Transmission Control logic

It's all about the code!
Trazique
Posts: 12
Joined: Wed Aug 06, 2014 6:17 pm
Location: Jacksonville Fl

Re: Transmission Control logic

Post by Trazique »

Shift Solenoid Current Results for the Mercedes, Jaguar and Chrysler 722.6 transmission.

Correction. the pic I posted earlier is not my TCU, no wonder I wasnt getting correct readingd.

My TCU is located in the cabin on the passenger side foot well. Much better because I was able to look at the meter while driving.

Most other Mercedes TCU is under the hood.

Confession #2; I didn't get a meter like the CA-60, which I'm sure would have been much better that the CEN-TECH that I found last minute locally.

However here are the results of the shift currents for the 722.6: Please see page 27 of this PDF: http://shop.ukrtrans.biz/wp-content/uploads/catalogs/722.6.pdf

Shift Solenoid 1-2 amperage reading @ Park or Idle in D: 0.00A
Shift Solenoid 1-2 amperage reading @ Shift/load: 0.11A

Shift Solenoid 2-3 amperage reading @ Park or Idle in D: 0.00A
Shift Solenoid 2-3 amperage reading @ Shift/Load: 0.16A

Shift Solenoid 3-4 amperage reading @ Park or Idle in D: 0.04A
Shift Solenoid 3-4 amperage reading @ Shift/Load: 0.16A

Shift Solenoid 4-5 amperage reading @ Park or Idle in D: 0.00A
Shift Solenoid 4-5 amperage reading @ Shift/Load: 0.10A

As for the Shift Pressure Solenoid it had a constant of 0.04-6A @ light throttle and drops to 0.00 between shifts.

I would do these tests again with better isolation of the wires with the clamp, or maybe try a different meter.

Let e know what you think.
Attachments
20161105_082147.jpg
20161105_082147.jpg (8.13 MiB) Viewed 25184 times
Last edited by Trazique on Sat Nov 05, 2016 12:33 pm, edited 1 time in total.
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Transmission Conrol logic

Post by kb1gtt »

Large clamps like that are commonly less accurate than the smaller clamps. I don't know why, but the MFG's seem to think that larger amp numbers is a good thing. However when was the last time you had to read 600 amps. I can't think of a time when I had to read more than 60 amps. Also these larger clamps tend to be better at 60Hz than at higher frequencies. The CA-60 is good until about 20kHz, then it's response starts to trail off above that.

With the meter can you verify the ohms of the solenoid?

I see the manual claims it ranges from full off to full on, with PWM for the intermediate points. I'm wondering what the PWM period is. Do you have a scope? If so can we measure the voltage pulses to get the PWM period.
Welcome to the friendlier side of internet crazy :)
Trazique
Posts: 12
Joined: Wed Aug 06, 2014 6:17 pm
Location: Jacksonville Fl

Re: Transmission Conrol logic

Post by Trazique »

kb1gtt wrote:Large clamps like that are commonly less accurate than the smaller clamps. I don't know why, but the MFG's seem to think that larger amp numbers is a good thing. However when was the last time you had to read 600 amps. I can't think of a time when I had to read more than 60 amps. Also these larger clamps tend to be better at 60Hz than at higher frequencies. The CA-60 is good until about 20kHz, then it's response starts to trail off above that.

With the meter can you verify the ohms of the solenoid?

I see the manual claims it ranges from full off to full on, with PWM for the intermediate points. I'm wondering what the PWM period is. Do you have a scope? If so can we measure the voltage pulses to get the PWM period.

I see..

Ok i will check it up.
Trazique
Posts: 12
Joined: Wed Aug 06, 2014 6:17 pm
Location: Jacksonville Fl

Re: Transmission Conrol logic

Post by Trazique »

I picked up a slightly smaller clamp today. Still couldnt find one as small as the CA50 here locally. However It is able to test Hz%. I have not had a chance to test fir PWM as yet

But i briefly tested the current on 2 of the shift solenoids and saw 0.06A on gear shift 2-3 and 4-5. Dont mark it as official as yet... More to come.
Attachments
20161105_163748.jpg
20161105_163748.jpg (3.57 MiB) Viewed 25178 times
Trazique
Posts: 12
Joined: Wed Aug 06, 2014 6:17 pm
Location: Jacksonville Fl

Re: Transmission Conrol logic

Post by Trazique »

More Information on how the TCU for my 1999 Mercedes CLK430 controls shifting on the 722.6 5G (5speed) Transmission. Please see pages 5 and 7.

I believe it would be the same for Chrysler vehicles starting from 2001 that carries the 722.6, including Jeep wrangles.
Attachments
722.6 2006-04_04.pdf
(2.5 MiB) Downloaded 446 times
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Conrol logic

Post by AndreyB »

While I am curios about this, the more I read the more I realize that realistically I would probably not have to dive into this until I win a lottery :(
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Trazique
Posts: 12
Joined: Wed Aug 06, 2014 6:17 pm
Location: Jacksonville Fl

Re: Transmission Conrol logic

Post by Trazique »

Lol, well it is a lot on information.

I know of two different TCU that were built using Arduinos to control these transmissions, but their prices are outragious!

ofgear.dk and i think the other is PCS.

I was excited to see you ask the question about "what do we know about Transmission logic" because i followed you since hackaday when you showed your first ECU.
That excitement excitement led me to do some research on my 722.6 transmission.
Kb1ggt suggested that i get amperage readings on the shift solenoids, I tried with clamp meters, but they were inaccurate.

So I decided to break the wires leading to each shift solenoid and placed my meter between the wires.

I have consistently gotten a reading of 0.50 - 0.54 Amps per solenoid on each up shift and down shift from 1st to 5th gear and back down.
The Modulating pressure solenoid drew a consistant 0.81 Amp (i am still monitoring that one to see what it would read at WOT)

Tomorrow I would continue the testing and het dutycycle readings also.
Trazique
Posts: 12
Joined: Wed Aug 06, 2014 6:17 pm
Location: Jacksonville Fl

Re: Transmission Conrol logic

Post by Trazique »

From reading the Mercedes factory info the 722.6 engages only when the shift is requested (it uses PWM for these solenoids s well) and "ratchets" through the gears rather than speed 1 Solenoid, speed 2 solenoid....... or even a combo.

The 1-2 shift solenoid is also used for the 4-5 shift.
The shift sequences are initiated and pressure are through the pwm solenoids. Oncea shift is completed everything returns to a rest state.

Commanding shifts is fairly straight forward but it must be in sequential and because of this things must be setup properly.

So the coding is where you guys come in :lol: :lol: :D .
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Conrol logic

Post by AndreyB »

Here's what works for @porelmundo

That's Eagle board + Android sketch.
Attachments
TCU_source_code.zip
(3.13 KiB) Downloaded 419 times
TCU_Eagle.zip
(58.81 KiB) Downloaded 391 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Conrol logic

Post by AndreyB »

https://github.com/rusefi/rusefi/issues/1454 is the epic for rusEFI build-in TCU

next story to develop could be https://github.com/rusefi/rusefi/issues/1597
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
porelmundo
contributor
contributor
Posts: 87
Joined: Tue Apr 26, 2016 1:15 am

Re: Transmission Conrol logic

Post by porelmundo »

I dont think this will be to much stress for processor, TCU's are simple operations.
What i learn trough the years is that they all work the same no mater wish brand, some of them have more solenoids some less,
and we have all we need for this already build on the hardware, with all the spare outputs i think this will be the next challenge and the best is that i know how to make it possible with your help.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Conrol logic

Post by AndreyB »

It looks like my Passat B6 has a 6 speed auto 09G Tiptronic with external TCU. Just saying :) (not my car on this picture)
Attachments
015.jpg
015.jpg (41.25 KiB) Viewed 23265 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Conrol logic

Post by AndreyB »

Status update: we are slowly moving forward with TCU, at the moment 2003 Mazda Miata is the most probably first test mule.

At this point discussing too many different transmissions is probably a distraction. There is always https://www.patreon.com/rusefi :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
dbh97
contributor
contributor
Posts: 84
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: Transmission Conrol logic

Post by dbh97 »

Just posting this here for future reference:

http://www.efidynotuning.com/trans.htm
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Conrol logic

Post by AndreyB »

David, in your opinion are we ready for manual operation of Passat B6 transmission Aisin TF-60SN code 09G?

https://github.com/rusefi/rusefi/wiki/oem_docs/VW/2006_Passat/automatic_transmission_09G_851503.pdf
image.png
image.png (128.38 KiB) Viewed 22570 times
Attachments
B6_TCU_connector_52_pin.jpg
B6_TCU_connector_52_pin.jpg (105.31 KiB) Viewed 22570 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
dbh97
contributor
contributor
Posts: 84
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: Transmission Conrol logic

Post by dbh97 »

Not even close, unfortunately. In addition to 2 on/off solenoids, there are also 6 pressure control valves that have to be operated (and tuned) somewhere on a linear scale, not just on or off. Getting this wrong could toast clutches, or worse.
First we need to figure out how to control a single valve reliably, which many transmissions have as the main pressure control, and we can get more wrong without ruining a transmission.

Copied from Slack:
Matthew Kennedy wrote: imagine you had two clutch pedals
you have to apply one while you release the other
older automatics work differently: each shift only requires applying OR releasing EXACTLY ONE clutch
clutch-to-clutch boxes require manipulating (slipping) multiple clutches at the same time
oh and if you engage the next clutch before the previous one has disengaged, it either:
  1. vaporizes the clutches
  2. locks up the gearbox (aka tire smoke machine)
And lots of closed loop
Based on shift time
And how hard the shift was
For example I think the 6l80e includes in its model the moments of inertia of some internal parts
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Conrol logic

Post by AndreyB »

Attachments
Aisin-Warner-09G-Electronics-Anatomy-of-Shift-Controls-Rostra-Powertrain-Controls.pdf
(2.14 MiB) Downloaded 322 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Conrol logic

Post by AndreyB »

Status update: @dbh97 has implemented all the code to switch basic solenoids using upshift/downshift wheel paddles - i.e. we paddle buttons input with debounce, have solenoid activation matrix and designated solenoid outputs.

As of today this code was never tested on a real vehicle, looking for a beta tested to collaborate with David @dbh97 via https://app.slack.com/client/T4NJHQ8QZ/C011D96LHT7

I do not remember if we have enough automated test coverage of this.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Control logic

Post by AndreyB »

KAIJU has Re5r05a
image.png
image.png (738.35 KiB) Viewed 21080 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Control logic

Post by AndreyB »

Jatco re4f03a
image.png
image.png (650.85 KiB) Viewed 21082 times
re5r05a
image.png
image.png (602.62 KiB) Viewed 21082 times
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Control logic

Post by AndreyB »

Kudos David

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Bluepower
Posts: 13
Joined: Mon Dec 23, 2019 4:42 pm

Re: Transmission Control logic

Post by Bluepower »

That looks promising!!
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Control logic

Post by AndreyB »

Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
dbh97
contributor
contributor
Posts: 84
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: Transmission Control logic

Post by dbh97 »

I'm putting some time into developing this again, under pressure from @nmstec.

Right now I'm trying to understand the control strategy for the TCC. I've asked @blundar for some screenshots of OE tunes, hopefully that will provide some insight. @nmstec posted a screenshot in Slack of a few tables from the SUPERMATIC transmission controller from Chevrolet Performance Parts. What I'm seeing there doesn't quite add up to or fully explain what I read in the factory service manual, so I'm hoping the OE TCM tune will shed a little more light.
image.png
image.png (130.6 KiB) Viewed 13973 times
Here's a link I found with a little information:
https://www.sonnax.com/tech_resources/214-20-years-later-what-s-new-with-4l60-e-pwm-modifications
https://www.schaeffler.com/remotemedien/media/_shared_media/08_media_library/01_publications/schaeffler_2/symposia_1/downloads_11/6_Torque_Converter_1.pdf
dbh97
contributor
contributor
Posts: 84
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: Transmission Control logic

Post by dbh97 »

@blundar doesn't have time/motivation to get me screenshots, but I found some that seem helpful:
https://ls1tech.com/forums/automatic-transmission/1416477-how-4l60-65e-trans-tuning-shifting-tcc-tm-w-pics.html
This makes a lot more sense than the SUPERMATIC controller tables, and mostly agrees with the factory service manual.
What I'm still missing regarding torque converter control is exactly how the PWM LOF (lock up feel) solenoid is controlled and tuned, but I think I can ignore that for now and implement what I do know.
dbh97
contributor
contributor
Posts: 84
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: Transmission Control logic

Post by dbh97 »

There are still a few things I don't know about how the TCC is controlled.
The service manual states, regarding the 4-3 downshift: The PCM de-energizes the TCC solenoid valve, and operates the duty cycle of the TCC PWM solenoid valve to release the converter clutch for a smooth disengagement, prior to initiating the 4-3 downshift.
How quickly does it back off the PWM? I we disengage quickly rather than using PWM, how long should we wait before shifting? I suppose I'll just make these tunable fields and let @nmstec worry about it.
Also, I can't find anything in the service manual regarding TCC unlock for the 3-4 shift. I assume there must be states where the TCC is locked in 3rd gear when the 3-4 shift is demanded.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Control logic

Post by AndreyB »

Excited to see fresh activity in this thread!
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
dbh97
contributor
contributor
Posts: 84
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: Transmission Control logic

Post by dbh97 »

I need some ideas for generalizing the configuration interface for transmission control solenoids. Currently I'm looking at the GM 4LXX series - 4L60E, 4L65E, 4L70E, 4L80E, 4L85E.
Control on these transmissions has a lot in common, but also quite a few differences. Currently I'm focusing on the 4L60E, and even it has changes between years.
At some point, I think at the 1995-1996 model year split, a PWM TCC apply solenoid was added in addition to the on/off TCC solenoid. It was removed again in the 4L70E.
Also changed, I think at the same time, was the PWM 3-2 shift solenoid was replaced with an on/off solenoid. From what I can tell, the PWM variation has a tuning curve associated with it.
Also, an input shaft speed sensor was added at some point. Having an input shaft speed sensor would let the TCU measure the shift speed, which would let us set the desired shift speed and let the TCU automatically tune the line pressure. If we don't have this, I suppose the shifting line pressure will have to be manually tuned by feel. This is also the case with my E4OD.

So basically, there will be a lot of fields and tables in Tunerstudio that are specific to a particular model of transmission, and even a certain range of years of a particular transmission. I wish there was a way to hide fields based on a dropdown selection in addition to just disabling them.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: Transmission Control logic

Post by AndreyB »

dbh97 wrote:
Fri Apr 15, 2022 3:20 am
I wish there was a way to hide fields based on a dropdown selection in addition to just disabling them.
and I believe there is such a way! We use it on trigger dialog for number of skipped tooth and to change trigger input control name

Code: Select all

		field = "Trigger type",							trigger_type
		field = "Total tooth count",					trigger_customTotalToothCount, {trigger_type == @@trigger_type_e_TT_TOOTHED_WHEEL@@}, {trigger_type == @@trigger_type_e_TT_TOOTHED_WHEEL@@}
		field = "Missing/skipped tooth count",					trigger_customSkippedToothCount, {trigger_type == @@trigger_type_e_TT_TOOTHED_WHEEL@@}, {trigger_type == @@trigger_type_e_TT_TOOTHED_WHEEL@@}



		field = "Cam Sensor (Primary channel)", 		triggerInputPins1, 1, {   twoStroke == 1 || ( !(@@TRIGGER_CRANK_BASED@@) && ( ( (@@TRIGGER_TYPE_WITHOUT_KNOWN_LOCATION@@) && skippedWheelOnCam) || !(@@TRIGGER_TYPE_WITHOUT_KNOWN_LOCATION@@)  ) )  }
		field = "Crank Sensor (Primary channel)", 		triggerInputPins1, 1, { !(twoStroke == 1 || ( !(@@TRIGGER_CRANK_BASED@@) && ( ( (@@TRIGGER_TYPE_WITHOUT_KNOWN_LOCATION@@) && skippedWheelOnCam) || !(@@TRIGGER_TYPE_WITHOUT_KNOWN_LOCATION@@)  ) )) }
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
dbh97
contributor
contributor
Posts: 84
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: Transmission Control logic

Post by dbh97 »

Well, that kinda works. The dialog stays the same size as if the fields were actually there.
image.png
image.png (26.65 KiB) Viewed 13589 times
Post Reply