A main issue we have to deal with is that these motors are designed for extremely high RPMs and rather low torque which is the opposite of what we need for most practical applications.
The machines tend to have rather complex control circuitry and it's probably easier to simply salvage the more valuable components from the control board and build a new circuit than spending ages reverse-engineering (unless, of course, you love a good hardware hack).
It 's a good idea to replace the taco coil on the back of the motor with a hall-effect sensor as this will give us a feedback signal that's easier to use.... pulses at exactly the frequency of revolution.
If we connect the field-coils and the armature winding as a self-excited series-wound motor, the speed runs up and up and up until it exceeds it's rated RPM and destroys it's own brushes and bearings.
Without a control circuit, when we apply just a little load the available torque is insufficient and the speed drops abruptly.
These motors require some means of regulation to use them safely.
There's two paths that can be taken with the control electronics:
- Keep the circuit itself very simple but this is going to take more advanced electronics knowledge to design.
- Keep the design principles simple, use a microcontroller but this will be more expensive and complex to build.
We can use back EMF for feedback to a simple thyristor based controller to keep the speed constant... But this circuit only provides half-wave power.
Better control can be obtained by wiring the motor as a separately excited DC motor and provide the armature with 10 - 30 volts to control speed and 1 - 1.5 amps through the field coils to control torque.
As increased torque results in less speed, we can operate the armature at a fixed voltage and use the field coil current to control the motor.
If we use a fixed voltage on the armature, we still need to switch it off to stop the motor, it will continue to run with virtually zero torque with no power applied to the field coils.