Paddle control and issues

[Turbofreakdotcom]

I broke this issue off into another post from "Toyota A340 control" since this is a universal topic.
On previous configuration with other trans controller, the paddles are hooked to the single cruise control wire from the cruise accel/coast stock buttons.
With cruise control power on, there is about 10 volts on the wire. Below are the values I get strait from the car:
Upshift button = 1.8 volt R=100 Ohm
Downshift button = 3.5 volt R=220 Ohm
No button pressed = 9.5 on wire.

With help from Daniel's (devojet), litneon, and James on this project, I used:
5 volt Zener diode to ground, 2.2 KOhm resistor to ground, and a 1 KOhm resistor on the 9.5-10 volt cruise control wire to limit the voltage to 4.9 volts to the TPS line (since TPS signal is CAN from MS2 ECU).
This takes the voltages to:
Upshift button = .8 volt R=100 Ohm
Downshift button = 1.5 volt R=220 Ohm
No button pressed = 4.9 on TPS wire from cruise line.

The issues
The down-shift and up-shift buttons are unpredictable which could shift gears up or down when either button pressed. So I changed the resistor combination in the paddles to give a larger voltage difference that gave me:
Upshift button = .8 volt R=100 Ohm
Downshift button = 3.8 volt R=680 Ohm
This helped down-shift to be much more predictable, but up-shift is still very bad by changing gears up and down unpredictably.

Changing [Trans controller setup] > [Shift settings] > "Minimum shift time" to .50 helps a little, but it starts getting too slow for me and doesn't help enough.

I found a "debounce" line in the "trans_init" C file set to "78" with a comment stating "10 ms debounce". When I finally figured out how to compile the code I change that value from "78" to "390" which, if I figured it out correctly, would change from 10 ms debounce to 50 ms debounce. Is this correct?
I didn't seem to make much difference,but before I go changing and flashing the CPU over and over again, I wanted to make sure I was on the right path with the debounce calculation and changes. Is there a limit to this debounce while troubleshooting, ex.. if I set "3900" would this be 500 ms and is this an acceptable change?

[Turbofreakdotcom]

Since downshifting worked great and up-shift unpredictable, I changed the resister combo on the paddles tonight.
Upshift button = 3.6 volt R=680 Ohm (resistor was on downshift paddle)
Downshift button = 2.0 volt R=220 Ohm

Upshift paddle will upshift now without it dropping down a gear at all. Downshift is now quirky. In my head, I assume 680 Ohm works so well at 3.6 volts, it is close to the 4.9 volts static voltage and 2.0 volts on downshift triggers upshifting a gear since it goes from 4.9 volts down past 3.6 volts to 2.0 volts and back up to 4.9 when released triggering upshift again.

I only have grounds going through the clock-spring to the cruise buttons, but from you circuit guys, can I change the resistors on my 5 volt limiting circuit to get a lower static voltage of 3.5 volts, leave one of the paddle resistors going to ground and have the other paddle resistor going up to 5 volts? The assumption is:
Downshift button = 2.0 volt R=220 Ohm
Cruise wire static voltage = 3.5
Upshift button = 4.5ish volts at R=?

Does it work that way?

[Turbofreakdotcom]

I feel silly answering my own stuff, but if it helps anyone else...... I removed the MS2 for testing changed the 1K resistor in the 5 volt limiting circuit to a 2K2 and made up a test switch for upshift with a 680 Ohm resistor to 5 volts which brought the voltages to:

Downshift button = 1.5 volt R=220 Ohm to ground
Cruise wire static voltage = 4.2 volts
Upshift button = 4.5 volts R=680 Ohm to 5 volts

With the TPS wire from MS2 back in the paddle circuit, this is working flawlessly! I messed with it for 5 minutes and couldn't make it shift wrong either way.
To get 5 volts into the steering wheel paddle area I will have to remove the older 3 wire clock-spring and install a later model 4 wire spring.

[Raymond_B]

Please do keep posting what you find! Thank you


Sent from my iPhone using Tapatalk

[Turbofreakdotcom]

Will do.
I may play with the resistor values to get to good even values on both sides of the static voltage to satisfy my OCD. Really, the only other thing I have to do is pull the 4 wire clock spring and steering column harness from the 89 donor car to bring the 5 volt line into steering wheel for the pull-up resistor on upshift and install in my 87 which will be a bit of labor. Unless, someone else has a better suggestion on what I can do to fix a solid circuit for the present wiring without bringing in the extra 5 volts for pull-up.
The other trans unit worked fine the way it was, but..... this last circuit with the pull-up seems to respond faster on the MS.

[Raymond_B]

A lot of this stuff is uncharted territory, especially the trans control it seems. So again, thanks for the information!

[Turbofreakdotcom]

On the 4th day I rest! After hours each day trying to figure out the best way to go about getting the manual shifting to work solid with the paddles, today I figured out how to get the voltage pull-up circuit in the steering wheel for the up-shift paddle. There was already a 12 volt wire on the clock-spring from the BRAKE tailight fuse.
With the circuit in the attached photo I was able to get these values on the TPS line (engine not running 12v BATT):
Downshift button = 1.5 volt R=220 Ohm to ground
Cruise wire static voltage = 4.0 volts
Upshift button = 4.5 volts (12v +BATT to pull-up circuit to cruise wire)



Here is a video for you to show to show how well it functions at this point:
https://www.youtube.com/watch?v=q6RYE17Fb0E

I apologize for the novice circuit drawing. The circuit was pieced together just seeing what worked using my voltmeter. If anyone has any advice, please let us know if this circuit could be simplified.

*EDIT:
Added a couple more photos inside the steering wheel just showing how I was able to fit the 1N4001 diode/680 Ohm resistor circuit in heat-shrink between plastic steering wheel rear trim and steering wheel metal spoke and another simply showing where I pulled the ground.

[dontz125]

Sticking my oar in late (you've already wired it up for 12v), but why not use the 5v Vref for your pull-up? Just on general principles, less to no chance of a wiring error slagging your processor ...

[Turbofreakdotcom]

Sticking my oar in late (you've already wired it up for 12v), but why not use the 5v Vref for your pull-up? Just on general principles, less to no chance of a wiring error slagging your processor ...

Thank you for the warning.
The stock cruise control wire is actually 9.5 volts in the steering wheel at the paddles so I used the 12 volt circuit to pull up the 9.5 volts to have higher voltage within my 5 volt limiting circuit. Seems convoluted, but it worked and I was fairly sure this was what I would have to do. This was really a plan to keep the stock cruise control circuit functional, though I do have fears if the zener diode and or resistor circuit goes south it would damage my MS2.

For MS2 safety purposes, I will re-design this today using all MS2 5 volt reference for paddle control. Not to mention how more controllable the 5 volt circuit is since it is regulated whether engine is running or not. This will cause me to cut the cruise control wire from the cruise computer and cut the 12 volt BRAKE line from the BRAKE fuse in order to re-route the MS2 wires through the clock-spring for the paddles.
This will mean eliminating the cruise control, but, in theory, could be added back using some relays to switch the clock-spring wire back and forth from the MS2 to the cruise computer and BRAKE line. I am not using cruise control now, though I had future plans to get it working again so I will not be going over the relay circuit to retain cruise control.

I will still be using the same principal dividing the circuit for a center voltage of around 3-4 volts, close to 5 volts for up-shift and around 2 volts for down-shift.

[Turbofreakdotcom]

I was trying to keep from swapping to the 4 wire clock-spring and column harness, but there is no way around it If I run the 5 volt reference. Misreading the diagrams, the 12 volts is actually for the horn which I thought was a simple ground circuit. I will revisit this thread when I pull the parts from the donor car and install. I'll have to keep the current configuration for now and live with the chance of a failure.

Will it only be the TPS ADC on CPU input that burns up if the TPS line gets 10 to 12 volts to it or will other components take on damage?

[Turbofreakdotcom]

Can you share which part of the code is responsible for shifting gears while holding down the paddle? I would like to remove that function.

[Turbofreakdotcom]

James,
I'm working on deciphering the code today to the best of my ability. Is this the main part that would cause MS to continue shifting while holding down the button?

Code: Select all

                        if ((pad_mode == 1) && (outpc.gear < max_gear)) {
                            outpc.gear++;
                            paddle_timer = 3906; // 500ms timeout/re-shift
                            paddle_phase = 1;
                        }

                        if ((pad_mode == 2) && (outpc.gear > 1)) {
                            outpc.gear--;
                            paddle_timer = 3906;
                            paddle_phase = 2;
                        }

[slow_hemi6]

Looks like there is a bit of room inside the wheel. if you have a switched +V (Horn is often not switched) and ground available then why did you not just use a 5v regulator and a couple of capacitors to have your own 5v supply inside the wheel. No clock spring changes then. A lm7805 or the nice low drop out lm2940 MS uses.

[Turbofreakdotcom]

The 12 volt is constant, not switched. In the clock-spring:
switched cruise wire - 9.5 v when cruise button on
a center terminal which seems to be a ground
and the un-switched 12 volt for horn
If the 12 volts were switched that is a great idea. Though, it can be at the cost of the horn being inactive at key off.

When I figure out more of the code I can configure it to better handle the pull down/pull down one wire setup.

[slow_hemi6]

lm7805 generally works fine with supply a little over 7v and a lm2940 works well with 6v supply. Would depend on if your CC wire can hold up that supply voltage under a little bit of load.

[Turbofreakdotcom]

Doesn't seem like much load for 16-18 gauge wire unless the wires of the clock-spring is an issue on a load. If I use the regulator powered by the 9.5 volt cruise wire, wouldn't I still need another wire going through to the MS?

[dontz125]

Load will be trivial; these are signals to an ECU, not driver pulses to a big honkin' fuel pump.

[slow_hemi6]

All seems too complicated how it is. The tps input has got a series 1K resistor in its circuit internally. The reg should keep the voltage steady. Roughly about 4.6v upshift, 2.5 resting and .4v downshift. The gang can debate this if they see a problem.

WIN_20160420_114618.JPG

[slow_hemi6]

Load will be trivial; these are signals to an ECU, not driver pulses to a big honkin' fuel pump.

My point was we don't know what is upstream of that CC wire and so don't know what is causing it to be 9.5v. Whatever IS upsteam may not be able to supply the reg without becoming a voltage drop. You have to try these things and see.

[slow_hemi6]

wouldn't I still need another wire going through to the MS?

You must be using something to get your variable voltage back to the MS already. You would use the same wire.