Trans control - 0020 release

[jsmcortina]

First release in a long while:

Changes since the last release
Tweak ini file so signature fetch matches size.
Add defines for hardware types.
Use more port pointers instead of if () {...} sections in main.
Add Microsquirt V3 wiring. (Lightly tested.)
Merge in GPIO/A340E port setup from Daniel deVoigt.
Add else clause for non-paddle per Daniel.
Change can_poll_id setting to a type in number.
Add tooltips for all settings and curves.
Use "Require Power Cycle" on some settings.
Fix serial buffer size.
Enable thermistor calibration.

The zip can be downloaded from:
http://www.msextra.com/doc/trans/files/ ... 140911.zip

A current TunerStudio is required. As a fair bit has been changed, it is easily possible that there will be breakage.

Successes, failures, bug reports and suggestions welcome here.

James

[prof315]

Well the timing on this couldn't be better! I have just started preliminary research on the possibility of using a V3 microsquirt tied in via CAN to an MS3-Pro and using the combination to control a 2.2L Ecotec and 4T40E that are installed in a custom trike. As far as I can tell from the info I have, the 4T40E is very similar electronically to a 4L80E the only major difference being the trans pressure switch. The 4T40E has only 3 switches in it compared to the 5 switches in a 4L80E.

[jsmcortina]

I'm back up to speed on the code now, so if you have the full specs on solenoid patterns and switch patterns etc, I can encode them. The way my trans code is written I've been trying to build in (hardcode) the selections with an aim to make the user interface simpler and with fewer settings.

James

[prof315]

I'm back up to speed on the code now, so if you have the full specs on solenoid patterns and switch patterns etc, I can encode them. The way my trans code is written I've been trying to build in (hardcode) the selections with an aim to make the user interface simpler and with fewer settings.

James

Here's the shift solenoid pattern for a GM 4T40E or a 4T45E ( the 45E has a higher torque rating but is otherwise the same beast)
1st 1-2 on 2-3 off
2nd 1-2 off 2-3 0ff
3rd 1-2 off 2-3 on
4th 1-2 on 2-3 on

I don't have the pressure switch pattern handy but I can get it in a day or so. And thanks James!


Jeff

[Matt Cramer]

I'm getting the attached error message when I try opening the default tune. Maybe a typo in the INI?

[prof315]

I set up a "sample project' using the .ini and default tune in the zip James linked in the first post and no problems with it.

[denmah]

i had no problem creating a microsquirt and TCU canbus project, all looks good!

[denmah]

anyway to add in shift by RPM only when over a certain TPS amount? kinda like hptuners stock pcm

if you are wide open, it shifts by RPM instead of mph

makes drag racing more consistent for sure when its not looking at mph at all

[denmah]

sorry to add all the posts, but is that MAX rpm? in gear under base settings?

[Matt Cramer]

i had no problem creating a microsquirt and TCU canbus project, all looks good!

Interesting, I don't get the error if I create a new project. I did, however, get the error when I tried opening the file by double clicking on it and using a temporary project.

[ap67et10]

I am excited about this! I will load this up as soon as I can. Are you still thinking upshift and downshift tables are going to be worked in?


Thanks,
Andrew

[jsmcortina]

Ooh.. lots of interest here. Do be aware that this release is an alpha ! Keep the questions and ideas rolling in. I intend to be looking at code again after the MegaMeet.

Some answers:
Matt C- I think I need to add a default value for tsCanid to the ini file to clear that up.

Jeff - if 4T40 is very similar to 4L80E, then it should be fine to add it. (For reference, 4L80E is the easiest trans, 41TE is a complete pig and doesn't work right.)

Matt H - yes, the shift method could be changed, it isn't difficult code. The "Max RPM" setting is a placeholder, it isn't currently being used.

Andrew - yes, I plan to add downshift curves too.

James

[jsmcortina]

I implemented "new Microsquirt wiring" as per: http://www.msextra.com/forums/viewtopic ... 23&t=55785

James

[ashford]

i have a few suggestions
1. divorce pc from tps and use map instead, its more realistic for turboed applications(this ended with the v5 gpio code)
2. downshift tables would be awesome, that gives infinitely adjustable hysteresis for gear shifts instead of a single one for all gears/loads
3. torque reduction and or pressure reduction for shifts(probably difficult to do) ie retard timing or set trigger in engine ecu for flatshift?

as it sits there is no support for 4r70w or i would of run this on my trans when i had it. the outputs are identical to 4l80 the inputs are different though. the gear lever position sensor is a potentiometer and give a voltage for gear, ther are only 3 physical forward gear selections 1-2-od, fords have a od cancel switch. newer trans have a digital input it actually has 4 chanels 2 of which can be used to determine gear position, i think their intent was to have redundancy and it also has positions in between gears so the ecu can tell which way the lever is going. i studied the template of trying to use a switch inline with the digital one to select between d and od but had conflicting patterns.
there are 2 vss types a 6 hole and a 24 tooth on the output shaft(direct drum actually), no input shaft sensor.

[devojet]

Good work James.

The down shift curves sound good. I have found the single downshift value to be a comprimise. I actually started coding that feature a while ago on the older version but never finished it.

Cheers

Daniel.

[ap67et10]

after opening it up, the only thing I don't see is the O/D on off selection. That would be nice for the 4r70w as ashford states. I see the selector voltages and pretty much everything else. I'm going to load it into the GPIO soon and test it on my 4r75w. I'm tired of some of the intermittent weirdness I get with megashift code. I'm assuming output shaft speed sensor is on default pin selection for GPIO?


Andrew

[prof315]

Ooh.. lots of interest here. Do be aware that this release is an alpha ! Keep the questions and ideas rolling in. I intend to be looking at code again after the MegaMeet.

Some answers:
Matt C- I think I need to add a default value for tsCanid to the ini file to clear that up.

Jeff - if 4T40 is very similar to 4L80E, then it should be fine to add it. (For reference, 4L80E is the easiest trans, 41TE is a complete pig and doesn't work right.)

Matt H - yes, the shift method could be changed, it isn't difficult code. The "Max RPM" setting is a placeholder, it isn't currently being used.

Andrew - yes, I plan to add downshift curves too.

James

Question: The TPT inputs.... Are these coming from the external (PRNDL) range switch or the internal pressure switches ? Either way I'll get the info for the 4T40E but which info is needed? The 4T40E even uses the same shift,TCC PWM and Pressure control solenoids as a 4L80E plus electrically identical input shaft and VSS sensors. The trans temp sensor is the same electrically as well.

[jsmcortina]

The 4L80E uses a number of internal pressure switches (three wires) to determine selector position.

James

[Paul_VR6]

Glad you are back at it. I was looking to do gpio for my 5l40e but I think this option would be better if you are adding more support. I will try and post details over the weekend.

Quick edit for the link: http://koti.mbnet.fi/pro-ht/Kuvat/autom ... /5L40E.pdf

[prof315]

The 4L80E uses a number of internal pressure switches (three wires) to determine selector position.

James

Ok that's the major difference from a control standpoint between the 4L80E and the 4T40E. They both use internal pressure switches to determine the actual gear engagement but have a different number of switches and different logic patterns. I_think_ they both use 3 wires however. Gear selector position is still done with an external switch that is moved by the selector shaft and also contains the neutral safety switch. So it looks like I need to find info on the 4T40E pressure switch logic pattern.