how things have progressed and maybe someone will resolve the ever-elusive IAC issue.

I know that the Megasquirt is a wonderful thing and it is proven itself in many instances and in many different applications!

But, at least for me in this juncture of my life, unfortunately it was not

Good luck y'all and thanx for the time spent here

JSC · Post by **JSC** » Thu Apr 20, 2006 9:14 am

Lance;

Thanks for the reply.

However, I have tried a number of code combinations and settings. Since I added jumpers to the resistors on the bottom of MS-II, I have had no issues at all.

That pretty well answers that question.

I have bench tested the system (see my post 4/10/06, this thread). It doesn't seen to matter what I do on the bench, it always works! This brings me to the conclusion that it's not the 2916 motor driver or the motor, so what's left? That what's driving the 2916.

I have a new scope on the way so when it arrives I'll investigate further.

Scott