Trans control release pre-1.0.2 beta3

'Microsquirt' Transmission Control development and support.

Moderators: jsmcortina, muythaibxr

Trans control release pre-1.0.2 beta3

Postby jsmcortina » Sun Oct 08, 2017 7:49 am

Transmission Control pre-1.0.3 beta2 Firmware Released

To announce the pre-1.0.2 beta3 version of the Microsquirt trans code. This has had limited field testing, so proceed with caution.

Changes since the last release
Bug fixes
-

New / changed features
Extend race lockup settings to have more per-gear config.
41TE support disabled.
Add per-gear shift timers.
Rework gear choice, gear shifting and lockup behaviour.
Allow shift and lockup states and timers to be datalogged.
Add BTR option and output pattern code from Jamie Staltari.
Apply new line pressure partway through shift.
Rework testmode to integrate better.
Add on/off lockup mode.
Add low-range options.
Change to use output-shaft speed for shifts and lockup internally.
ini changes from PT to support output-shaft.

Minor / internal
Change to stdint typedefs.
Re-arrange code to use more functions and simplify main()

The zip can be downloaded via:
http://www.msextra.com/downloads/dev/trans/pre-1-0-2-beta3/
I can supply, repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk

My Success story: viewtopic.php?f=104&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
jsmcortina
Site Admin
 
Posts: 33917
Joined: Mon May 03, 2004 1:34 am
Location: Birmingham, UK

Re: Trans control release pre-1.0.2 beta3

Postby ashford » Mon Oct 09, 2017 7:46 am

much better. i have a few new settings that threw my old tune into stupid mode in a few places, ill sort them out before i comment on it.

what i have observed though is the the wait time for lockup is not being used on a shift, as soon as gear goes from 3 to 4 lockup starts immediately or observing the shorter wait for unlock. ill get the tune sorted better and post a log
ashford
Super MS/Extra'er
 
Posts: 1331
Joined: Sun Apr 27, 2008 4:29 pm

Re: Trans control release pre-1.0.2 beta3

Postby ashford » Tue Oct 10, 2017 7:40 pm

i have noticed a typo in ini or bug. delay before shift is associated with wrong gears. i was making changes in 1s and 2s and not noticing changes in upshift times so i inputted very different numbers in each. for the 1-2 shift it uses 3s and 2-3 shift it uses 4s. same with downshift assuming that
1-2 would use 2s
2-3 3s
3-4 4s
4-3 3s
3-2 2s and
2-1 1s

these were the settings
settinsforshift.png
settinsforshift.png (78.63 KiB) Viewed 482 times


this is the log
shift time test.msl.zip
(310.11 KiB) Downloaded 4 times



it seems to find the lighter thottle and downshift times i will need to enable lock tcc during shift.

tcc lock delay is not observed when doing a 3-4 shift it only seems to be observed during non-shifting engagement.
another thought i have is to use map for tcc engagement threshold as it take about full engine load at low speed to disengage at lower speeds using a setting that works well at higer speeds
ashford
Super MS/Extra'er
 
Posts: 1331
Joined: Sun Apr 27, 2008 4:29 pm

Re: Trans control release pre-1.0.2 beta3

Postby ashford » Thu Oct 12, 2017 8:03 pm

i am pleased with the shifting strategy now(other than the gear delay time association).
i think tcc needs a bit more depth in its tuning.
here is a long log with mixed driving. tcc off during shift in beginning and end, on in the middle. resets should give it away

EDIT: i thought 1meg was limit, will cut it down a bit
ashford
Super MS/Extra'er
 
Posts: 1331
Joined: Sun Apr 27, 2008 4:29 pm

Re: Trans control release pre-1.0.2 beta3

Postby ashford » Thu Oct 12, 2017 8:16 pm

chopped in half
Attachments
2017-10-12_15_mod1.csv.zip
(521.44 KiB) Downloaded 4 times
2017-10-12_15_mod2.csv.zip
(519.23 KiB) Downloaded 5 times
ashford
Super MS/Extra'er
 
Posts: 1331
Joined: Sun Apr 27, 2008 4:29 pm

Re: Trans control release pre-1.0.2 beta3

Postby jsmcortina » Sun Oct 15, 2017 9:26 am

Could you post your MSQ as well please.

James
I can supply, repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk

My Success story: viewtopic.php?f=104&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
jsmcortina
Site Admin
 
Posts: 33917
Joined: Mon May 03, 2004 1:34 am
Location: Birmingham, UK

Re: Trans control release pre-1.0.2 beta3

Postby ashford » Mon Oct 16, 2017 6:24 am

here it is at the end of the log. this could also be used as a base tune for a 4r70/75w as it is a stock trans. this trans also has an input speed sensor i don't know if it has any useful info or not.
Attachments
CAN1_CurrentTune.msq
(30.33 KiB) Downloaded 7 times
ashford
Super MS/Extra'er
 
Posts: 1331
Joined: Sun Apr 27, 2008 4:29 pm

Re: Trans control release pre-1.0.2 beta3

Postby knightrous » Mon Oct 16, 2017 10:46 pm

Can we please have the Toyota A340E/341E selector circuit updated in the documentation as the currently one has a flaw (outlined here: http://www.msextra.com/forums/viewtopic.php?p=513448#p513448)

Image
'89 Toyota MR2 AW11 - 1MZFE 3L V6 - Still building the car, DIYPNP ready for it, eventually MS3PRO!
'89 Toyota MR2 AW11 - 16V 4AGZE - MS2 waiting to be installed!
User avatar
knightrous
Experienced MS/Extra'er
 
Posts: 152
Joined: Tue Oct 06, 2009 4:58 pm
Location: NSW, Australia

Re: Trans control release pre-1.0.2 beta3

Postby slyrye » Mon Oct 23, 2017 6:40 pm

James if I may request please don't stop supporting Chrysler 41TE Tranny we are currently working on it using MS_Trans Controller... we jumped over from MegaShift since we are having trouble getting support for months now...!

Re: Attempting a Chrysler 41TE Trans Control With Microsquir

Sent: Mon Oct 23, 2017 4:59 pm
From: apalrd
To: slyrye
Well, yes and no. The pressure switches are used, but not exactly.

In each shift, there are two 'elements' (clutches) which are involved. One is released, one is applied. To shift from 1 to 2, the LR element is released and the 24 element is applied. This does not necessarily mean that one solenoid is energized and another is de-energized, since in this example the LR solenoid is de-energized and the 24 solenoid is also de-energized, since the 24 solenoid is inverted (turning on the solenoid releases the clutch). First, for each shift, you need to establish which element is the apply element and which is the release element. There are 10 valid combinations between forward gears. (1->2, 1->3, 2->3, 2->4, 3->4, 4->3, 4->2, 3->2, 3->1, 2->1). There are also 'garage shifts' (to/from neutral), and R-D rolling shifts (which skip neutral) which require additional software.

Mechanically, to get a 'good' shift, the total clutch capacity needs to equal the input torque. For example, if the engine produces 100 Nm of torque, the the sum of the torque capacities of the apply and release clutches needs to equal 100 Nm (if you know all of the geometry, you can back calculate the pressure required to get a given torque capacity).

During the shift, one solenoid is fired 'ballistically' (turned on/off at the right time) and the other is duty cycle controlled at 143hz to modulate the pressure on the clutch. The clutches all have a pressure accumulator which smooths pressure pulsations due to 143hz PWM. The solenoid flow rates and clutch volumes are modeled to calculate the pressure on each clutch (indirectly - Chrysler models the volume of fluid in each accumulator, hence the 'clutch volume index' for each clutch unique to each transmission). The solenoid which is fired ballistically is modeled as the pressure reduces or increases, and the other solenoid is PWMed to match the loss of torque capacity (so as the torque capacity falls off on the releasing clutch, the apply clutch is PWMed to match).

The shift is broken into four phases, two of which are closed loop. The order changes depending on the type of shift (up or down, and how much load).

For powered upshifts:
-Fill phase - apply element is filled to minimum volume so the clutch touches but does not transmit torque. This is done by turning the solenoid on 100% for a learned amount of time, then turning it to a low duty cycle
-Torque phase - Release element is vented ballistically, and apply element is PWMed to maintain the same gear ratio (both open loop based on the model, and closed loop by calculating the actual gear ratio from the input to output shaft). At the end of this phase, the release clutch is completely empty and the apply clutch is carrying 100% of the torque but slipping to maintain the old gear ratio. If the apply element cannot be filled fast enough (duty cycle reaches 100%), Chrysler controls will sometimes fire a 'bump along' pulse on the release element solenoid (where they apply for 10ms to slow down the vent rate)
-Speed phase - Release element is entirely released, so apply clutch duty cycle is increased to pull the transmission to the new gear ratio (by pulling down the turbine RPM). This is done with closed loop acceleration control. The accel rate target starts high and drops as the target speed is approached. The engine torque is reduced using spark retard during this phase only.
-Hold phase - Apply element is set to 100% duty to lock the clutch and the shift is completed

For powered downshifts (kickdowns):
-Fill phase - Release element is vented to the hold point. Apply clutch is not filled, but the time it takes to fill it is calculated
-Speed phase - Release element is vented with PWM control to allow the engine to accelerate (the transmission slips to the new gear ratio). As the speed approaches the target speed, the duty cycle on the release element is increased again to control the engine acceleration (to reach the new gear ratio, by slipping the release element). The apply element is fired sometime during the speed phase, so that it is filled when the speed phase ends.
-Torque phase - Apply element is filled ballistically, release element is vented to maintain the new gear ratio until it is fully vented
-Hold phase - Once the release element duty cycle reaches 0, the apply element should be fully engaged.

For unpowered upshifts (i.e. during fuel shutoff, such as shifting into OD):
-Fill phase - Release element is vented to the to the hold point
-Speed phase - Release element is vented with PWM control to allow the engine speed to fall to reach the target gear. This is the same as the speed phase of a kickdown, except the engine is decelerating.
-Torque phase - Apply element is energized. They may also PWM control this one to smooth the clutch lockup.
-Hold phase - Apply element is fully energized

For unpowered downshifts (i.e. while braking)
-Fill phase
-Torque phase
-Speed phase - similar to a powered upshift, except the engine is requested to increase torque (by exiting fuel shutoff) to accelerate to the target speed
-Hold phase


As you can see, the software is not trivial. The MS Trans hardware should be more than capable of the control, as I believe there are 4 PWM outputs with sufficient current.

apalrd
MS/Extra Newbie

Posts: 14
Joined: Sun Sep 13, 2015 6:54 pm
Location: Flint, Michigan
MSnS'n'4G15
slyrye
Experienced MS/Extra'er
 
Posts: 207
Joined: Sun Aug 12, 2007 2:41 pm
Location: Philippines

Re: Trans control release pre-1.0.2 beta3

Postby jsmcortina » Mon Oct 23, 2017 11:40 pm

The resources aren't available to support such a niche market at this time. To adequately support this trans would require days or weeks of testing too.

James
I can supply, repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk

My Success story: viewtopic.php?f=104&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
jsmcortina
Site Admin
 
Posts: 33917
Joined: Mon May 03, 2004 1:34 am
Location: Birmingham, UK

Re: Trans control release pre-1.0.2 beta3

Postby slyrye » Fri Oct 27, 2017 10:46 pm

jsmcortina wrote:The resources aren't available to support such a niche market at this time. To adequately support this trans would require days or weeks of testing too.

James


Alrightee then .. Are the solenoid outputs from the Previous releases still logically functional?
MSnS'n'4G15
slyrye
Experienced MS/Extra'er
 
Posts: 207
Joined: Sun Aug 12, 2007 2:41 pm
Location: Philippines

Re: Trans control release pre-1.0.2 beta3

Postby jamies » Sat Oct 28, 2017 1:18 am

Hi James,
I only just saw this thread,
I will bench test the BTR stuff in the next week or so, when time permits
jamies
Experienced MS/Extra'er
 
Posts: 339
Joined: Tue Apr 07, 2009 12:22 am

Re: Trans control release pre-1.0.2 beta3

Postby kikkegek » Thu Nov 23, 2017 11:26 am

I have a GM coolant sensor connected and it will not read it. CLT is stuck at 77C. I disconnected the sensor and reading didnt change. I know there is no temperature influence at thismoment of coolant or trans temp. But does this firmware support reading of the CLT sensor?

2017-11-20_15.55.16.trip2schiphol.msq
(30.35 KiB) Downloaded 2 times


log file is too big. sorry
kikkegek
Helpful MS/Extra'er
 
Posts: 37
Joined: Thu Jan 14, 2016 12:14 am
Location: Netherlands, The Hague

Re: Trans control release pre-1.0.2 beta3

Postby jsmcortina » Sun Nov 26, 2017 8:17 am

Coolant temperature is not currently used by the trans control code.

James
I can supply, repair or upgrade Megasquirts in UK. http://www.jamesmurrayengineering.co.uk

My Success story: viewtopic.php?f=104&t=34277
MSEXTRA documentation at: http://www.msextra.com/doc/index.html
jsmcortina
Site Admin
 
Posts: 33917
Joined: Mon May 03, 2004 1:34 am
Location: Birmingham, UK

Re: Trans control release pre-1.0.2 beta3

Postby kikkegek » Sun Nov 26, 2017 12:34 pm

jsmcortina wrote:Coolant temperature is not currently used by the trans control code.

James


OK, but it shows in 3.3.4 how to wire it, in 6.4.7 that Megalogviewer van read the value and its in the wiring table in 4.1.........so it suggests that its functional, at least a readout.

So maybe make a more clear statement about function in the manual?

edit: 11-29-2017
the reason I am asking is because I just run the Microsquirt for trans control and have a carb for fuel and HEI ignition. And I want to use a Android tablet for dashboard. would be awesome if I could get a CLT reading in there.
kikkegek
Helpful MS/Extra'er
 
Posts: 37
Joined: Thu Jan 14, 2016 12:14 am
Location: Netherlands, The Hague

Re: Trans control release pre-1.0.2 beta3

Postby kikkegek » Wed Nov 29, 2017 1:15 pm

I found some minor text inconsistencies in the firmware and software:

My trans is of 1996. And I found that they changed the frequency in 94 from 292 to 614Hz . Funny thing is that you can only select 290 or 590 Hz in Tuner Studio MS and the HELP BALLOON in the firmware shows "later model 597Hz (most common) and early used 297Hz. And a 4L80E manual I found shows a frequency of 292.5Hz for the Force motor.

I did some more testdriving tonight.

raising the linepressure up toe a minimum of 50% and more for all gears, results in exactly the same THUD in 1-2 and 3-4 shift and 2-3 shift stays smooth.

then I went to a seperate line pressure table for each gear and lowered the 1st gear pressures a lot. only 7% at low loads and maximum of 50% at full load. I gave all other gears a minimum line pressure of 50% at low loads up to 100% line pressure at full load. 1-2 and 3-4 shift will still have a THUD and 2-3 will be smooth.

Kinda lost here why different line pressure dont seem to have any noticeable influence in shift firmness

any thoughts?
kikkegek
Helpful MS/Extra'er
 
Posts: 37
Joined: Thu Jan 14, 2016 12:14 am
Location: Netherlands, The Hague


Return to Transmission Control

Who is online

Users browsing this forum: No registered users and 1 guest