Sony PS-X65 Vintage Turntable Brushless motor drive repair

[mjkuwp]

hi All!

I have a really nice vintage Sony PS-X65 turntable of the "PS-X" family.  It's broken and I am trying to learn electronics as I repair it. It's not as bad as it sounds.  I'd like to discuss the brushless slotless motor that is used in this family.  It has never held speed properly (lots of wow and flutter) and only recently have I made some headway in figuring out why.  I think it is the "hall element" sensors inside the motor which are supposed to do the commutation and which do not appear to work.  I have obtained a parts donor and research table; a PS-X40 which does work.

My plan is to reverse engineer the specifications of the hall element in the motor and then replace the elements with a modern sensor such as an Allegro 1360 (Hall Effect Sensor IC).  My problem will be that it is unlikely to be a drop-in replacement and that I do not have enough Electronics know-how to easily create a work-around. (I am a mechanical engineer and very new in the electrical hobby).

Here is what I assume about the original hall element:  I assume it is a wheatstone bridge with one or more of the legs being magnetoresistive.  I have been studying these sensors and they appear to be quite challenging to use due to temperature and strain sensitivity drift problems.

I furthermore assume that with the replacement IC the basic characteristics should be much more stable but I am not sure how to adjust my circuit for this amplified, ratiometric sensor. 

I have one concrete question:  Is it reasonable to use the excitation voltage intended for the ?wheatstone bridge? hall element as the power supply for my ratiometric IC? Seems like a good idea to me...but then I am new at this.

I have some pictures from photobucket: the thumbnails are shown but you can click the image for full size

Here is a picture of the motor winding:  Two phases driven by +30V/-30V



Here is a close-up of the hall elements.  I think one was dead/flatline and during my troubleshooting I think I killed the second one (hence the PS-X40 donor machine)



Here is the circuit:  Sorry for the poor quality.  I'm not concerned about the left-hand side.. probably works.  The RH side starts with what I think is inside the hall elements indicated in the blue rectangles... then op-amp to drive the PNP/NPN.  There is feedback to the opamp




the circuit is laid out in US Patent 4135120 and there is an ok explanation in that document.

I think it is simple as this: 

Hall Elements provide sine-wave for commutation
feedback on motor winding to opamp compensates for the zero crossing of the NPN and PNP that would otherwise create notchiness
PLL Circuit obtains speed feedback from the rim of the turntable platter and adjusts the gain which is the excitation voltage of the hall element sensors

My first post!
Mark

[mjkuwp]

answering my own silly question here:  it's totally unreasonable to replace the differential output hall element with a hall effect IC because the hall effect IC requires something very near 5VDC.  it is ratiometric but only within a small range.  The excitation voltage actually crosses zero and I am positive I will not find an IC that works that way.

Also, the modern hall effect IC chips are single-ended output.  So it looks like what I will need to do is build an op-amp circuit to drop into the existing circuitry in order to simulate the hall effect IC.

I suppose another thing would be to find a more appropriate substitute but my searching has led me to believe such devices are out of production.

I am trying to use LTSpice to model the circuit but I am having trouble figuring out how I can replicate the wheatstone bridge (if that's what it is) which would have one or more resistances that change in order to provide commutation to the motor.

[mjkuwp]

I am still making slow progress.  I need to create a new motor commutation circuit.  I now know that the magnets are 1500 gauss at the surface and there are some options for magnetic field sensing IC.  The bare hall element that the turntable originally use are just not made anymore so a direct replacement is not easy.

In the original design the GAIN signal from the speed feedback is placed across the wheatstone bridge that was responding to the magnets.  this combination created a sine and cosine wave that varied in amplitude according to the gain (excitation).   the differential output fed into an opamp and then into NPN/PNP to drive the motor.  The concept is somewhat simple.

My plan is to use an analog multiplier chip to multiply the gain signal by the motor commutation signal since I cannot directly create this.  The hall effect IC are single ended and create 2.5V at 0 magnetic field. 

Does anyone have comments on this general plan? 

I have a rough schematic but I still need to do some reverse engineering on my good turntable to attempt finding the impedence of the circuit so I can size the voltage dividers and hopefully get my circuit to a state where it will not disturb the original motor feedback circuit.
When I retrofit this to the table I will take the +30V supply -30V supply and also the GAIN signal from the original board.  The rest of the design will come from this new circuit.

Click to go to photobucket and view bigger size.


Another question:  Since the hall IC will create a signal centered around 2.5V... and the rest of the circuit is referencing ground, I had the idea to supply my two hall effect IC with a +2.5V and -2.5V rail.  I am using an adjustable regulator for the negative rail.  To me this seems easier than adding a stage and moving both signals down.  I don't need a 5V rail in my board - just the 15V/-15V   for the opamps and multiplier IC and 30V/-30V for the motor coils.    Does anyone see any pitfalls with this or have any comments?

please note that i do not have formal education in electronics so my terminology could be a bit off.  I am a mechanical engineer and so I have had basic circuits only..