Microsquirt/MS2 based transmission controller

[Joe]

OK,

Made a few minor tweeks.Bench test OK. Installed Micro_Trans in truck. Initial checks ok. I will be road testing tomorrow.
I still need to connect my brake switch.

EPC makes some racket, but I'll deal with that later.

Looks good so far.

Joe

[jsmcortina]

Made a few minor tweeks.Bench test OK. Installed Micro_Trans in truck. Initial checks ok. I will be road testing tomorrow.
I still need to connect my brake switch.

EPC makes some racket, but I'll deal with that later.

I agree on the EPC. What is strange is this is being run at the quoted PWM frequency. I'm inclined to think that the specs are may in the wrong units? i.e. should be kHz not Hz.

Keep us posted (and send me any code tweaks please.)

James

[Joe]

Initial road test.

Very pleased overall. Drives nice. Didn't have a chance to make a long road test, but I had to at least drive a few minutes.

Quirks I'm going to try and address.

1)light throttle upshifts. Nice. Downshifts good at 0 TPS. After initial shift out of 1st gear Ii would not down shift again to 1st on a nice easy coast down. Had 2nd gear starts until a key cycle. I need to review logic (hysteresis??) for 1st gear shift.

2)Going to attemp to tune pressure table with a gauge, initially thought heavy throttle shifts(75%) were way to harsh but after reviewing data log I actually hit 1st gear for a brief second.

Need to sort out 1st downshift before going much further.

James,

I'll send you the mods I made anytime but I'm hoping to clean things up a bit before forwarding.

Joe

[jsmcortina]

Excellent!

A few warts to fix, but that's great news.

James

[Joe]

Update-

Lowered downshift hysteresis from 5 to 4
Raised 1-2 shift from 5 to 6 mph.

Result-

Nice, coming to a stop shifts to first gear @ 2mph. I now have first gear start.

Freeway run- works great. TCC ramps on nicely. At first wasn't sure I had apply.
Steady throttle touched the brake. Yep TCC was applied. I can feel the release. Apply is very smooth.
Can also see it in the datalog.

Working on an updated ini. Adding a few things to the datalog. I will test in the next few days.

Hard to believe but my poject is not fully assembled yet. I don't have a shifter boot for my t-case installed at this time.

Sitting at a light the EPC chatter is getting to me. I may have a chance to check a 4l80E with a GM pcm this friday.
When I gather some data I may need some help getting the EPC frequency set correctly (Help James).

If anyone is following this thread I have to say an upgrade from a TH400 to a 4l80E is worth the trouble.

The Micro/Trans setup is starting to work well. Very pleased so far.

Joe-

[jsmcortina]

Thanks for the updates, I am looking forward to driving with it too. I had hoped to revisit the EPC frequency today, but it is too late now. It should be a matter of changing some of the PWM parameters, but I need to get the Microsquirt out of the car and onto the bench to confirm before I send any diffs or test code.

James

[Joe]

James,

I have my microsquirt back on the bench and I don't mind testing. I know you are busy with MS3 and I'd hate to distract you.
(I'm looking forward to that also). How would I go about upping the HZ on the EPC timer by a factor of 10??

Spec is 292.5hz. On the bench I measured 30hz. If you can post some some updated timer settings I'll try them out.
I'm looking into what to change but I'm not real sure right now.

Thanks,

Joe

[Joe]

Still studying,

Looks like reducing PWMSCLB will do the trick. It's late tonight but I am going to experiment friday after work.

I am having trouble making the math work. (Matching what I see on the scope)

pwm HZ = bus clock/ preiod value/ prescaleB / scaleB

ClockA Pins 1,4 or 5
ClockB Pins 2 and 3

Joe

[jsmcortina]

I rechecked the PWM parameters and clocks A&B were transposed.

I believe I've fixed the code and have tested on the bench with my Microsquirt. I checked the frequencies on the scope.
EPC = 296Hz, TCC = 32Hz.

I updated the MS2 and GPIO sections of the code, but haven't tested. The GPIO frequencies are a bit of a problem due to the pins chosen. EPC and TCC are on the same clock, so you can't get both frequencies right without compromising precision. Presently I've set EPC to 732Hz (just over double the spec) and kept TCC at 32Hz. Happy to adjust if anyone uses it.

New code:
http://www.jsm-net.demon.co.uk/trans/files/alpha/trans20090626.zip

Joe,
did you change any code or have you been tuning. I'm keen to roll any enhancements you have made into the official code.

Any other testers out there?

James

[atomic6]

James,

We have tested it on the nova, worked great before the recent changes, the problem we had was the lock-up would not work the way it should, it was always locked-up. We just put a switch on it,Kyle(1320 video) has some footage with us on the dyno(508 rwhp @10 lbs)

I had a long reply all typed out with links and data, then went to post it and the site timed out,or something happened . I intended to post again the next day,but I got side tracked.

The setup we have is not using the EPC, it is converted to vac. modulated line pressure, but will load up the newest code,and try the lock-up again.

We are very happy with it, the logging of the trans is great on the Dyno, and soon the dual fuel will be controlled by the MSII, so then the two will be connected together by can bus, and I can merge the logging(right now we use two computers)

Thanks for all of your work.
I have a bunch of logs,from the dyno if you want to look at them, but there is not much to see.

[atomic6]

first time at dyno,

http://www.1320video.com/gallery/DynoSport4-25?page=3
http://gmmodernmuscle.com/forums/viewto ... &start=100



second time at dyno. less pictures and data
http://gmmodernmuscle.com/forums/viewto ... &start=120

Kyle has video some where. and I have all the logs from the two days.

we will be going back to the dyno soon,to the track soon, and on the street testing soon to.

[Joe]

Atomic6,

Same issues I had.

EPC & TCC bug still there.(Inverted Signal, Not hz)

James,

They can be fixed with the following lines of code-

// Pointers used for more flexibility with hardware

*pwm_tcc_addr = ~outpc.tcc;
if ((pg4_ptr->setting1 & 0x30) < 0x20) { // 4L60E or 4L80E
*pwm_epc_addr = outpc.epc; // GM trans work that 0 duty - full pressure, 40% = min pressure.** ~
} else {
*pwm_epc_addr = ~outpc.epc;


I'm not familiar with the differences between the MS2 INJ driver VS MicroSquirt so I didn't post the changes.(On my list of things to do).

I was affraid that inverting TCC and EPC might break MS2 code.
I was thinking an additional condition for pointer assignment for board type and trans type is needed.

The above lines work on the Microsquirt. EPC and TCC.

Joe-


Edit: New issues. I installed a line pressure gauge and an ammeter for EPC check out. Didn't check with the old code installed
(Running .011 with mod above) Seems line pressure is not as linear as I would have thought.
Percent in software vs actual EPC amps. Pressure tables are going to need some dial in.(Reduction)
Verified EPC 0% no line pressure, 100% full line pressure.
The GM Line pressure spec is Pressure Vs EPC amps. Next step I think would be some data that would correlate %EPC vs EPC amps. (Agree? disagree? comments??)

Any chance of getting a 0-5v channel to log? I was looking into the pressure sensor the GPIO project was recommending.

Right now I have a mechanical gauge installed. Would be nice to have an electric one that I could log.

[devojet]

Hi James,

Just thought I would post to let you know I am still following the progress of this code. My Hot Rod project is taking much longer than I thought it would. Haven't done anything to it for the last 6 months as I bought a new race boat. When I get some time I will check out the GPIO changes you made and bench test them.

Cheers,

Daniel.

[jsmcortina]

I was wondering what had happened on the Nova, nice to hear the trans controller is still on there. (Shame we aren't _quite_ there yet with MS3 because that will eventually be able to give you 8 channels of sequential gasoline and 2 channels of methanol control.)

I think the EPC inversion can be fixed by changing the PWM polarity and I'm going to try that now. I'm sure I had it correct at one point during development and must have broken it later.
The outputs were already inverted for the MS2, which is where I was testing a few months back, I have now used PWMPOL to invert the Microsquirt outputs. New code posted. http://www.jsm-net.demon.co.uk/trans/files/alpha/trans20090627.zip

What is the comment about Hz ? I changed the prescale values and believed that the TCC and EPC frequencies where correct on the scope.
If I'm missing something obvious do let me know!

The GM tech guide on the 4L80E has a graph of EPC pressures against duty. Is this really critical though? i.e. can you not tune it out through the tuning curves?
If the consensus is that we _do_ need a table it isn't that hard to add one and I'll do it.

100_0290small.JPG

(EDIT, reduced image size)

James

[Joe]

James,

.012 looks good. On to tuning.


Joe-

[atomic6]

Tryed the new code, the convertor is now not locked all the time , but now I can't get it to lock at all.

I may be doing something wrong, I am putting all the settings the way I think they need to be so it locks at a low mph,this is so I can test it on the jack stands.

I will attach my last log, and msq.

I also went into the trans ini, to log more things, and that works

next step for me is to bring my external trans pressure sensor in on ADC1, and log that.

Getting closer,every day.

I also configured Tunerstudio to work on the trans controller, loving that...

[jsmcortina]

Checking the code, I have it set so it will only lockup in 4th, your log only shows 1,2,3.

Is lockup in 3rd desirable too?

James

[atomic6]

Checking the code, I have it set so it will only lockup in 4th, your log only shows 1,2,3.

Is lockup in 3rd desirable too?

James

In my Grand National, I lock it up as soon as it hits 2nd, some people don't like it locked in wot, but I think it is benifial if you have stuff that won't break, and a Turbo.(the torque multipication helps some cars, but it is better on turbo car to have that extra load.)

At the very least I would have it lock in third, most cars will run till the end of the track in third.

If we could lock up in any gear that would be great, the trans hydraulicly won't let it lock in first, but we have a triple disc lock-up and would like to use it to get the most to the ground.(that the builder dared us to break!!!!)

we can always just lock it up with a switch if it is alot of work to change, most people needing this controller are going to modulate the lock-up, for us that would be bad...

[jsmcortina]

Changing the code is not a problem, I'll probably add "allow" options to set which gears are allowed. Any thoughts on smooth operation or delays after shift in these other gears. I want to ensure it doesn't stay locked continually - we need to unlock during shifts at least.

James

[atomic6]

Maybe add a high performance mode, lock in any gear, unlock before shift, and then x rpm after shift. or stay locked thru shift(with warning)

Jim(Parish) use to keep his 4l80e locked thru shifts, untill he broke his input shaft that way. But,his truck was 5000lbs,and making over 1000 hp, one of my cars was as fast as his truck, but not as heavy, or powerful. that car has a 200r4, I would lock manually, and it would stay locked thru the shifts.

The chip in the factory ecm, would only lock at 60 mph and above in forth gear, and not at WOT.(normal mileage type lockup useage)

But, we(turbo buick people) would install a switch to turn on at the start of a race. The 200r4 also does not allow lock up in first gear ever, then the shift to second allows the lock up to engage, and it stays locked untill the trans returns to first,or the switch is turned off.(returning to ecm control of the lockup)

Lockup used this way is not for everyone, everyday. But very usefull in racing.

Thanks,
John