Success!!!! I hooked the TPS to the TCM, using a diode to prevent the 8V output of the TPS from being applied to the 5V input of the TCM. Worked like a charm. Here's a little video I made showing what it does:
Like I mentioned, I used a diode to block the 8V from being applied to the inputs on the TCM. The inputs have weak 5V pullups, and putting 8V to them could cause damage. Here's a schematic of the TPS to TCM connection:
I only connected it to the "Managed Input" pin for testing, but I am about 99% positive that it needs to be connected to the "Unmanaged Input" pin as well, otherwise the TCM will throw a code and inhibit TC lockup :rasta:
From what I understand, the Unmanaged Torque is the torque being requested by the driver (basically equal to throttle position), while Managed Torque is the torque actually being delivered by the engine. If no torque reduction is enabled they
should track each other. Torque reduction on a computer controlled engine could come from a request from the TCM, or from the traction control system, or from the ECM itself due to certain conditions. When that is the case, the torque requested by the driver may be quite different from what is actually being delivered. Since shift speed is dependent on throttle position (unmanaged torque) and shift firmness, trans pressures, and TCC operation are dependent on actual engine output torque, the trans would need to know both for proper operation.
I mentioned in the video that I wasn't sure why the torque readings went negative. Well, thinking about it, this is the actual torque being delivered by the engine. If you are cruising and let off the throttle, the engine torque is now
negative - the transmission is driving the engine, rather than the other way around. With that in mind, the readings make sense. Obviously those are not real values. They are just numbers programmed into the TCM to equate to a certain input pulse width.
In the case of a mechanically controlled engine (or an electronic engine with no torque reduction enabled), unmanaged torque and managed torque should always track each other. By connecting the TPS output to both inputs, were faking it. The TCM shoudln't care - it's going to base it's shift firmness off of the managed torque value and adapt to the rest. All that should matter is that the inputs are consistent, which they will be. The only thing we need to do is disable torque management in the TCM, otherwise it will request torque reduction for shifts, and when that doesn't happen it will throw yet another code. Luckily, EFI Live lets us disable TR