"Cogging is a function of the number of poles; the rotor experiences a
regularly irregular rotational force due to the naturally varying
intensity of the magnetic fields produced by the stator. The rotor is
therefore constantly inconstant in its speed. There is a regularity to
it that is said to be audible to some, and that's "cogging". The
tendency can be ameliorated by using a stator with a lot of poles, the
more the better." This is only partially correct. Cogging is caused by the change in variable magnetic reluctance as the PM rotor passes the metal pole pieces of the stator. Adding more poles does not decrease the frequency or amplitude of cogging. If you turn the motor by hand you will feel the cogging and it "feels" finer with a 24 pole motor vs a 12 pole motor because the cogs are closer together. A 24 pole motor turns at half the speed of 12 pole motor so the frequency of the cogs is identical in both (120Hz). The magnitude of the vibration caused by cogging will be identical in motors with identical power ratings and the vibration is directly proportional to the power consumed by the motor. In most cases, the 24 pole "upgrade" motor is higher power than the 12 pole motor it replaces, so the 24 pole motor will actually produce more cogging than the 12 pole motor it replaces. This was investigated in the link below: https://www.diyaudio.com/forums/analogue-source/309925-hurst-motors-300-rpm-vs-600-rpm-upgrade-myth....Coreless motors have no metal pole pieces in the stator windings so they produce no cogging. |
"Phoenix, Thank you for that explanation. Do DC motors cog at all?" Any motor with iron pole pieces will exhibit cogging. In a DC motor, the magnets are stationary but the rotor coils are wound on steel laminates with poles which also produces cogging. Some of the better BLDC motors have stators with skewed poles (angled slots rather than vertical) so the rotor sees more or less a constant magnetic reluctance and produce much less cogging. |
From my experience, the highest performance for either belt drive or DD is a 3 phase BLDC motor with the caveat that it is run as a 3 phase AC synch motor and not a DC motor (they can be operated both ways). If done right, they have little or no cogging, more torque than a comparable AC synch motor and the speed is determined by the frequency so speed control is fairly simple, though the drive circuitry is not.
|
Most of the direct drive motors (at least the better ones) are coreless so they have no cogging. An 1800 RPM motor is a 4 pole induction motor; the stator does not have the poles and gaps that an AC synch motor will have. When you turn an induction motor by hand there is no cogging because there is no permanent magnet in the rotor. The rotor is magnetized by the rotating field, but that requires a certain amount of "slip" to operate and is a function of the torque load. Because of this slip, induction motors are not truly synchronous which introduces another variable in speed control. Induction motors are speed sensitive to voltage as well as torque, where AC synch motors are unaffected by either. DC motor speed is affected by voltage, temp and torque load. As stated previously by others, there’s no free lunch and there are strengths and weaknesses to each design choice.
Stylus drag is fairly constant so in most cases, it is inaudible. While there is a measurable change in speed caused by stylus drag that varies from the start of a record to the finish, it is rather small and extremely slow changing and mostly inaudible. Changes in speed with groove modulation exist in theory, but I have never seen any data that claims to measure or quantify it; it must be extremely small. Heavy platters with lots of inertia will have a positive effect on this phenomenon, because changes in groove modulation are short duration (by definition) and more inertia will reduce short term speed variations.
In my experience, the biggest change in speed of BD tables is caused by the warming of the belt and bearing oil viscosity; it is not uncommon to see speed drift of 0.2~0.3 RPM over a 45 minute playing time which is audible to those with pitch sensitive hearing and if corrected all at once, audible to just about everyone.
|
"Dear friends: I would like to know if each single LP was recorded/cutted
at exactly/accurated 33.333..rpm and if for any reasons exist tiny
deviations from 33.333..rpm accuracy why or how can we or not detected
through an accurate TT that spins at exactly 33.333...rpm during play
time of LPs?" The most popular motor for the cutting lathes is a Technics SP02 direct drive motor which is quartz locked for speed accuracy and weighs ~110lbs with plenty of torque. W&F are rated at 0.0084% RMS. In terms of absolute speed accuracy, it should be more accurate (and more stable) than most of the DD tables playing the LP it cuts. http://pspatialaudio.com/lathes.htm |
"But I was a bit puzzled by your statement: "A 24 pole motor turns at
half the speed of 12 pole motor so the frequency of the cogs is
identical in both (120Hz)." In a direct-drive turntable, doesn't the
motor have to turn at 33.3333 rpm, regardless of the number of poles?
And therefore might there not be a theoretical advantage to having
double the number of poles?" My statement was in response to the assertion that adding more poles reduces cogging (it doesn't). Doubling the number of poles does not change the vibration signature because it will also cause the motor to turn at half the speed (assuming the same 60Hz drive signal). Both a 12 pole and a 24 pole motor will produce vibration from cogging at 120Hz (10 revs/sec x 12 poles at 600 RPM and 5 revs/sec x 24 poles at 300 RPM). The amplitude of the vibrations is directly proportional to the power consumption of the motor, not the number of poles. If you go to the link I provided, you will see that the measurements bore this out. The motor on a direct drive table does turn at 33.333 RPM so the number of poles will determine the drive frequency needed in that case (RPM=Freq x 60/pole pairs or Freq=RPM x pole pairs/60). If the motor uses steel pole pieces, it is subject to cogging and the frequency of vibration will be 0.555Hz x number of poles; in most cases, this would be below 20 Hz so it will show up as rumble.
To run a DD motor from 60Hz would require 216 poles.
From what I've seen, most of the DD tables use a DC motor with servo control (feedback) to maintain proper speed. This of course, comes with its own set of problems. |
"In my industrial world, traditional motor absolute speed control was obtained by either servo motors/ controllers or motors with speed encoder feedback.
However in past few years drive controllers have become so sophisticated that now best speed regulation is obtained by running " open loop" with zero encoder feedback and using current feedback at the controller itself.
Now is this a possibility for TT speed control or is this how some are already controlled?" It depends on the type of table. For belt drive (the LFT1 is BD), the motors are relatively high speed (300/600 RPM) with low inertia rotors so open loop speed control is possible. The SOTA Eclipse package does this with a BLDC motor run as a 3 phase AC synch motor and the motor speed is very stable and accurate. The RR tach is used to counter long term speed drift as the table warms up. Running a direct drive motor open loop is much more difficult because of the slow speed and high inertia of the platter. With no feedback, the platter speed will wobble considerably. Most of the DD tables that I’ve seen use a rotary encoder for speed feedback and a DC servo control to drive the motor. The VPI HW40 uses a magnetic ring encoder and drives the BLDC motor as a DC type using block (trapezoidal) communtation and Hall sensors. Because of the LPF in the feedback loop, the platter speed is still susceptible to oscillations, although with a heavy platter, it will move the oscillations lower in the audio band vs the light platter DD tables of the 70’s and 80’s. The HW40 does respond to variable drag on the platter, but it is quite sluggish, slowing down for ~250mS before compensation is applied and takes another 250mS to correct, so it does little to affect W&F. Current feedback is still feedback, but if implemented correctly, it can eliminate both the encoder and the delay in the feedback loop. Field Oriented Control (FOC) monitors the current in the windings and can compute the rotor flux position on a de-rotated frame of reference so the control loop operates at DC. The current control loop regulates the torque of the motor and a speed loop is wrapped around the current control loop (changing torque changes acceleration and therefore speed). The speed feedback comes from an estimator circuit that derives the back EMF signal (rotor speed) without external sensors and works well at low speeds as well as in the presence of high noise. I don’t know of any mfr that uses this method, but it would be interesting to try. |
"....why do you think VPI use trapezoidal commutation?
Surely this is not optimum if they want to minimize torque ripple." The controller is made by Elmo Motion Control (Gold Twitter Solo), and while Elmo's PC control software is fairly comprehensive, I'm not sure if the controller is capable of sinewave drive; it is not capable of sensorless operation (FOC) which would use sinewaves. Sinewave drive is considerably more complicated and would result in smoother operation, but for whatever reason, VPI chose to operate the motor in the simplest mode available, as a DC motor which uses block commutation and Hall sensors. They do some other rather strange things: The motor is designed for high power (500W) so the windings are 0.3R; not the easiest load to drive especially from a class AB output amp. The Elmo controller uses PWM output at 10kHz, which IMHO is a bad choice for use around a sensitive audio circuit like a cartridge pickup (especially MM). |
Unfortunately, I haven’t had the chance to "look under the hood" on very many of them, but based on my limited exposure:
Belt drive: American Sounds AS2000. David Karmeli not only understands high end audio, he gets the physics and engineering part right as well. His design philosophy may be unique in the industry. Full disclosure: I had a hand in designing the motor control for this table, based on the Eagle PSU and RR tach.
Direct Drive: I’ve not auditioned the new Technics tables and I don’t have a lot of technical info on them, but from what I can gather, they are doing some really sophisticated things in motor control and they have the engineering HP to do things right. It wouldn’t surprise me at all if they were doing FOC.
I’m basing my observations more on an engineering perspective than an audiophile one. YMMV.
|
@richardkrebs-
ThinGap does make a different series motor (LS51, same price ~$600 in qty.) with higher pole count (higher driving freq), lower power (48W) and higher winding resistance (4Ω); it is still coreless and would be much better suited for a DD table application the the TG23x series and could be driven from a class AB amp. It’s been available for a couple of years, so I’m not sure why it didn’t get used in this case?
The HW40 gets great press because it does sound good and even measures respectably, but no where near the numbers VPI claims for W&F, speed accuracy or torque. Of all the reviews I’ve read, I don’t believe I saw a single measurement; making matters even more difficult, you need the software from Elmo (you can it download free from their website) and need to construct a serial cable to connect a PC to the controller in order to get some of the relevant measurements. Luckily, the parts are readily available thru Digikey. With the software, you can also do 78 RPM or adjust the tempo for off speed records such as the Stones Beggar's Banquet, in steps of ~0.07%.
|
|
"ignore the technology and measurements.
i did......or i do every day." I never understood this position. While the sound produced by a component might be the only thing that matters to some people, perception is highly subjective and dependent on many external factors (mood, daily hearing condition, type of music, room acoustics, blood alcohol content, etc.). It is also personal to the individual and not easily or accurately reproducible or even communicated to others. To ignore science and technology when designing audio components is wrought with peril, and makes you a hobbyist, not a designer. Design decisions should be informed by established theory and practice, and the actual sonic results (and measurements) should confirm those choices and correlate with the design theories. The more connections between the two realms (rules of correspondence), the more certain you can be that the sonic results are accurate and not due to some subjective factor, bias or confusion of the senses. Measurements are used by manufacturers in their specifications and marketing materials to promote their products as "better" than the competition. Verification of these measurements prevents mfrs from making outlandish or misleading statements about their products and serves to protect consumers. |
A couple of corrections: I no longer sell audio components or make any money from designing them. This is purely a hobby for me as I'm interested in the technology (for now).
The 5 or 10% of your post I responded to was the only relevant part of your rather declaratory statement; the rest of your post had little or nothing to do with addressing your bold claim. I did a direct copy and paste quote, so I'm not sure how I twisted it?
Your last post does a much better job of explaining your position regarding the proper role of measurements vs listening. Thanks for clarifying.
It's certainly not my intention to be inconsiderate towards others who may not fully understand the technology, quite to the contrary, my intention is to help educate. If someone has questions, I'd invite them to ask or to do a google search (it's how I find a great deal of the information I seek).
|
