After doing some more thinking and research in my Haynes manual, I located the wire for the TCC solenoid at the transmission end this weekend. To make sure I located the right one, I wired up an LED and pullup resistor to a pin and ground lug, pushed the pin through the wire insulation and routed the LED through an opening in the left fender so it's visible through the windshield.
I finally drove the car this morning, and it worked perfectly -- no harm done, and the LED lights up when the TCC locks up.
I confirmed on the way to work that the TCC unlocks whenever there's no pressure on the gas pedal (it's apparently released by the brake switch on some other cars.) I still need to confirm that the TCC doesn't lock in 3rd, which I'll do on the way home.
The next step would be to splice in a rectifier diode in series with a wire on either side. Upstream of the diode would go to the LED; downstream would go to a momentary-contact switch on the shifter, and from there to 12V. This should allow me to manually lock up in 3rd in the ~27-39 MPH range or 4th in the 39-45 MPH range.
I'll need to confirm that the PCM is sending 12V and not 5V before wiring in the switch; I should be able to do this using the switch lead and a DMM.
Most of the speed limits around here on secondaries are in the 25-35 MPH range, so it should buy at least an extra 1-2 MPG. It'd make the biggest improvement in FE when climbing hills at 30-35 MPH. I also wouldn't need to pulse to 45 MPH to lock up the TCC for cruising at 40, which should reduce the risk of a speeding ticket.
The PCM limits the maximum gear to second and third below water temps of 101°F and 137°F respectively, so using the TCC switch could also help efficiency during warm-up.
The LED upstream of the rectifier would let me know when the PCM is sending a lockup signal so I could let off the switch at highway speeds.
I expect keeping the TCC locked while shifting would be detrimental to the powertrain's reliability, so I'd plan to limit the speed, throttle and gear selection to make a shift unlikely while the TCC is locked. Putting the switch on the gearshift should help avoid an accidental shift in lockup, since I'll need to let it go to change gears.
I'm looking forward to installing this and trying this out. It should add a whole new dimension of hypermiling fun to my commute, while still leaving the car driveable for anybody else.
I tried keeping it in 3rd up to 45MPH on the way home last night...the PCM locked the TCC around 42 MPH, maybe 1-2 MPH less than when it'd normally lock up in 4th. The PCM lockup signal appears to be a function of engine load (at a given RPM) and speed, regardless of the gear selection -- I've occasionally gotten it to lock up in 4th at 42 MPH when I really feather the accelerator.
It would definitely benefit from manual control. I have a friend with a Saturn sedan; his AT locks up in 3rd at low speeds when 3rd is manually selected, so it should be okay on this car.
Given csrmel's thread on the Chrysler a604, I'll definitely need to check the voltage/current output with a DMM before I start hooking up an override switch.
I finally hacked into the wiring last night. I hated to splice into the OEM harness, especially given it's location, but it's the only way I had to accomplish what I wanted. It's my car, darn it, and the warranty is long gone anyway.
The connector for the TCC wire is right behind a gap between the splash guard and left fender, forward of the front wheel -- perfect for getting hosed down with water. I cut the wire, spliced a long piece of ~12 AWG speaker cable onto the ends (to minimize added impedance), shrunk some heatshrink over the solder joints, ziptied the speaker cable to the harness and wrapped everything with electrical tape. Hopefully, it'll be good enough. I may go back later and dab some RTV onto the ends of the heatshrink to ensure they're sealed, but I don't want to redo the solder joints -- there's not much free wire on the harness to work with, and it's in a tight location.
For now, I've routed the speaker cable along the fender, through a gap between the fender and door pillar, along the rocker panel and under the driver's seat to the passenger side. I'll probably eventually route it under the dash to the shifter instead.
I twisted the ends of the speaker cable together, hooked up an analog voltmeter and the TCC LED from last week to the joint, and connected the LED and voltmeter to ground.
As of this morning's commute, everything is still working as expected. At lockup, the LED comes on and the voltmeter is indicating alternator voltage (~12-15V), so everything is still working and now I can see that the computer is feeding 12V. Whew. Now that I have the wiring remoted to the interior, it'll be a lot easier to play around with it.
Unfortunately, this morning's was another commute that would've been perfect for a lockup switch -- I was following somebody uphill at 35 MPH and getting 22 MPG instead of 30 because the converter was unlocked.
When I got to work, I hooked up a DMM inline on the 10A range; I'll check out the current draw on the way home.
After that, I can check the impedance of the solenoid; on Thursday's drive, I'll check for any AC waveforms in the lockup signal with an amplified speaker. So far, it just looks like a simple 12V will work for an override switch, but I want to be absolutely sure.
Current draw is ~1.0±.03 Amps DC. I'll set up the speaker tonight to check for AC waveforms.
The cable I used appears to be 16 AWG (26/30); at a current draw of 1 Amp, 20 feet would come to a voltage drop of about 90 mV. If I'd used the 22 AWG used in the harness instead, the drop would be more like .3 V, which is starting to get significant -- especially if I wind up using a .6-1V diode in series as well.
The scary part will be the current draw from the solenoid when breaking the circuit -- it should easily be several amps.