Page 7 of 13

Re: Microsquirt/MS2 based transmission controller

Posted: Fri Aug 13, 2010 12:01 am
by UnaClocker
It's all thanks to James. He's done an excellent job of getting support for the Chrysler 41te added to his firmware. In the process, he's made the code better able to be adapted to other transmissions as well. Soon the firmware will have a VSS output, paddle shifter support, plenty more, I'm sure. I definitely appreciate the time James has put into this.

Re: Microsquirt/MS2 based transmission controller

Posted: Fri Aug 13, 2010 4:19 pm
by devojet
Great to see james's code in action. I have been playing with it since james released the first version. I still haven't finished building my hot rod yet but I do have the gpio already set up to run the gearbox. Are you using this code on a GPIO?

Cheers

Daniel.

Re: Microsquirt/MS2 based transmission controller

Posted: Fri Aug 13, 2010 4:41 pm
by UnaClocker
Yup, using a GPIO. I'm going to write up a guide on setting up a GPIO for use with the Chrysler 41te and James's code when we get all the features in place. :)

Re: Microsquirt/MS2 based transmission controller

Posted: Tue Aug 24, 2010 1:07 pm
by slow67
I'm not totally new to Megasquirt (I've built/tuned one on a friends car, but its been ~ 2 years). I'm used to tuning GMs with HPTuners (which brings most of my questions/comments).

What is everyone's opinions about the following questions:
1. Why not use 1 big table for upshift/downshifts (would look similar to a map table, with X being TPS and Y being the particular upshift or downshift line)? My thought is it would be a little easier, as it would put all your up/downshift tables in 1 bigger table. Also what about a little bigger table (perhaps use the existing MAP table size that is used in MS2/MS2 extra?

2. Why not command the TCC by gear and mph/tps instead of just purely TPS/MPH? I'm asking because I have my 4L80E setup where if I'm in manual 3rd, it will lock the converter, but if I'm in manual 4th, it will not lock in 3rd (my 3rd gear lockup MPH is set higher than my 3>4 shift point). I do this so it will lockup cruising around at ~50 mph, as to not build heat. You could also get rid of the "Race" mode by just simply programming it into the TCC tables.

3. What about having a "shift-lock" option ( a simple yes/no) for locking the converter during a gearshift? I noticed this is in "Race" mode, but can also be handy for part throttle shifts as well (in my opinion).

4. What about having a 2nd shift table (GM calls it "Performance Mode", aka Tow-Haul)? What about manual hold gear option? (if the trans is in manual 1st, it stays in manual first, and if in manual 2nd, it stays in manual 2nd, etc)? Maybe use the MS2's table switching and have either a manual mode (where the TCM commands the gear that the shifter is in) or a 2nd shift table (maybe have an either/or) and 2nd TCC table?

5. I can make the 4L80E a 6 speed (and a 2.75 gearset isn't all that expensive). How hard would it be to modify the code for this (a 3rd solenoid for the OD section, would shift like this: 1, 1 OD, 2, 2 OD, 3, 3 OD, etc assuming the shifter is in the "D4" position)? I know the hardware is capable, as there is provisions for a 3-2 solenoid on the 60Es. Perhaps a cheaper alternative to TCI's "6L80E"!

BTW I am very well versed in 4L80E hydraulics (I helped design the Jake's Performance Transbrake) if anyone has any questions on 4L80Es.


Edit: I attached a sample table of what I'm envisioning for a shift/TCC table. 255/256 is pretty much considered maxed out, as it will likely never be reached.

Re: Microsquirt/MS2 based transmission controller

Posted: Tue Aug 24, 2010 2:33 pm
by UnaClocker
slow67 wrote:4. What about manual hold gear option? (if the trans is in manual 1st, it stays in manual first, and if in manual 2nd, it stays in manual 2nd, etc)?
James is actually adding a paddle shift mode to the code as we speak. That should cover this feature pretty well. It will ONLY be in the gear you paddle shift into, nothing short of maybe a calculation to prevent a down-shift over-rev should override the paddle shifter/manual mode.

Re: Microsquirt/MS2 based transmission controller

Posted: Tue Aug 24, 2010 2:48 pm
by jsmcortina
Manual hold should already be there. If the selector is in 2nd it won't shift up out of 2nd.

James

Re: Microsquirt/MS2 based transmission controller

Posted: Tue Aug 24, 2010 3:02 pm
by slow67
UnaClocker wrote:
slow67 wrote:4. What about manual hold gear option? (if the trans is in manual 1st, it stays in manual first, and if in manual 2nd, it stays in manual 2nd, etc)?
James is actually adding a paddle shift mode to the code as we speak. That should cover this feature pretty well. It will ONLY be in the gear you paddle shift into, nothing short of maybe a calculation to prevent a down-shift over-rev should override the paddle shifter/manual mode.
One downfall of a paddle shifter (in a higher-hp application) is the over-run clutches aren't on (if in Gear Position D4) in gears 1-3. Also no engine braking (with a stock Valvebody you have to be in Manual 1st for engine braking in 1st gear, manual 2nd for 2nd gear, etc).

With no engine braking, you could let off the gas, the engine would go to near-idle, and once you hit the gas again (assuming WOT), it could blow up the overdrive roller clutch. It is feasible to build a valve body that has engine braking in 1st-3rd if you have the shifter in D3 (Doing it in gear position D4 is more difficult, but feasible as well).
jsmcortina wrote:Manual hold should already be there. If the selector is in 2nd it won't shift up out of 2nd.

James
That's excellent news! What about if you are at a stop and the shifter is in 2nd? 2nd gear start is a very nice feature. First, it lets you take off in 2nd in a low traction situation (snow, ice, mud, etc).
2nd, a proper way to do a burnout with a Th400/4L80E is to start off in 2nd (with the tires wet) and very quickly shift to 3rd, and let off the gas before the tires chirp. This is to protect both the low roller clutch and the intermediate sprag.

I'm collecting parts right now to build a 4L80E controller. This will go in a transmission test-mule car, and then possibly my own car. Looking to make the aftermarket a bit better for 4L80Es (more options, such as spragless 2nd, a 6 speed alternative to the PRICEY TCI one, etc).

Re: Microsquirt/MS2 based transmission controller

Posted: Tue Aug 24, 2010 3:05 pm
by devojet
Here are my comments,

1. Lance used the full table as you suggest in the megashift code for the GPIO. (see www.msgpio.com) I think James's implementation gives the same information but would use less memory to store the information.

4. I like the idea of a manual mode that follows the shifter input, but doesn't shift up or down ie if you select 2nd it wont shift up or down. This would be handy on a dyno when tuning the engine.

Re: Microsquirt/MS2 based transmission controller

Posted: Tue Aug 24, 2010 3:21 pm
by UnaClocker
slow67 wrote:One downfall of a paddle shifter (in a higher-hp application) is the over-run clutches aren't on (if in Gear Position D4) in gears 1-3. Also no engine braking (with a stock Valvebody you have to be in Manual 1st for engine braking in 1st gear, manual 2nd for 2nd gear, etc).

With no engine braking, you could let off the gas, the engine would go to near-idle, and once you hit the gas again (assuming WOT), it could blow up the overdrive roller clutch. It is feasible to build a valve body that has engine braking in 1st-3rd if you have the shifter in D3 (Doing it in gear position D4 is more difficult, but feasible as well).
Ahh, yeah, the Chrysler transmissions don't have that "problem" (for lack of a better word), our shifter/valve body really only selects forward/backward/neuteral, the 4 forward positions don't change anything hydraulic in the trans. So yeah, while the paddle shift is a solution for the Chrysler guys (at least this one model of Chrysler trans), I understand why it won't for the GM trans.

Re: Microsquirt/MS2 based transmission controller

Posted: Tue Aug 24, 2010 4:44 pm
by slow67
devojet wrote:Here are my comments,

1. Lance used the full table as you suggest in the megashift code for the GPIO. (see http://www.msgpio.com) I think James's implementation gives the same information but would use less memory to store the information.

4. I like the idea of a manual mode that follows the shifter input, but doesn't shift up or down ie if you select 2nd it wont shift up or down. This would be handy on a dyno when tuning the engine.
Agreed. I wouldn't think that memory would be too much of a restriction, as MS2Extra can have dual tables for map, spark, and boost. I was thinking you could have a choice of a 2nd table or manual mode (to save inputs and memory). Not sure if thats a requirement either though (but it would be nice to still be able to configure when the TCC locks up in manual mode as well). Its not hard to configure manual mode with a table though, as long as it has manual hold 1st and 2nd. If so 1st and 2nd is taken care of. For third, set it to upshift to third at 1 mph, downshift at 0. Do the same for 4th.

UnaClocker wrote:
slow67 wrote:One downfall of a paddle shifter (in a higher-hp application) is the over-run clutches aren't on (if in Gear Position D4) in gears 1-3. Also no engine braking (with a stock Valvebody you have to be in Manual 1st for engine braking in 1st gear, manual 2nd for 2nd gear, etc).

With no engine braking, you could let off the gas, the engine would go to near-idle, and once you hit the gas again (assuming WOT), it could blow up the overdrive roller clutch. It is feasible to build a valve body that has engine braking in 1st-3rd if you have the shifter in D3 (Doing it in gear position D4 is more difficult, but feasible as well).
Ahh, yeah, the Chrysler transmissions don't have that "problem" (for lack of a better word), our shifter/valve body really only selects forward/backward/neuteral, the 4 forward positions don't change anything hydraulic in the trans. So yeah, while the paddle shift is a solution for the Chrysler guys (at least this one model of Chrysler trans), I understand why it won't for the GM trans.
GM did it for 2 reasons:
1. So you don't have harsh downshifts when slowing down.
2. Its less complex to mearly engage another clutch than to dis-engage a clutch and engage another.

As long as you dont hit the sprag hard (too hard of a shift, or load/unload sticky tires under power) its fine.

Re: Microsquirt/MS2 based transmission controller

Posted: Wed Aug 25, 2010 7:00 am
by slow67
Just remembered, My table is bigger than it needs to be, as there is a hysteresis so the downshift tables/TCC unlock tables I posted aren't necessary :oops:

Re: Microsquirt/MS2 based transmission controller

Posted: Wed Aug 25, 2010 8:38 am
by UnaClocker
Yeah, there's a MPH hysteresis, and, at least for the Chrysler trans, there's a time hysteresis as well. (Wait at least .5 seconds (user settable) between shifts, to give the hydraulics time to react)

Re: Microsquirt/MS2 based transmission controller

Posted: Wed Aug 25, 2010 3:03 pm
by slow67
a few more questions....

Would it be easier hardware/codewise to just have the TCC on/off instead of PWM?

Has anyone thought of using the Map Sensor instead of TPS for load (would be much easier for retrofitting carbed vehicles)?

How is the engine coolant temp referenced in the code? Modern GMs don't really take into account ECT into any equations.

Any thoughts of adding a Trans temperature Bias table to adjust the PCS for trans fluid temp?

Re: Microsquirt/MS2 based transmission controller

Posted: Wed Aug 25, 2010 4:55 pm
by UnaClocker
slow67 wrote:a few more questions....

Would it be easier hardware/codewise to just have the TCC on/off instead of PWM?

Has anyone thought of using the Map Sensor instead of TPS for load (would be much easier for retrofitting carbed vehicles)?

How is the engine coolant temp referenced in the code? Modern GMs don't really take into account ECT into any equations.

Any thoughts of adding a Trans temperature Bias table to adjust the PCS for trans fluid temp?
I think the TCC just ramps up slowly to be easier on the clutch. With a high stall TC, there can be quite a jolt when it engages. While it might be simpler to not have that, it's already there, so it's simpler to just leave it there. ;) heh
I'm pretty sure you can use MAP or TPS for load already.
I don't think it currently does anything with the engine coolant temp, but it does use the trans fluid temp sensor (at least datalogs it, not sure if it has a table to make use of the data or any way to lock out problematic gears when the coolant is cold (I know older Chrysler EATX's had some issue with 4th when the fluid was below room temp, but that was back in ATF+2 days (or earlier), they fluid itself has come a long way since then).

Re: Microsquirt/MS2 based transmission controller

Posted: Wed Aug 25, 2010 5:13 pm
by slow67
UnaClocker wrote:I think the TCC just ramps up slowly to be easier on the clutch. With a high stall TC, there can be quite a jolt when it engages. While it might be simpler to not have that, it's already there, so it's simpler to just leave it there. ;) heh
I'm pretty sure you can use MAP or TPS for load already.
I (and many other tuners I know of) just bring on the TCC in an on/off fashion, as it increases clutch life. You just have to be more careful of when you bring it on (usually try to keep it 400 or less RPMs between locked/unlocked to keep it from shuddering). Also aftermarket TC Clutches will shudder badly if you ramp it in slowly (this is a big problem with the 6L80Es right now, as no-one has found the tables to modify the TCC PWM yet that I know of).

Good to know about the MAP, I didnt see anything about it in trans.ini (and switching from meters to inches for tire size doesn't appear to work either).

Re: Microsquirt/MS2 based transmission controller

Posted: Fri Aug 27, 2010 4:28 pm
by slow67
One note on line pressure on the 4L80E, the Line Pressure in reverse is useless, as the reverse boost valve controls pressure in reverse. Maybe replace it with a Neutral line pressure to control how hard the forwards come on? Or is it used by other transmissions?

Re: Microsquirt/MS2 based transmission controller

Posted: Sat Aug 28, 2010 11:05 am
by slow67
Quick question on the controller, are the directions based off of a normal MS2 build or an MS2Extra build?

Re: Microsquirt/MS2 based transmission controller

Posted: Wed Sep 22, 2010 8:27 am
by oysteinbno
Hi guys. I'm from Norway, and have build a couple of MS and hooked up to both Rotary, Ford v6(Cossworth) and US V8's. Started in 2005 with the first MS and have now a MSII 3.57 toghether with a relayboard that I use on my 36 Chevy.
I'm now going to use a GM 4L60E transmission, and I have read here at this forum that I can use the MS to controll the transmission.
But can I use the MSII 3.57 to controll both engine and transmission without the GPIO-board?
And is it best to use the GPIO board together with the MS.??

I've read : http://www.msextra.com/doc/trans/ms2-wiring.html
and I wonder if all those modifications is necessary for use MS as a transmission-controller??


c",)

Oystein Bardevik

Re: Microsquirt/MS2 based transmission controller

Posted: Wed Sep 22, 2010 9:31 am
by racingmini_mtl
You can't use the same MS device to control the engine and the transmission. So you can use your MS2 V3.57 setup to control the engine but you'll need to have another device to control the transmission. If you want to use the code discussed here you will either need another MS2 setup or a Microsquirt. Or you can use a GPIO with the other transmission code.

Jean

Re: Microsquirt/MS2 based transmission controller

Posted: Wed Sep 22, 2010 9:40 am
by Tjabo
UnaClocker is using a GPIO.