Long-time member hobbit described the thing pretty thoroughly here (in the context of building a hack to control the blower while the car is off, for camping).
You can't really test the blower controller that way. It has to be connected like it was in the car- power, ground, and control signal (variable duty cycle squarewave) going in, and a load (blower) on the output. Then see if it works. Posted via the PriusChat mobile app.
In a pinch you might use something like a headlight for a load. I see in hobbit's write-up that about 5 kHz works for the control signal, and more low time in the duty cycle translates to more output. Pinning the control input to ground could be an easy way to get max output for a quick test.
I've now realised how stupid I am. I was talking about the hybrid battery controller and thinking it was the same as the A/C blower controller which it isn't. It's the black one in there
The simple in car test of a gen2 ac blower controller is a dc voltage check on the two pin connector located bottom left of the blower case at the floorboard. It will be low dc voltage on low, higher (around 12v) on high. I bench tested a gen2 ac blower controller yesterday while replacing the blower for a woman. A 1% duty cycle to the controller produces the highest speed but 20% (medium high) is shown in the test pdf. A 12v constant voltage to the controller input will run the blower on high but is not the normal control mode. However the more common problem is the blower brushes, not the controller. 20% Duty Cycle Med High Speed (ZT-703S Meter) 50% Duty Cycle - Medium Speed
Hmm, post #5 says pinning the input to ground will run the blower on high. That's also consistent with saying "1% duty cycle to the controller produces the highest speed", as that's an input that spends 99% of the time low. Post #5 also says that a duty cycle input with more low time is what produces higher speed. #5 is also based on a very thorough report by hobbit, who built a manual speed control override for his gen 2. I'm still pretty sure the input works as #5 says.
