I am restoring a vintage sewing machine equipped with an electric motor.
It's a 100W brushed motor operating at 220V (AC), unfortunately no other information is available from the plate. Having a woven stator and rotor, it is very likely to be an universal motor, unless someone else can object to that. Current speed control is achieved using a foot-operated rheostat, which becomes quite hot after sewing at slow-speed for some minutes.
I would like to replace the rheostat with an Arduino-driven speed controller to achieve higher efficiency or at least to eliminate the nasty overheating issue. Ultimately, I would also like to computer-control the thing, but that will be trivial as soon as I have the microcontroller in place.
I have researched two possible approaches, with some pros and cons:
- phase-angle partialization (AC): simpler and cheaper circuit, but lower efficiency and higher EMI
- high-frequency PWM (DC): more complex driver circuit, higher efficiency, lower EMI but increased wear on brushes
Regarding brush wear, this is what I infer from a high-level motor control presentation by ST Microelectronics, although no justification is given for the fact. If it wasn't for that, I would definitely go for the latter solution.
Any buzz noise is not a big issue since the machine is already super noisy (it's fully mechanical).
I don't plan to mass-produce or even sell the speed controller, so the cost is not an issue, as long as it is not prohibitively high. Same goes for construction complexity.
And now for the questions:
- Are there any other options to consider?
- Is the impact on brush wear actually relevant/correct?
- What is the overall best choice for the intended purpose?
Finally, I would like to save some collateral discussion by saying that I know how to safely handle mains electricity (e.g. that I need opto-isolators and such), but any consideration is welcome :)