>Well, if there is any way of getting it to work, I think that the
>transmitter expects to find a variable resistance at several of the pins of
>the buddy cable. I don't think that the trainer transmitter actually
>generates any signals, I think that the buddy cable simply reads the
>control pots.


I doubt that it's sending resistance through the cable. Most likely it's
encoded
pulsetrain output, which gets switched in or out by the master box.

>I'd bet that a PIC wouldn't have aproblem generating the PWM that is the
>actual modulating signal. If so, you could just canibalize a transmitter
>for its RF unit and feed it the signal from the PIC.


Well.. A pic could do it, but I just sent dontronics my "getting started"
code for the AVR8515 which happens to include code to generate eight servo
outputs. It would be an absolute no-brainer to hack that to output the
channels all on one pin. It has programmable position for all eight
channels, plus programmable frame rate. I was thinking along the lines of a
robot controller or some such, with the servos local to the CPU, but this
application would work as well.  It would be a very quick hack to make this
an RS-232 to servo controller, as int driven serial comms are also in the
package. Feedback would be possible, in that you could read back the current
channel settings. You have no way to know where the servos are at the other
end of the link, but you could know where you last told them to be.

It also includes a 19 bit PN generator, and stepper motor control, and a few
other little functions.
With all that, the VAST majority of the CPU time is idle.