For a simple voltage controlled motor the drift is normal, you can reduce that with a controller with negative output resistance as stated above, but even this will have some speed deviations because of the temperature coeffiecient of the copper windings of the motor.
A DC motor's dynamic friction changes noticable as it warms up, typically reducing as the bearings heat up. However as the windings warm up their resistance increases which can have the opposite effect. Using a much larger motor than necessary can reduce thermal effects like these at the expense of more expense...
In control systems theory, after the controller and before the device being controlled is a summing point. In your case the controller is your DC power supply or adjustable DC voltage. The device being controlled is your DC motor. In between the two is the summing point. What you are summing to the motor speed is called disturbance. The latter is anything. Temperature influence, bearing drag, drag caused by the stylus. Anything. And it will influence the motor speed.
Conclusion: it is impossible to accurately maintain speed without some kind of feedback control.
The reason that it it takes so long for your platter is similar. The DC voltage you use is good for maintaining the speed after the platter has been turning a long time. Before all parameters (disturbance) has reached an equilibrium a long time has to pass.
The feedback will regulate away any influences of disturbance. Depending of the type of controller circuit, you might have a small and negligible error, or no error at all. As a bonus the time to reach nominal speed will be much shorter.
This does not change when using an AC motor for frequency control. Just read "frequency" instead of DC voltage. Designing a stabilized speed control is fun, the results are amazing.
Conclusion: it is impossible to accurately maintain speed without some kind of feedback control.
The reason that it it takes so long for your platter is similar. The DC voltage you use is good for maintaining the speed after the platter has been turning a long time. Before all parameters (disturbance) has reached an equilibrium a long time has to pass.
The feedback will regulate away any influences of disturbance. Depending of the type of controller circuit, you might have a small and negligible error, or no error at all. As a bonus the time to reach nominal speed will be much shorter.
This does not change when using an AC motor for frequency control. Just read "frequency" instead of DC voltage. Designing a stabilized speed control is fun, the results are amazing.
Agreed mostly, of course there is drift from several sources, BUT, there are several products out there with just a simple stabilizer, I do not think who use those would adjust all the time.
That is correct. When you say stabilizer there is some feedback. Some stabilizers even cut off the power to the DC motor and then measure the back EMF of the motor and use this as speed information for the controller. While ingenious, the principle is still the same.
An asynchronous AC motor like in old turntables do not use feedback. But the source impedance of the driving source is very low as compared to the disturbances. Still those motors never reach synchronous speed due to something called slip. Slip might differ due to disturbances like bearing drag, in most cases negligible. But then again, for DIY asynchronous motors are much more difficult as the nominal speed is set by mechanical design. And there is simply no adjustment possible. Yes, you'll find an example where there is, but it is not trivial.
I don't know whether synchronous AC motors ever have been applied in turntables. Maybe direct drive actuators are considered synchronous.
I mean stabilizer like an LM317, with a helitrimmer at the voltage control, but there is no feedback from the motor in an LM317 circuit.
You find synchronous AS motors in Rega, Systemdek, Voyd, Linn ... TTs.
JG
You find synchronous AS motors in Rega, Systemdek, Voyd, Linn ... TTs.
JG
I cannot imagine there is not some (less obvious) feedback. It simple is not possible to have a DC motor run at a constant speed by applying a DC voltage. Some motors have a centrifugal regulator inside. I can imagine you still want to adjust the DC voltage to get a run as smooth as possible, so maybe that is it. I would be interested to see a schematic and motor specification of what you mention.
I checked Linn, and yes those are synchronous motors according to the label.
Progressing with the motor drive, yet DC motor version. The feedback is from the motor pulley, 1 per revolution.
Instead of reducing P, when the RPM is stable around the desired speed, it reduces the maximum rate of change to 0.2 (MaxR in the 4. row), meaning it's DAC output (driving the motor) will change with 1 bit in every 5th measurement only. I could reduce it even further, stable with 0.02 also, that is also enough to compensate the temperature drift etc. When I touch the platter or similar, it falls out from this slow mode and MaxR jumps back to the quick control mode. It controls fast again, than switch to slow mode.
Problem is, the belt is bad quality and too wide for the pulley. Some sections are touching the pulley side, making noise and fluctuating speed. I have new belt coming.
On the other hand, I also want to try it with driving on the platter directly, not on the sub platter. I have never had a TT driving directly on the platter. All the NAD, Rega, Systemdek I had was driving a sub platter. Please help me with what is a typical belt size for driving on the 300mm platter. What is the length, what is the platter to motor axle distance, what is the pulley diameter and if possible, pulley cross section as well (V or round? for round belt?).
I would buy a belt for driving on the platter directly to play with it, but it would be better if I can start with an often used size, or something what I can find at all on ebay/ali.
Thanks!
JG
Instead of reducing P, when the RPM is stable around the desired speed, it reduces the maximum rate of change to 0.2 (MaxR in the 4. row), meaning it's DAC output (driving the motor) will change with 1 bit in every 5th measurement only. I could reduce it even further, stable with 0.02 also, that is also enough to compensate the temperature drift etc. When I touch the platter or similar, it falls out from this slow mode and MaxR jumps back to the quick control mode. It controls fast again, than switch to slow mode.
Problem is, the belt is bad quality and too wide for the pulley. Some sections are touching the pulley side, making noise and fluctuating speed. I have new belt coming.
On the other hand, I also want to try it with driving on the platter directly, not on the sub platter. I have never had a TT driving directly on the platter. All the NAD, Rega, Systemdek I had was driving a sub platter. Please help me with what is a typical belt size for driving on the 300mm platter. What is the length, what is the platter to motor axle distance, what is the pulley diameter and if possible, pulley cross section as well (V or round? for round belt?).
I would buy a belt for driving on the platter directly to play with it, but it would be better if I can start with an often used size, or something what I can find at all on ebay/ali.
Thanks!
JG
There is no DC current. V3 is biased in class C. At the collector of V2 is a square wave. At the base of V3 appear positive and negative pulses. The positive pulses are clamped by D1 to +12V. The negative pulses cause V3 to conduct momentarily. At the collector of V3 appear those pulses which are integrated by C3+C4 to a DC voltage which represents the motor RPM.
It is described in the attached article, but unfortunately it is in Dutch. The diagram of fig, 5 might shed some light.