IAC tuning. Am I crazy? (Solved!)

Post by **Bruce Bowling** » Wed May 03, 2006 6:29 am

BottleFed70 wrote:
I checked my wiring today and I seem to have the "o-ring style" GM IAC with the inline 4 pin connector. I measured coil resistance of approx 47ohm/ Please see the attached picture as to how I have it wired into my relay board. I believe I jumpered everything properly during the build, but I must admit I found some of the IAC wiring instructions a little confusing.

If you measure 47 ohms across the inner two terminals, and another 47 ohms or thereabouts on the outer terminals, and nothing from the inner two to the outer two then you are O.K

BottleFed70 wrote: One thing I couldn't tell from the picture was wether I'm supposed to be looking at this from the perspective of looking at the IAC pigtail connector or the IAC itself. Perhaps I have the polarity of both coils backwards? I took it from the perspective of looking at the IAC.

In this case it does not matter, but follow the IAC actual pins. If you had both coil's polarity reverse the motor would still turn correctly. You would have to get one coil reversed and the other correct, and then the motor would just turn opposite from what you expect. In this case just swap one coil polarity, either one.

BottleFed70 wrote:
I'd love to do a continuity test from the IAC pigtail connector to the MSII daughter card(or elsewhere on the ecu) if you can give me some instructions on what I'm looking for? (IE which IAC pin should go to which MSII pin).

The thing to note is that one driver pair is X2 and X3 (respective pins 38 and 37 on the 40-pin processor socket) and the other is X4 and X5 (36 and 35 on the processor socket). Pair X2 and X3 connect up to one coil. Pair X4 and X5 connect up to the other coil. Keep this pairing straight and it should be OK.

One test for the driver is to use two LEDs in "anti-parallel" arrangement. All this means is to take two LEDs and hook one LED anode to the other's cathode, and vise-versa on the other leads. So you have two LEDs pointing in different directions. Put in a series resistor, say 470 ohms, and hook this arrangement to X2-X3 and make up another LED pair and hook to X4 and X5. Depending on the current direction one or the other LED will glow. So when the motor moves the LEDs will flicker back and forth. You could even wire this up in parallel with the stepper motor and run the wires into the passenger compartment and use this as a visual of direct stepper driver activity.

- Bruce

Post by **grippo** » Wed May 03, 2006 7:29 am

BottleFed70 wrote:
Hi Al,

I gave the 2.35x1 code a try today with mixed results.

For some reason this code didn't work? The IAC was not moving in 4 step intervals. Luckily I datalogged the warmup, and you can clearly see in the datalog that it was moving in random steps from 1 to 4. 1,2,and 3 step movements were the most common, but it did take a few steps of 4. Please see the attached datalog.

I will take a look at this tonight - earlier if I can find the source code.

Post by **grippo** » Wed May 03, 2006 3:41 pm

BottleFed70 wrote:
Hi Al,

I gave the 2.35x1 code a try today with mixed results.

For some reason this code didn't work? The IAC was not moving in 4 step intervals. Luckily I datalogged the warmup, and you can clearly see in the datalog that it was moving in random steps from 1 to 4. 1,2,and 3 step movements were the most common, but it did take a few steps of 4. Please see the attached datalog.

I tested the code very carefully and it worked perfectly. But, I think I know what happened. I set the no. accel steps = 4 in the defaults section of the code. You loaded your msq on top of this and it had 0 for no. accel steps, so it wiped it out. I think all you have to do to get it to work is go into the idle control settings in MT and set the no. accel steps to 4. You don't have to re-download or anything.

BottleFed70 · Post by **BottleFed70** » Wed May 03, 2006 6:58 pm

grippo wrote: I tested the code very carefully and it worked perfectly. But, I think I know what happened. I set the no. accel steps = 4 in the defaults section of the code. You loaded your msq on top of this and it had 0 for no. accel steps, so it wiped it out. I think all you have to do to get it to work is go into the idle control settings in MT and set the no. accel steps to 4. You don't have to re-download or anything.

Hehe.. If there's a way to screw it up, you'll bet I can find it.

Unfortunately I can't give this a try until tomorrow evening, but when I do I'll make sure to post another datalog.

Philip Lochner · Post by **Philip Lochner** » Wed May 03, 2006 11:43 pm

grippo wrote:I set the no. accel steps = 4 in the defaults section of the code. You loaded your msq on top of this and it had 0 for no. accel steps, so it wiped it out. I think all you have to do to get it to work is go into the idle control settings in MT and set the no. accel steps to 4. You don't have to re-download or anything.

According to the Megatune manual "no. accel steps" and "accel step size" is not used. Does this apply to 2.35?

Post by **grippo** » Thu May 04, 2006 6:28 am

Philip Lochner wrote:
According to the Megatune manual "no. accel steps" and "accel step size" is not used. Does this apply to 2.35?

The manual is correct, but I decided to use the no. accel steps for the purpose of forcing 4 (or any user input) steps per move. This is ONLY in the experimental code I put out yesterday. If this works out we will put it in the official code and then update the manual. It was my fault for forgetting to tell Bottlefed to set this input. If you set it to 1, the program should operate like v2.35.

BottleFed70 · Post by **BottleFed70** » Thu May 04, 2006 9:53 am

grippo wrote:
Philip Lochner wrote:
According to the Megatune manual "no. accel steps" and "accel step size" is not used. Does this apply to 2.35?
The manual is correct, but I decided to use the no. accel steps for the purpose of forcing 4 (or any user input) steps per move. This is ONLY in the experimental code I put out yesterday. If this works out we will put it in the official code and then update the manual. It was my fault for forgetting to tell Bottlefed to set this input. If you set it to 1, the program should operate like v2.35.

Do I need to worry about the accell step size? Right now I have it set to 0.5ms with my regular step size being 2.7ms.

Post by **grippo** » Thu May 04, 2006 4:52 pm

BottleFed70 wrote:
Do I need to worry about the accell step size? Right now I have it set to 0.5ms with my regular step size being 2.7ms.

The accel step size should have NO effect - it is not used in the program. Just set the regular step as you have been using and the no. of accel steps to 4. If this helps, but not quite there, try settin to 6 or 8. Do not change anything else when you do this - we only want to change 1 variable per test.

BottleFed70 · Post by **BottleFed70** » Thu May 04, 2006 6:17 pm

grippo wrote: The accel step size should have NO effect - it is not used in the program. Just set the regular step as you have been using and the no. of accel steps to 4. If this helps, but not quite there, try settin to 6 or 8. Do not change anything else when you do this - we only want to change 1 variable per test.

I set it to 2.7ms just in case.

So I set the # of accell steps to 5 and datalogged a complete warmup again. Nothing else has changed since the last datalogg.

This seemed to help a little. Looking at Tuesday's log, at 221 steps my idle was about 1500rpm (Tuesday already seemed to be an improvement as I has been seeing 1600rpm). Today at 223 steps my idle was about 1500rpm. BUT..again after simply turning the key off and starting the car again, my idle dropped down to normal at about 800rpm.

Please see the attached dataloggs. I've also included my MSQ just in case. I had to remove Tuesday's datalog in order to post todays. Let me know if you need it and I'll email it or something.

Tomorrow I'm going to try again, this time with # of accel steps set to 10

Post by **grippo** » Thu May 04, 2006 7:57 pm

The code is working, but its not having the intended effect. Your msq looks fine, I can't find anything in the software or datalog that would explain this. Is there anything you can think of that is different on restart. As an example, there was some lingering warmup enrichment on run 1 which was gone in run2 because iit started at oeprating temperature. But this is only a few percent enrichment. Now if your car was starved for fuel and that's what made it idle lower on the second run, that's a possibility. But it would sound like crap and I'm assuming it idles smoothly on restart. In run 1, if you had let it run another 5 min would the idle go down ?

I very much doubt that going to 10 steps would make any difference since there is no improvement with 5. I will talk this over with Bruce in the morning. He has a board he may want you to try.