[Publius]

So my MMF5 has what appears to be a single-phase AC synchronous motor. The 120V cables goes right into it, and there does not appear to be any significant power supply circuitry on it.

I strongly suspect - nay, I know for certain, numerically - that this has virtually zilch power supply rejection, and probably has an impact on the sound. There are 60hz harmonics that are controlled by only whether or not I turn the turntable on. And I tried powering it by my UPS once and it made a beautifully loud buzzing sound.

I suspect that most other turntables in that price range are going to be AC synchro too, and even if they weren't, I have no intention of getting another turntable. And given all the factors involved I'd say the best way to improve this situation is to build my own AC power regenerator.

This of course is quite hard, and I've come to the following conclusions about the proper design:

• Output impedance must be 1 ohm or less.
• I can't use an output transformer because it will increase the output impedance too much
• I can't use lowpass filtering for the same reason.
• Oscillator output must be digitally controlled to properly control both the frequency and amplitude of the signal.
• The load (a motor) is largely inductive.
  • 
    For these reasons... I'm strongly considering a PWM switching output as the best solution. Run a comparator at a very high speed (20khz or above) between the oscillator and the regenerated output, and switch between +-209V DC rails that are generated from the source AC.
    
    Any comments or obvious flaws with this plan? Besides the fact that switching power supplies are hard?

 

[Gontran]

How much power does your turntable need? I'm guessing on the order of 10 or 15 watts.

You could build a nice stable op amp based 60 Hz oscillator, followed by an audio power amp (a 100 watt pawn shop special could work) and feed a step down mains power transformer, but in reverse. Like a 120VAC to 24VAC toroidal that you actually drive the 24V side with the amp output(s)

I have never tried this but it sounds like a fun experiment and cheaper/easier if it works...

 

[Publius]

I'm worried that a transformer would increase the impedance too much. But then again, a synchronous motor would be a highly linear load anyway, so that's not quite as much of a concern.

I've reviewed op amp oscillators and they are sh***y. Very little amplitude control and very temp dependent on the frequency. Realistic THD numbers are hard to come by too. Digital looks much, much more accurate for the time I want to spend on the design.

 

[Garbz]

With careful circuit design it should be possible to get a very acurate and stable wave. I'd be more inclined to find an occilliator source made from logic gates then amp it with the opamp though.

The impedance shouldn't matter providing at 50hz there's a low resistance. Afterall many power filtering units rely on high impedance above the frequency of the mains current.

 

[Francis_Vaughan]

I'm curious to know why you have specified the output impedance so low. It seems very stringent.

Since it forms a system with the mechanical properties: frictional losses, platter rotational mass, drive belt spring constant, motor mass, and then into the electrical world, all coupled together, this all needs working through.