Model train throttles?

[kevinkr]

I wasn't intending to tease. As has been pointed out, there are two main problems with model cars and trains. Intermittent high resistance contacts and stiction. The solution in both cases is the same; short high voltage pulses. Of course, it's not going to be EMC compliant (not with those pulses connected to an aerial), but it will give good control. What you would need would be a line scan valve as your series pass element, driven by variable width pulses from a 555 or similar. A simple triode amplifier between the 555 and the PL519 would suffice. Poor smoothing before the controller would be useful.

The 317 approach is the exact opposite of what is needed.

I use homebrew PWM with my model trains (another foamite revealed ) I generally use big mosfets or bipolars. 556 makes a cheap and nifty PWM controller with a small number of parts. Hint: Use the second timer as a comparator looking at the fixed rep rate ramp generated by the first.. Incidentally a 40 - 50Hz repetition rate works great with just about every motor I have tried. Performance seems to diminish above 60Hz or so with typical can and older style open frame motors.

Maximum voltage at 100% duty cycle should not exceed the dc voltage rating of the motor to avoid overheating problems. Use sweep tubes for power elements (CF) if you must use tubes, and put them inside a feedback loop with an op-amp (OPA445) to control the output voltage under varying loads. (Open circuit voltage hazard otherwise.)

PWM results in oodles of torque, guaranteed starts, and good low speed operation. Buzzing at low frequencies is minimal. Higher frequencies result in inductive instead of mechanical integration in the motor and produce lower torque and a lot more noise. I dream of diy dcc someday..