Quadrature phase motor drive

[PeterBell]

Hi, my electronics skills are a little rusty - it's forty years since I took my degree.

I have emigrated from UK (240V, 50Hz) to Philippines (230V, 60Hz), with my Thorens turntable. I don't wish to listen to all my vinyl records playing 20% too fast!

The turntable uses a 16-pole synchronous motor with two windings at 90 degrees phase. The original design uses a single capacitor as a low pass filter, generating the 90 degree phase shift for one winding. The label states that the turntable is rated at 2.5W.

It seems, to me, that an effective solution, which should also bring aural benefits, is to generate a clean 50Hz signal electronically. The XR-2206 chip would seem to be an ideal and cost-effective solution - I could simply take a sine wave from that into a transformer via a power amp, and feed the existing motor circuitry. I would incorporate a pot to allow small speed adjustment. By dispensing with a 240-115 dropper resistor, I would only need 100-115V.

However, it is generally believed that the phase relationship of the signal feeding the two motor windings needs to be accurate for the motor to run most smoothly (with aural benefits). How can I generate an accurate 90 degree phase shift, without needing a 'scope to adjust? If I can do this, simply doubling up on power amp and transformer would be very simple.

As a final, and ultimate, enhancement, it would be nice to add electronic speed switching, requiring 37, 50 (and 86.6?) Hz. I know that this complicates the phase shift, but is there a way to do this without a Hilbert transformer (requiring high precision resistors/capacitors).

Can the XR-2206, or a similar device, be phase-locked to a triangle wave, so that two XR-2206 chips could be phase-locked at 90 degrees?

All ideas and suggestions gratefully received.

 

[crutschow]

Using the XR-2206 to generate a sine-wave and amplifying that to power the turntable should work fine. You could use a power transformer to increase the voltage from the power amp to the required 110V.

But I think you concern about the phase-shift is unwarranted. I would think the built-in capacitor would give a phase-shift close enough to 90° that you could never hear a difference from having an absolutely accurate 90°.

Offhand I don't see any way to sync two XR-2206 chips together.

 

[PeterBell]

Thank you for your response - it is much appreciated!

My concern over accuracy of the phase-shift was prompted by this site. However, I'm not sure whether it is a real problem, or simply 'smoke-and-mirrors'. However, I can see that deviation from the true 90 degree phase difference will lead to 50Hz torque pulses.

Yes, I didn't see a way of syncing two XR-2206 chips, but thought that it was worth asking the question of someone more experienced.

Can anyone suggest other devices which can produce a clean and accurate sine wave in the 30-100Hz band, and which could be triggered/phase-locked?

Perhaps I should just get on and build the simple version - at least I can then listen to my vinyl - and decide whether to tinker further afterwards.

 

[alec_t]

However, I'm not sure whether it is a real problem, or simply 'smoke-and-mirrors'. However, I can see that deviation from the true 90 degree phase difference will lead to 50Hz torque pulses.

Here's the result of simulating the Fourier spectrum of the product of two 60Hz waveforms (which I reckon would be indicative of motor torque) with (a) a 90 degree phase shift and (b) a 120 degree phase shift.

In both cases the secondary peaks are ~30dB or more below the main peak at 60Hz; so IMO there is no real problem, particularly when motor inductance and motor/turntable inertia are taken into account.

Edit: Depending on how much of a purist you are another option might be to record from vinyl to PC then frequency-convert the results using Audacity or some other audio processor. I've had acceptable results from ancient 78 rpm records played at 45 rpm doing that.

 

[Sangoma]

Is it belt driven?
If so, having a new motor pulley made may be a simpler solution.

 

[PeterBell]

Here's the result of simulating the Fourier spectrum of the product of two 60Hz waveforms (which I reckon would be indicative of motor torque) with (a) a 90 degree phase shift and (b) a 120 degree phase shift.
In both cases the secondary peaks are ~30dB or more below the main peak at 60Hz; so IMO there is no real problem, particularly when motor inductance and motor/turntable inertia are taken into account.

Thank you for the thought, but I'm not convinced that plotting the product of two out-of-phase waves is really representative of torque for a motor with two windings in quadrature.

Edit: Depending on how much of a purist you are another option might be to record from vinyl to PC then frequency-convert the results using Audacity or some other audio processor. I've had acceptable results from ancient 78 rpm records played at 45 rpm doing that.

Interesting idea, but being able to play the vinyl and staying in the analogue domain is the real aim. Using this trick on 78s might be okay, but doing something similar for hi-fidelity modern vinyl is unlikely to yield satisfactory results.

 

[PeterBell]

Yes, definitely belt drive and 60Hz pulleys used to be available - however, finding one and getting it to Philippines may pose a problem. In any case, the quality of the utility supply here is rather poor - I've seen 5% deviations from the nominal 60Hz and I'm not convinced that the waveform is all that pure, either.

 

[Sangoma]

Yes, definitely belt drive and 60Hz pulleys used to be available - however, finding one and getting it to Philippines may pose a problem. In any case, the quality of the utility supply here is rather poor - I've seen 5% deviations from the nominal 60Hz and I'm not convinced that the waveform is all that pure, either.

Yes, good point, and that is assuming any supply at all!

Thinking of the brownouts and blackouts, maybe a DC to AC inverter running from a battery, and a charger running from the mains when it does.

One thing there, is having a pulley made would be cheap.

What about a pic, atmega or similar, programme two outputs

 

[PeterBell]

Yes, the inverter was my 'plan A'. I brought a 'pure sine' inverter with me from UK, but it is no longer functioning - the oscillator doesn't run. I've requested a schematic from the manufacturers, but they don't even do me the courtesy of replying. I could, perhaps, spend a few hours sketching out the schematic myself.

As for brownouts - there's no point in running the turntable if the amp isn't powered! In any case, I do have a generator.

Yes, I could look for a machine shop to turn me up a pulley, but getting a decent job done here, as opposed to a bodged job, is difficult. In any case, the original pulley is a two-piece nylon assembly, incorporating a friction clutch.

Again, a pic solution had occured to me, but my first task would be to obtain, or build, a programmer - a fun project for when I have lots of time. In the meanwhile, the simple XR-2206 solution may be the pragmatic option.

 

[Sangoma]

Yes, the inverter was my 'plan A'. I brought a 'pure sine' inverter with me from UK, but it is no longer functioning - the oscillator doesn't run. I've requested a schematic from the manufacturers, but they don't even do me the courtesy of replying. I could, perhaps, spend a few hours sketching out the schematic myself.

As for brownouts - there's no point in running the turntable if the amp isn't powered! In any case, I do have a generator.

Yes, I could look for a machine shop to turn me up a pulley, but getting a decent job done here, as opposed to a bodged job, is difficult. In any case, the original pulley is a two-piece nylon assembly, incorporating a friction clutch.

Again, a pic solution had occured to me, but my first task would be to obtain, or build, a programmer - a fun project for when I have lots of time. In the meanwhile, the simple XR-2206 solution may be the pragmatic option.

I bought a pic programmer for about £7 on fleabay, and it works very well, however for this use, a picaxe would be fast enough, simple to programme in basic, and doesn't need a programmer, add a 16X2 lcd display, and you could have a choice of speeds and fine adjustment.

Should be able to use a 7474 to give two outputs with a 90deg shift as well, as another possibility.

 

[crutschow]

If the microprocessor has two PWM outputs then you could generate two sine-wave outputs with a 90° phase shift using a sine lookup table.

Note that the voltage applied to the 90° winding for normal operation may be different than the primary voltage.

 

[Mr RB]

The turntable motor is usually designed to operate at a standard fixed RPM. Do you know the RPM?

As a repairer year ago people would bring me expensive turntables they imported, or brought themselves, and I would simply replace the motor with one at the same standard RPM but suiting the local mains Hz.

That is probably still your best option, provided you can source a motor of the correct RPM and correct pulley diameter, the motor frame sizes should be easy enough to match.

 

[alec_t]

Do you know the RPM?

You can calculate the motor rpm as turntable-rpm * turntable-diam / motor-pulley diam.