Turntable speed control


Author Message
Mixtel90

Guru

Joined: 05/10/2019
Location: United Kingdom
Posts: 8939
Posted: 01:58pm 07 Jan 2023      

This is just theory at the moment, but...

Any of you into hi-fi will know that switching between 33-1/3rpm and 45rpm isn't always quite as easy as you might think. Early systems are mechanical, using different diameters of pulley on the motor shaft. The posher systems use a synchronous motor but change its supply frequency.

The 24-pole motors usually used rotate at 250RPM on a 50Hz supply (300RPM at 60Hz). This is geared down by a belt drive to give a turntable speed of 33-1/3rpm. To get 45rpm you need to increase the supply frequency from 50Hz to 67.5Hz (or from 60Hz to 81Hz).

Of course, we are talking about sine wave supplies here. The mains is a more or less reasonable sine wave, the motor "cogs" a bit and the give in the belt, coupled with the inertia of the turntable, even things out. The idea then is to produce a sine wave at about 90V that can be switched between 50Hz and 67.5Hz. Additionally, a capacitor is used on one winding of the motor and ideally this should have a slightly different value for each speed to minimise motor vibration.

My idea is to use a PicoMite with a R2R ladder, at least 8 bits and preferably 10bits or more, to generate the sine wave from a lookup table. The resulting output would feed a class D bridged power amp module from ebay - probably about 30W - to feed a step-up transformer to get the voltage required for the motor. I'd be relying on the crystal oscillator on the PicoMite for the master frequency. I've not worked out whether I can get the correct division values yet, or whether the lookup table can be processed fast enough for higher bit rates.

Things similar to this are already done - usually with boards of expensive logic rather than microcontrollers. They also have the usual 1500% audiophool price markup. This solution costs a PicoMite, a £6 amplifier board and about £20 for the transformer plus a few odds and ends. It shouldn't, in theory, perform any worse than £500 motor speed controllers.

Any comments?