Turntable Speed Stability- Makes Sense?

Just in general, a question for turntable audiophiles- my thoughts are that speed stability is best achieved without a speed monitoring / feedback control device and the use of very large platter pulleys (or the platter itself).
The "high end" turntable designs tend to have the largest diameter driven pulley off of the belt.
Example- the belt wraps around the edge of the platter, a very large diameter when compared to the pulley on the motor.
This would put more available torque and resistance to drag at the platter and record surface and translate into less variations because of changes in drag on the stylus.
It also means that small variations in motor speed are less noticeable at the platter by virtue of the difference in pulley sizes- the pulley multiplier / ratio effect. This means that small variations of the motor speed would be reduced to a fraction of their actual percentage because the size of the motor pulley is so small compared to the platter size.
Example, motor pulley is 1/6th of the diameter of the driven platter, therefore variations in motor speed will be communicated to the platter at 1/6th of their speed difference magnitude.
Contrast that to a smaller driven pulley - example a small sub-platter- where the sub platter pulley could be 2 or 3 times the size of the motor pulley.
The smaller sub-platter pulley would have less available torque against drag, and be much more susceptible to speed variations of the motor pulley- instead of the platter seeing 1/6th the magnitude of the motor deviation, the platter could see as much as 1/2 the magnitude if the sub platter is only 2X the diameter of the motor pulley.
Smaller driven pulley turntables - e.g. Rega with small sub-platters- tend to drive the need for external speed control units.
Speed control units tend to have the "hunt and peck" phenomena, where the constant monitoring and feed back in minute speed changes can show up audibly. I believe I notice this (on some occasions) with my direct drive table.
So does this make sense? Is the most speed stable way to drive a table is to wrap around the large pulley size of the platter edge?

You seem to have a good grasp on things.

Speed control units tend to have the "hunt and peck" phenomena, where the constant monitoring and feed back in minute speed changes can show up audibly. I believe I notice this (on some occasions) with my direct drive table.

Click to expand...

True when poorly instigated, at the least bi-directional speed control is needed.
If not stated it likely won't have it.

There are too many generalities in your discussion. Speed stability is affected by many factors. Speed stability is addressed differently depending on the drive design. For instance, for belt-drive systems using AC synchronous motors, the speed is determined by the line power frequency (60Hz in the US) and to a lesser degree the line voltage (120 Volts AC). The line frequency is a very stable reference. Line voltage can vary and affects the power being applied to the motor which can cause variations in the torque being generated by the motor potentially speed variations. There is not much that can be done about this other than to use a power conditioner to control any voltage variations and spikes. A UPS with a built-in power conditioner works very well to smooth out the incoming power being supplied to the motor. This very well may be a good application for a "speed box" in the case of the AC synchronous motor driven belt drive.

There are other drive methods, such as DC Servo belt drive, direct drive, and direct drive with quartz lock. There is a really good explanation of different drive types here:

http://www.kabusa.com/ttdrive.htm

You either rely on platter mass to eliminate small variations of speed due to needle drag, or should use feedback in speed controller. Any controller without feedback can only achieve long term stability. And feedback should be taken from platter, not motor when belt drive ius used.

If you want to see how stable is speed, you need test disk with 3150 test tone and high resolution spectrum analyzer (down to 0.1Hz). It allows directly see effects of short term speed instability (wow and flatter).

Decibel_116 said:

There are too many generalities in your discussion. Speed stability is affected by many factors. Speed stability is addressed differently depending on the drive design. For instance, for belt-drive systems using AC synchronous motors, the speed is determined by the line power frequency (60Hz in the US) and to a lesser degree the line voltage (120 Volts AC). The line frequency is a very stable reference. Line voltage can vary and affects the power being applied to the motor which can cause variations in the torque being generated by the motor potentially speed variations. There is not much that can be done about this other than to use a power conditioner to control any voltage variations and spikes. A UPS with a built-in power conditioner works very well to smooth out the incoming power being supplied to the motor. This very well may be a good application for a "speed box" in the case of the AC synchronous motor driven belt drive.

There are other drive methods, such as DC Servo belt drive, direct drive, and direct drive with quartz lock. There is a really good explanation of different drive types here:

http://www.kabusa.com/ttdrive.htm

Click to expand...

agreed. and i was speaking in general terms. all things being equal a belt driven turntable would have more speed stability if the platter pulley were larger. the belt wrapping around the platter rim (example music hall mmf-7.1) would be more stable than the belt wrapping around a smaller diameter sub-platter (example, rega RP-*).

Totally agree, and this also allows the use of a much heavier platter given the same size drive motor. The main advantage to a larger pulley (platter) is the use of lower powered drive motors, which in theory run quieter and have less of a cogging affect on the platter. For instance, the motor on the AR XA is barely able to start the platter without a little assistance, if the belt were placed around the platter an even smaller motor could be used, or a heavier platter. Another reason why more modern cartridges with their lower tracking forces sound superior.

avanti1960 said:

agreed. and i was speaking in general terms. all things being equal a belt driven turntable would have more speed stability if the platter pulley were larger. the belt wrapping around the platter rim (example music hall mmf-7.1) would be more stable than the belt wrapping around a smaller diameter sub-platter (example, rega RP-*).

Click to expand...

Yes, I agree. Any eccentricity in the motor shaft or pulley is reduced by making drive pulley as large as possible. Once per revolution motor shaft rotation often shows up in pitch modulation spectrum IME.

Not sure if anyone's aware of my plots of pitch stability in terms of position of platter rotation? Once round the circle is one platter revolution, radius of plot is 'instantaneous' pitch for a 3150Hz test tone. Red and Blue lines are successive revoloutions. It's easy with this to identify factors which might contribute I find, but also inset spectral analysis which provides pointers and often throws up 'once per pulley rev' for belt drives. The example below is for a very small motor shaft, actually it's a rim drive in this case where an idler couples the shaft rotation to the platter, but similar principles apply. I think it's clear to see avanti1960's point here awprint:

avanti1960 said:

Speed control units tend to have the "hunt and peck" phenomena, where the constant monitoring and feed back in minute speed changes can show up audibly.

Click to expand...

Your equipment is malfunctioning.

Kingshead said:

....the motor on the AR XA is barely able to start the platter without a little assistance, if the belt were placed around the platter an even smaller motor could be used, or a heavier platter.....

Click to expand...

Although I've run analysis on many 3rd party fine TT samples, I've yet to see posted any 3150Hz test tone sample from an AR-XA. I'd be intrigued to run analysis and post it if anyone is willing to post a link to a short sample 10s or so of a 3150Hz or so test tone, preferably in .wav or .flac format? See what all the fuss is about, seems there's so many AR TTs about on this forum someone must be able to kindly post a test tone sample ?

awprint:

The general idea of open loop control with large time constants (which is the translation of the OP's idea in technical terms) makes perfect sense in the low tech world where design costs have to be kept low. However my years of learning electronics in a university utterly and completely contradict the idea that closed loop control is instable by nature. It's quite the opposite really, closed loop control is the very reason why rocket don't do loopings and modern airplanes fly more smoothly than the old manual-command crates, among many other applications. You don't achieve this stability with open loop, long time-constant control. The same goes for speed control, and there are many thick boring books out there that explain you how to do it right and take into account exterior influences (in this case: platter imbalance, needle drag & bearing friction) - and decades of experience taught us that the automatic control theory works. That said, I fully agree that common sense is not really satisfied with it, but then so isn't it with how your GPS or your colour TV works: you just have to admit it or to spend a few years at the university if you want to understand it.

Avanti1960, FWIW cheap/low-skill design and/or poor condition are the culprits for the speed instabilities in your turntable. In particular, there is no reason whatsoever for a DD motor to cog or "hunt and peck" as you put it, unless it is defective.

The smaller sub-platter pulley would have less available torque against drag, and be much more susceptible to speed variations of the motor pulley.

If this is a PM syncro motor driving the drive pulley, this does not apply.
The torque is delivered by the total mass of the platters. If the mass of the platters is equal in both systems, the torque will be the same at 33.33rpm.

jlb2 said:

The general idea of open loop control with large time constants (which is the translation of the OP's idea in technical terms) makes perfect sense in the low tech world where design costs have to be kept low. However my years of learning electronics in a university utterly and completely contradict the idea that closed loop control is instable by nature. It's quite the opposite really, closed loop control is the very reason why rocket don't do loopings and modern airplanes fly more smoothly than the old manual-command crates, among many other applications. You don't achieve this stability with open loop, long time-constant control. The same goes for speed control, and there are many thick boring books out there that explain you how to do it right and take into account exterior influences (in this case: platter imbalance, needle drag & bearing friction) - and decades of experience taught us that the automatic control theory works. That said, I fully agree that common sense is not really satisfied with it, but then so isn't it with how your GPS or your colour TV works: you just have to admit it or to spend a few years at the university if you want to understand it.

Avanti1960, FWIW cheap/low-skill design and/or poor condition are the culprits for the speed instabilities in your turntable. In particular, there is no reason whatsoever for a DD motor to cog or "hunt and peck" as you put it, unless it is defective.

Click to expand...

Yes I agree that properly designed servos offer the best chance for ironing out true platter speed stability, but there's not much hope when overall bandwidth of the servo is really low, as it is in most belt drive systems.........

IME the causes of pitch instability in descending order (biggest first)

Eccentricity of record/centre hole/spindle bent/spindle hole slack
Record warp/platter not level
Eccentric belt drive motor shaft/pulley eccentric/bent
Cartridge/arm resonance instability
Motor pole vibration
Suspended chassis vibration
Odd belt resonances

Far below these IME are the effect of needle drag (in a healthy TT), and mains frequency variation (synchro motors).

Servos in a belt drive can only have enough BW to take care of the top two if you're lucky, so I reckon. Attempting to deal with out of BW errors can only end in tears, I'm still with avant1960 on this one.
awprint:

marcmorin said:

The smaller sub-platter pulley would have less available torque against drag, and be much more susceptible to speed variations of the motor pulley.

Click to expand...

It's not a good sign of health if needle drag variation affects pitch stability at all, really. Any chance of a sample 3150Hz tone for an AR -XA, MM? You must have the most healthy examples, by account. awprint:

luckydog said:

Yes I agree that properly designed servos offer the best chance for ironing out true platter speed stability, but there's not much hope when overall bandwidth of the servo is really low, as it is in most belt drive systems.........

Click to expand...

I agree, belt resonance is a big difficulty. OTOH I am not aware of a belt-driven, closed-loop controlled turntable. I do know belt-driven turntables in which the motor speed is controlled (Thorens TD125 & TD126 being notable examples), but then the transmission to the platter remains in open loop and the speed consistency is entrusted to the design and build quality (main bearing and belt) so these are not really closed-loop controlled drives. The only fully closed-loop designs I know of are direct-drive, and the motor is designed to avoid motor pole vibration. In such a design the only remaining error sources are those related to the record/needle system, especially the record eccentricity (very difficult to take into account).

Also note that a closed-loop drive system the platter speed does not use a mains-driven synchronous motor since the motor speed command is controlled by the measured platter speed. Mains-driven synchronous motors are open-loop drives by definition.

jlb2 said:

The only fully closed-loop designs I know of are direct-drive, and the motor is designed to avoid motor pole vibration. In such a design the only remaining error sources are those related to the record/needle system, especially the record eccentricity (very difficult to take into account).

Click to expand...

Yes, in all my measurements of pitch stability DD offers best overall results I'd say, especially with higher inertia platters and this is what I use. And I agree there still remain sources related to record centring/warp, and cart/arm resonances which pretty much dwarf any mains, motor or drag variation in a healthy TT, typically IME. awprint:

luckydog said:

Yes, in all my measurements of pitch stability DD offers best overall results I'd say, especially with higher inertia platters and this is what I use. And I agree there still remain sources related to record centring/warp, and cart/arm resonances which pretty much dwarf any mains, motor or drag variation in a healthy TT, typically IME. awprint:

Click to expand...

When You checked my DD, if I recall correctly, there were only some small influences from record warps, some record eccentricity and som minor arm vibrations. Nothing were really seen that could be attributed to W&F, in principle.

gusten said:

When You checked my DD, if I recall correctly, there were only some small influences from record warps, some record eccentricity and som minor arm vibrations. Nothing were really seen that could be attributed to W&F, in principle.

Click to expand...

Hi gusten - yes IIRC you'd pretty much reached the point where all that was left was either in the record or inevitable and slight due to arm, and nothing notable could be attributed to platter W&F at all :thmbsp:

awprint:

luckydog said:

It's not a good sign of health if needle drag variation affects pitch stability at all, really. Any chance of a sample 3150Hz tone for an AR -XA, MM? You must have the most healthy examples, by account. awprint:

Click to expand...

The quote you list in my post was not mine. My post said if the platters weighed the same, and were going 33.3rpm they would have the same torque, regardless of being driven outside the main platter or being driven with belt on subplatter. Unless physics has changed when I was in school, energy is mass/velocity. Same mass, same velocity=same energy.