4g63 "Unipolar" stepper IAC(6 wire stepper)

md95 · Post by **md95** » Tue Dec 04, 2007 5:22 pm

I want to start playing around with idle control, and my engine has what's called a "unipolar" stepper motor, with 6 wires. Two are 12v power wires, the rest are coil wires, each of which need grounds to move the stepper. In the Idle valve documents, it mentions the unipolar stepper, but does not really go into use/wiring of it. I'm assuming the MS2 card will drive this stepper just fine? Is there a certain configuration it needs to run, eg special settings? I do have a Bosch 2 wire idle valve as well, which may be easier to set up and get running, but it will require some fab work..I'd like to use what the engine already has, of course.. I'm not in huge hurry to run closed loop idle, as when the motor is fully warm, idle is fine, and dont have A/C..want most control during cold warmup..

Here is a pic of the OEM wire schematics for the stepper:

Any insight to using this stepper, would be greatly appreciated!

Post by **muythaibxr** » Tue Dec 04, 2007 6:11 pm

Just so you know, the 2 wire PWM valve is going to respond faster, and work better with PID closed-loop idle speed.

The stepper motors move very slow, so I might need to make some changes to make those work better, and they may not ever work as well as the PWM idle valves..

md95 · Post by **md95** » Tue Dec 04, 2007 6:20 pm

muythaibxr wrote:Just so you know, the 2 wire PWM valve is going to respond faster, and work better with PID closed-loop idle speed.

The stepper motors move very slow, so I might need to make some changes to make those work better, and they may not ever work as well as the PWM idle valves..

Hmmm..well i would like to get into closed loop idle before too long, but my main concern is cold start/warmup idle control. How do the steppers do in that respect? I keep re-reading the docs for Idle valves, and it just mentions the unipolar stepper, but does not go into actually implementing it..maybe these valves cannot be used? If not, then it's the Bosch 2 wire valve for the win. Plumbing it will be the hardest part..since it's a turbo car, i need to to tap the intake pipe before the TB for the air inlet, then plum the outlet to the manifold..the valve even when fully shut, still flows air..

md95 · Post by **md95** » Wed Dec 05, 2007 3:24 pm

Anyone have any insight to using 6 wire steppers? I've been searching all the docs, and it mentions them, yet doesnt go into using them, or even say that they WONT work at all..rather confusing.

turbotiger · Post by **turbotiger** » Wed Dec 05, 2007 3:55 pm

For the short amount of time I had the car running with megasquirt so far, I had no problems using the stock ecu in parallel controlling the idle. But yah, it would be great to see MS control the 4g63 6 wire stepper motor.

racingmini_mtl · Post by **racingmini_mtl** » Wed Dec 05, 2007 5:08 pm

I have read of someone simply connecting the MS 4 IAC connections to the connections 1, 3, 4, and 6 on your schematic. I'm not sure exactly how it was connected and I don't have a link but hopefully this is some starting point. It was not on a 4g63 but it was a unipolar stepper with 6 wires.

Jean

md95 · Post by **md95** » Wed Dec 05, 2007 5:18 pm

racingmini_mtl wrote:I have read of someone simply connecting the MS 4 IAC connections to the connections 1, 3, 4, and 6 on your schematic. I'm not sure exactly how it was connected and I don't have a link but hopefully this is some starting point. It was not on a 4g63 but it was a unipolar stepper with 6 wires.

Jean

I wondered about that..just not connecting the 12v "center taps", and just drive the coils like a 4 wire stepper..I just wasnt sure if this is possible, or even if it will work like normal..I'm all about trying it, i just dont want to burn up the stepper chip! I will measure the coil's resistance and see how it compares to a 4 wire stepper..Is there a certain coil resistance "limit" that needs to be met, in order to not overdrive the stepper chip? Or is this chip pretty robust for current draw?

racingmini_mtl · Post by **racingmini_mtl** » Wed Dec 05, 2007 5:33 pm

Looking at the datasheet (UDN2916LB), it has some clamping and thermal protection so it should be pretty robust. It can drive 750mA continuously and 1A peak.

md95 · Post by **md95** » Wed Dec 05, 2007 6:02 pm

Wow, 1 amp max? Is that all these normal steppers use? I saw the GM stepper is anywhere from 35-56 ohms on the coils, so if mine measure out to at least that, then i should say it will be safe to try..If the resistance is really low, then i'll opt to not risk burning the chip up..

md95 · Post by **md95** » Fri Dec 07, 2007 9:28 am

As soon as I get my hands on a stepper for the car(mine was removed and blcoked off before i got the car), i'm going to test this out just by using the 4 coil wires and see if it works..hopefully this weekend i'll have some feedback on any success or failure... Acording to the tech docs for the OEM stepper, resistance from the center tap to each end of the coil is 26-35 ohms, so resistance from coil end to end should be plenty, and *should* be fine for the stepper chip to drive..

md95 · Post by **md95** » Sat Dec 08, 2007 12:49 pm

I've got a spare stepper to work with now...Is there a good way to test it's operation on the bench? For example, can i just power up the MS, without it seeing the CAS sensor to tell it it's running, and adjust the coolant temp to watch the stepper operation? Or does the code need a running engine? I really want to test this on the bench to see if the MS will even operate this stepper, as i dont have a IAC on the car..it was removed and blocked off, so it will take some work to get it back on.. The stepper's pintle does twitch when i just use the 4 coil wires, connected to a drill battery, so it looks positive so far. Coil resistance is around 60 ohms per coil..good there as well. Any help?

md95 · Post by **md95** » Sun Dec 09, 2007 11:08 am

Ok, got the 1G 4g63 stepper wired up on the bench, and initial results is it works! The pintle of the stepper moves in when MS powers up, and with my test CAS, the pintle comes out a bit for "cranking", and after rpms go up, it goes out a bit more to the right step for coolant temp. I still have yet to install this in the car, hopefully sometime this week, but if any other MS2 Extra 4g63 users want to try it out, here is some info...

For wiring: (this was on a 2.2 PCB)

X2 goes to Stepper wire color white(stock ECU pin 67
x3 goes to Stepper wire color black(stock ECU pin 68
x4 goes to Stepper wire color blue(stock ECU pin 58
x5 goes to Stepper wire color Yellow(stock ECU pin 59

You do NOT connect the 2- 12volt power wires at the stepper, leave this UNPOWERED. You MUST cut the 2 power wires RED, at the stepper connector!

If you connect the MS stepper wires at the ECU, be sure to cut and isolate the wires from the ECU that go to the stepper, dont just tap the wires...

Basic settings remain the same, and from my initial testing, its *around* 135-150 steps from full retracted, to full open(iac port is closed, no air). This does need some in car fine testing to verify full step range, but it should be real close.

Of course, this needs in car testing still, but this should get people started anyway. From what i know, the 2g DSM stepper is pretty much the same, so it *should* work on the 2g 4g63 as well...

turbotiger · Post by **turbotiger** » Mon Dec 10, 2007 1:07 am

Good to hear that it works does indeed power the stepper motor. I'll have to try it out when I get my car running a bit better at idle.

md95 · Post by **md95** » Wed Dec 12, 2007 7:33 am

Well I got it installed and wired up, and it works..sort of.. The problem is that the pintle doesnt seem to move during the warmup cycle..It moves just fine at power up to full retract, and when the engine is warm, it goes fully closed, but the small warmup steps just dont seem to react..I have step size set to 5.0 ms, and min steps to move at 5. Any ideas?

robiniddon · Post by **robiniddon** » Mon Jan 14, 2008 3:02 pm

EDIT: I think the suggestion I posted here might break your motor or controller circuit, so I have deleted it to be safe!

Did you get this to work?

Looking at the diagram you posted, it isn't entirely clear which of the four driving inputs corresponds to which phase.

See this diagram:

uni-polar steppers

The diagram is a 5-wire type (like mine) but you have a 6-wire type, so you have two commons.

Do you know whether your coils represent 1+4 and 3+2 or 1+2 and 3+4 ...

Cheers,
Robin

md95 · Post by **md95** » Mon Jan 14, 2008 3:21 pm

robiniddon wrote:Did you get this to work?

I was thinking that it doesn't work because:

The 2 phases are wired with three wires each (as you know).
Each phase is wired to produce the opposite field - so you pull down one side to get N and the other side to get S. If you pass a current end-to-end on one phase (like you've wired it) you will get the N and S at the same time - of course they won't be perfectly balanced, so it will sort of work, but not properly.

You could instead take the center tap (the two wires you cut) and wire those in place of two of the ends (i.e. only use half of each coil). In that way you'll reverse the current through half the coil only generating an N or S as required. You could probably verify this by snipping 3 & 6 and then wiring to 2 & 5 instead.

Does that make sense?

Meanwhile, has anybody ever got a 5-wire unipolar driver to work? I think it will work just fine by connecting it up as normal to IAC1A, IAC1B driving either end of one coil and IAC2A, IAC2B driving either end of the other coil, then leave the +12v common connected. The H-bridge will then raise both ends of one half-coil to +12 (i.e. zero or negligible current) and setup a current flow across the other end.

Incidentally, I also think your 6-wire stepper would work like this too - reconnect 2&5 to the +12 supply and it becomes a 5-wire stepper again.

Cheers,
Robin

Yes, i've got it to work pretty reliably so far. Maybe only once in a while will it not fully close, but i'd say 90% of the time it works fine. I use 6.0ms step size, and 8 steps min to move. I also had to jumper the 2 resistors on the MS2 card. That helped. I dont think i'm too clear on how you described the workings of the stepper, but as far as i know, the 2 power wires are center tapped, and the ECU sends grounds to each of the 4 ends of the coils. Basically this makes for 4 independant coils out of 2. Only one half of each coil is energized at one point in time. By connecting the 4 wires only, this effectively makes it a 2 coil stepper, as the entire coil is now being used to create the magnetic field. I thought about this for a bit, and wondered if this would have an effect of less resolution per step, but i'm not sure if it mattered or not. It works, and works fine for the most part. The only real issue i have now, and have not yet had time to really get into it, is that during cranking, the IAC steps immediately go to full hot running steps..this mostly happens when the engine is not overnight cold though. Every morning when i start it, it seems to go to my specified cranking steps. So basically, the IAC is doing what it's being told to, i just cant figure out yet why the MS is telling the IAC to go to hot running steps during cranking, instead of the specified cranking steps.

I think i kind of understand what you are saying, but by using only the 4 wires(each whole coil), it's not creating N and S at the same time, since current from the stepper chip is only going one way at a time. The MS is reversing the polarity on each end the coils, to create the N and S, where as the OEM was using the center taps and half of the coil to create the N and S. Does that make sense? There was 12v on the center of each coil, and the OEM grounded each coil's end. So half the coil was N, then it sent ground to the other end of the coil, to create the S. The MS is just making one end of each coil +, and the other -, to make one pole, then it reverses this to make the opposite pole.

It's in my opinion, that this type of stepper needs just a bit more current than what the stepper chip can supply, hence the occasional not fully closing issue. In the OEM configuration, 12 volts is supplied by a relay, and the ECU grounds the coils with higher current transistors. If there is a way to amplify the stepper chip's outputs, i'm all for it, although i dont think that is easily accomplished. The fact the outputs change polarity, make that idea a bit more unrealistic, as you couldn't just use simple transistors to amplify the signals. I *could* try as you described, by only using half of each coil..this may or may not work..This effectivly halves the impedance presented to the stepper chip(this may put more current demand on the chip, more than what it's capable of), and also halves the potential magnetic feild inside the stepper, although it works like that stock..its easy enough to try though.

robiniddon · Post by **robiniddon** » Mon Jan 14, 2008 4:00 pm

As long as it works ;-)

If you think about it, though, the OEM was driving each wire in turn to create (for the sake of argument) N on one coil and nothing on the other 3; the motor armature spins to point at the N.

In your application (centre taps disconnected) you are passing current through one half coil the same way the OEM was, so that will generate an N. However you're passing current the opposite way through the other half of the same coil (because you've connected it to +12 instead of 0) so you'll generate an S in that (half) coil.

If your half coils are opposite in the motor, this will generate:

Code: Select all

   -

N<-o<-S

   -

and all will be well; if they are neighbours in the motor you will have:

Code: Select all

   S
   |
N<-o-

And I guess this will either weaken the affect of the N or skew the armature off to the left or right a bit - this almost certainly won't matter from an accuracy point of view, but if the full motor torque was needed to reliably turn the armature you might find it misses steps and stuff.

Cheers,
Robin

md95 · Post by **md95** » Mon Jan 14, 2008 6:41 pm

robiniddon wrote:As long as it works ;-)

If you think about it, though, the OEM was driving each wire in turn to create (for the sake of argument) N on one coil and nothing on the other 3; the motor armature spins to point at the N.

Ok, i get this point, and makes sense.

robiniddon wrote:In your application (centre taps disconnected) you are passing current through one half coil the same way the OEM was, so that will generate an N. However you're passing current the opposite way through the other half of the same coil (because you've connected it to +12 instead of 0) so you'll generate an S in that (half) coil.

I dont understand how i'm passing current the opposite way through the other half of the coil..The MS supplies + on one end, and - on the other. The entire coil becomes one pole, not just half of it. The 2- 12 volt taps are not connected to anything. There really are not any "half" coils in the stepper, there are only 2, but the center taps automatically make for 4 coils, being able to change poles by reversing the polarity when one ground turns off on each end of the coil. Minus the center taps on this stepper, it's physically built like any normal 4 wire stepper. The OEM uses center taps to keep the internal circuitry simple, and it provides a way to easily change each coil's polarity to change it's magnetic pole. The MS stepper chip changes polairty on each end of the coil for us, so it's effectively doing the same thing- changing poles on each coil.

In the OEM configuration, it's using only half of each coil to make a pintle movement, and in the MS config, it's using the entire coil to make a movement. Thats really the only difference. Maybe i'm missing something here?

I'm willing to try it out and see what happens, but my only major concern is the input impedance to the MS stepper chip. IF i only connect half of each coil, then it will be in the 25-30 ohm range. I dont know if the chip can handle that much current draw. I'd rather not risk it, esp since i jumped the 2 power resistors.

devojet · Post by **devojet** » Mon Jan 14, 2008 8:16 pm

I've been watching your progress on this thread as I have a 6 wire stepper IAC on my toyota V8.

I can't see any reason you couldn't add some power transistors to the IAC circuit. If you reconnect the center taps to 12V and use 4 PNP power transistors then it should work the same as the factory ecu. With the PNP's the output will float when the IAC driver chip outputs 12V and the PNP's will ground the coil when the IAC outputs 0V.

To connect the PNP's, the Base connects to The IAC chip via a resistor. The collector is connected to ground and the emitter is connected to the IAC coil.

Cheers

Daniel

md95 · Post by **md95** » Mon Jan 14, 2008 8:59 pm

devojet wrote:I've been watching your progress on this thread as I have a 6 wire stepper IAC on my toyota V8.

I can't see any reason you couldn't add some power transistors to the IAC circuit. If you reconnect the center taps to 12V and use 4 PNP power transistors then it should work the same as the factory ecu. With the PNP's the output will float when the IAC driver chip outputs 12V and the PNP's will ground the coil when the IAC outputs 0V.

To connect the PNP's, the Base connects to The IAC chip via a resistor. The collector is connected to ground and the emitter is connected to the IAC coil.

Cheers

Daniel

Thats a good idea..I dont see why it wouldnt work..Any specific PNP's or resistor values for a start? I thought PNP's were meant to switch positive voltage, not ground?