For an RCA 45 player I used the DC brushless video head motor from a VCR with a basic closed loop speed control circuit that works ok. Runs the 45 player just fine.
I recently acquired a VHS machine from Freecycle (wanted the capstan motor) and was surprised to discover the 12 pole three-phase motor for the upper drum. Being a PAL machine that wraps just over 180 degrees, it will be designed to rotate at 1500rpm. The PCB even had significant signals labelled, so I may try driving it from a function generator and use its servo rather than ditching its electronics and driving it open loop with sine waves.
Hi,
Indeed , there will be variety between these motors.
Nevertheless 12 poles is 4 pole pairs @ 3 phase. It should run 750rpm @ 50 Hz which is very good.
If you run outer platter as I do with 300mm OD, and pulley is lets say 20mm, ratio is 15 *33.33333 = 500 rpm, that's what you need.
So for 33.33 you run it at 33.33 Hz
and for 45 rpm art 45 Hz
Very good and convenient as any frequency meter will show rpm speed
Hi, 12 poles will be good, just replied to Ec8010 above about it. But still its watermellon as you said, likely its 2 phase motor normally running with 90deg shifting capacitor from mains, and almost sure running at over 100V rms..... This will all work fine, but with many kg = $ in power drive , not necessary any more IMHO.
Here's how I got the motor I mentioned to work. The circuit is dependent on the tach signal changing in amplitude as the speed changes. The tach out is so that once you get the speed set right with a strobe disk, you can measure the frequency and set the speed using a frequency counter instead of having to use a strobe disk. The circuit is also somewhat temperature sensitive. The transformer is needed to isolate the DC voltage riding on the tach signal as I've tried it with just a cap and no transformer and it didn't work properly. I find with mine once set so long as the average room temperature doesn't change much, I don't have to adjust the speed at all.
Those motors usually run on 12Vdc, but check the driver chip datasheet to see what the maximum voltage it is rated at and don't exceed that voltage.
They also only turn one direction with no easy way to reverse direction and don't try to use one without a speed control circuit for things where a constant speed is needed.
The VCR video head motors also don't have that high of a torque and don't run that well real slow.
For slow high torque applications a DC brushless capstan motor would work best. Those require both 12 and 5 Vdc, can be reversed and can be stopped as well.
Also my finding.
Also correct.
But if you discard drive electronics and take bare naked motor, it is happy with 2.5 VRMS, which in sense of square wave feed from driver chips would mean that 5VDC PS is all that's needed. For pure sinewave drive (that is much preferred IMO) , 9 VDC PS is fine.
With those standard chips one will never get sinewave 3 phase supply, with Supaspin from RichB its so easy and with less than 50 euro cost, even less if you have some common parts in drawer.
Supaspin looks interesting.
I replied to the thread about it to see if it will work with a VCR video head motor.
I replied to the thread about it to see if it will work with a VCR video head motor.
Last edited:
Hi, it does.
Little bit earlier in this thread I showed figures with drum head motor from Sony HD pro Digital recorder, That one works. I think any BLDC motor that equals 3 phase permanent magnet synchronous motor, will work.
It is about amplifiers you use after Supaspin, Supaspin at its own delivers very accurate 3 phase line level signals adjustable in frequency by 0.1 Hz steps, also includes frequency ramp up for soft start.
It is up to one choosing the motor to design - choose proper 3x power amps to deal with motor current demands.
Little bit earlier in this thread I showed figures with drum head motor from Sony HD pro Digital recorder, That one works. I think any BLDC motor that equals 3 phase permanent magnet synchronous motor, will work.
It is about amplifiers you use after Supaspin, Supaspin at its own delivers very accurate 3 phase line level signals adjustable in frequency by 0.1 Hz steps, also includes frequency ramp up for soft start.
It is up to one choosing the motor to design - choose proper 3x power amps to deal with motor current demands.
Does it also have a tach function so that it can keep the motor at a constant speed with various loads once the initial speed is set or is a tach function simply not needed due to how accurate the circuit is?
I'm tempted to try it with three small class D amplifiers.
I'm tempted to try it with three small class D amplifiers.
That's all interesting stuff. But I would have expected the drum servo to be reliant on an FG (Frequency Governor) signal's frequency rather than amplitude. Leastways, that was the case in broadcast VT machines. The built-in servo ought to be quite good. After all, reproducing vertical lines on a television screen requires quite accurate timing of the upper drum along the track.
It does not have feedback if that's what you mean.
Feedback is needed with DC and asynchronous motors to hunt the right speed, but that is bag of worms at its own. Speed of happenings with mechanical things is so much slower than with electrons running in wire, so many chances to oscillate. Think of PID function.
Synchronous motor will lock in the exact speed given by generator and keep it regardless of the load, and Supaspin will give exact frequency to support that.
Deviations might occur only because of belt slipping, that's why I recommenced min drive pulley OD of 15mm. Another possibility is belt elasticity making vibrations: good belt and good platter bearing with tight tolerances and some drag will make it. Heavy platter also helps a lot on this subject.
Supaspin however has tacho function (when you connect the sensor) to give you reading (on convenient display) of exact speed averaged over one rotation of the table so you can correct it in 0.01 Hz steps (you can choose also 1/2, or 1/4 of rotation, but it makes no sense in my opinion).
What Supaspin lacks at the moment is sensor reading of very many points from one rotation. That would give frequency reading whose deviations could indicate micro speed osculations and resonances.
I thing that could be programed, what is the issue is what sensor is needed to give such precise signals. Only thing I can think off is to connect optical shaft encoder to the turntable, industrial servo type that will be large and several hundred $ in cost.....
If someone has no better idea, I think we need to rely on good belt with little or no stretch, perfect belt to pulley contact and excellent bearing with some drag (not friction less bearings IMO)
In the VCR that's what it was reliant on, however the circuitry was on the main VCR board itself somewhere and I don't know enough to design a digital feedback dependent on frequency which is why I used the circuit I did. Works great though as I can put load on the motor with my finger and feel the torque increase nearly immediately.
I'd require feedback as the motor would lose speed as the reject cycle occurred and if it fell out of sync it would just stop. I'm guessing the motor spins closer to the speed it would run in a VCR. That said there's possibly more current that can be provided with the external driver versus the built in driver chip so the motor likely will have more torque and may not slow down much if any during reject.
Wouldn't be too hard to do.
The way most video head motors do it is they have a magnetic ring around the bottom of the rotor and they have a trace on the motor board that is like a squarewave which goes around the whole rotor area with the trace connected to the chip which buffers the signal. That generates a frequency dependent on the motor's speed.
Don't see why that cannot be scaled up to measure the speed of the platter.
Another option is to look at how the quartz locked record player motors did the feedback and duplicate that.
Thought of perhaps magnets on the platter and a hall effect sensor, but then thought no, the magnets might induce a field into the cartridge which would be audible.
Presently, only one trigger spot is attached to the platter, dark area sensed by optical sensor. Sensor triggers reset signal to Supaspin on RPI. In between reset signals (one turn of platter) Supaspin counts RPI computer clock and gives very precise speed , but average of one rotation of the platter.
To get micro speed deviations we need to do opposite, set the counter reset time to some fixed value like 10mseconds and count sensor triggers in that period. If we take 10 mili second and say that at least 9 signals are needed to measure, that gives us 5000 trigger points per rotation.
Problem is that those 5000 points must be absolutely equally spaced as otherwise we measure deviation in our trigger points and not speed deviations.
Is it hal sensor, or magnet, or optical, problem remains the same.
Industrial optical disc could manage that, or one can glue CD recorded with say 1khz and install laser underneath platter 🙂
To get micro speed deviations we need to do opposite, set the counter reset time to some fixed value like 10mseconds and count sensor triggers in that period. If we take 10 mili second and say that at least 9 signals are needed to measure, that gives us 5000 trigger points per rotation.
Problem is that those 5000 points must be absolutely equally spaced as otherwise we measure deviation in our trigger points and not speed deviations.
Is it hal sensor, or magnet, or optical, problem remains the same.
Industrial optical disc could manage that, or one can glue CD recorded with say 1khz and install laser underneath platter 🙂
I thought of using a dividing head on mill to drill holes in the perophery of the platter to mount magnets, then realised that was a stupid idea because it would drive a signal into the cartridge. A safer idea would be to put shiny cylinders in the holes and use an optical sensor instead.
If you use reflective optical sensor, for me difference in reflection between white(ish) stone and black insulation tape is good enough. You can just print stroboscope on white paper.
With reflective sensor problem will be different, they need some minim distance between reflective and black to operate. I did not dig deep into it so cannot confirm how small is too small.
If you have access to precision divider head, I would think of drilling small holes around and use laser opto coupler. Or ready made optical encoder disc with laser coupler....
Nevertheless i dont think any of this is actually needed. Using servo feedback to correct motor speed is very tricky business, and not needed neither. Precision synchronous motor with accurate sinewave drive will rotate on fixed speed, no need for corrections there.
To measure TT speed deviations for my info I use 1khz test tone from test LP, even I must say Im not satisfied as results are worst than I would like. But than again who says that test record is perfect....
If one day I could borrow or buy cheap an shaft encoder, I could attach it at spindle on the top of the platter, and check speed in details and accuracy, to test the test LP and TT.
My thinking is to solve this problem first , before thinking of servo feedback.
In my case I could get away with just using a sensor to measure the motor speed and having the code be adjusted for that.
The slowdown is normal for reject using the motor I used given the reject cycle is fast and puts extra load on the motor.
The slowdown is normal for reject using the motor I used given the reject cycle is fast and puts extra load on the motor.
@Drbulj: Yes, in the end, running the thing open loop from a synchronous motor driven from a stable frequency is perfectly adequate. I have a dividing head, so it was really a solution looking for a problem.