Project to modify a record turntable's speed control unit

[keanureeves99]

Hi everyone,

I have an unusual, and perhaps foolish, project I'm working on, as an absolute beginner. What I'm attempting to do is modify a yard-sale record turntable to extend the range of its pitch control from +/- 20% (give or take) total speed to (as close as possible) to 0-100%, so that I can slow down and stretch samples continuously for a music composition so that I can do it in real time and "play" the sample pitch/speed like an instrument. I'm aware that modern turntables offer this feature, and that they use micro controllers to do so, but the price point for a Technics SL1200, for example, is rather steep.

Someone on reddit told me that I probably couldn't do this closed-loop, but I digress. Here's what I've learned so far:

Here's the schematic and wiring diagram as provided by the manufacturer. This appears to be for a European or Asian version, but it's the same model, a Sanyo TP-B2:


The first thing I tried was to put a higher resistance pot in place of the pitch wheel, which made it go faster. Ok. Next step. I realize that the motor uses something like a servo to control wow and flutter. Then, I quickly discover that neither does this schematic describe the servo motor control, nor does there exist an easily available diagram online. Well, here's the label on the back, anyway, in case someone is a schematic-finding expert:


I can't find anything at all, so I crack it open and try to make my own schematic. Here's the top of the board with the space between the traces indicated in black:


And here's the schematic I drew:


This schematic represents everything inside of the area indicated as "Motor" on the factory schematic, with the pitch wheel and 33/45 rpm selector represented by a variable resistor in my diagram.

Shamefully, I do not own a transistor testing device, so I can't confirm anything, but I did note the part numbers for them and managed to find the data sheets online.

After studying this for a while, I come to the (perhaps wrong?) conclusion that they're set up as common emitter amplifiers, and they're configured to cancel each other out to keep the speed from fluctuating at all, which seems very important for audio playback.

I also took an oscilloscope reading the base of C2298 to ground, I think. Sorry for the vagueness here, I did it quite a while ago. To me this looks like the motor is correcting its speed by trying to cancel out the amplified signal from 2SD603.


And that's where I'm stumped. Or maybe defeated. Is there any way to do what I'm attempting without just handing over speed control to a micro controller or something? Is that even possible?

Although I do have an end goal here, which is a cheap turntable that has continuous speed control from (close to) 0 rpm to 33pm, approximately, I'm also a curious neophyte and would appreciate any enlightenment on how this stuff works, or what I should investigate next.

Thank you for reading.

[Andy Chee]

One schematic shows the motor has five wires, whereas the other schematic shows four wires.  Not sure what's going on.

[keanureeves99]

One schematic shows the motor has five wires, whereas the other schematic shows four wires.  Not sure what's going on.

(Attachment Link)

my mistake. In reality, these combine to make a single loop with a VR and switches between other resistors that add +/- 2k ohms. That's all between the blue wire (VR) and the M wire (white)

[Terry Bites]

A stepper motor will impress pulsating torque on the disk. This may feed vibration into the audio track. You may be better of with a small brushless motor. Simple low speed open loop control can be achieved with a micro. Platter weights can be used to smooth out the torque by increasing the system inertia.

I watched interesting little video this a while back. . Just the job, reverse too.
Code here https://github.com/juanpablocanguro/BRUSHLESS-MOTORS

It's part of a 5 episode tutorial including closed loop control of a small brushless motor and postion control.
You could combine the velocity and speed functions for precise motor velocity and track seeking.

[keanureeves99]

A stepper motor will impress pulsating torque on the disk. This may feed vibration into the audio track. You may be better of with a small brushless motor. Simple low speed open loop control can be achieved with a micro. Platter weights can be used to smooth out the torque by increasing the system inertia.

open loop control does seem like the most straightforward way to approach this, and very much potentially the ultimate solution I'll use

However, I'm also curious to understand how the current setup works since I'm a beginner to circuit design. I had a thought today: It seems the VR/speed selector circuit is biasing Q1 (2SD603). Is this maybe what limits the range of speed control?

[Andy Chee]

Assuming that your turntable motor is a servo motor, NOT a stepper motor....

It operates exactly the same as a thermostat.  The feedback winding is essentially a miniature alternator/generator whose voltage output is directly proportional to speed.  If the feedback voltage is low, the circuit increases motor voltage/speed.  If the feedback voltage is high, the circuit reduces motor voltage/speed.  Unlike a thermostat which is just on-off, a servo motor is infinitely adjustable.

If you're looking to adjust down to 0% turntable speed, you will definitely need a new motor.  The original motor isn't designed to have much torque at such low speeds.

[jwet]

I redrew your schematic to make it a little more intuitive as to what's going on.

Its a control loop, the only tricky part is the capacitor and resistor feedback around the second transistor.  Since the load is completely inductive, this provides frequency compensation to stabilize gain at higher frequencies where inductors open up and do strange things like oscillate at ultrasonic frequencies.  If you grind through all the math, you would see a transfer function that had some problems area with stability.  The direct capacitor from collector to base decreases gain at 20 db/decade starting at 0 Hz (dominant pole like an internally compensated op-amp).    The RC in parallel adds and additional high frequency roll off starting at RC to create a total HF roll off 40 db/decade.  The component values were probably arrived at empirically with the motor, inertia of the platter in place and step load and line changes were applied.

The only thing that seems to be missing here is a clear reference.  I think the LED acts as the reference, its voltage drop is somewhat stable.  The feedback from the control winding is compared in the first transistor to this reference and closes the loop.