I therefore asked Mr. Moncrieff the authorisation to publish his article in my site. Here it is. If you want to read the original version, split in more pages, you will find it in the pages of the excellent IAR site.

Rockport Sirius III

If you love vinyl, you want this turntable.

This is a whole new kind of turntable for Rockport. This is a whole new kind of turntable for anybody. This is a whole new kind of turntable, period.

The most essential job for a turntable is to spin your record at a constant speed. Really constant. This is the first turntable to do it right. Really right.

In order to fully understand what this turntable does right, we first need to candidly reveal all the things that other turntables, in contrast, do wrong. Candid revelation requires detailed analysis, so that we don't gullibly swallow the sales hype of past turntables, whose broad generalizations gloss over the details which constitute that little thing called the truth. We need to penetrate beneath the surface gloss of sales hype.

But every cogent analysis, even a detailed one, should be logically organized, and based on a simple conceptual foundation as a key starting point.

Our conceptual foundation, our key starting point, is simple. Speed. Speed is the last great hurdle in turntable design, and no one has really conquered this hurdle. Until now.

Why is speed so important? As you know, the primary job of any record player, including turntable, arm, and cartridge, is to accurately reproduce the waveform of the music as it was originally recorded onto a vinyl record.

But how exactly is this job divided up among the turntable, arm, and cartridge? Most people would say that the cartridge has the job of reproducing the entire music waveform, and that the turntable and arm have lesser passive roles, merely responsible for being stable platforms for the record and cartridge, so the cartridge can do all the active work, reading the entire music waveform from the record.

That's wrong. The cartridge does not read the entire music waveform from the record. It can't. Why not? Because the vinyl record contains only half of the music waveform.

Where then does the other half of the music waveform come from? It comes from the turntable. That's right. The turntable is fully responsible for actively supplying half of the music waveform, and the other half comes from the cartridge.

This puts turntables in a whole new light. If a turntable's job is to actively supply half of your music's waveform, then it had better be doing its job right, otherwise your music will obviously and dramatically suffer -- to an extent you wouldn't have imagined when you thought that music's entire waveform came from the cartridge.

What do we mean by saying that half the music waveform comes from the turntable? You know that music's waveform can be plotted on a graph or on an oscilloscope. The graph of a music waveform has two axes, as do most common graphs. The vertical axis represents amplitude, and the horizontal axis represents time. The music waveform needs both these axes to exist, since it is a two dimensional entity in nature. If one dimensional axis or the other were somehow missing, there couldn't be any music waveform, and there couldn't be any music. Music itself is a time bound art form, and its existence depends on the two dimensions of time and amplitude as surely as a sculpture depends on three spatial dimensions (but not on time in the case of static sculpture).

In a record player, the vertical amplitude axis of the music waveform comes entirely from the cartridge. But the horizontal time axis of this music waveform comes entirely from the turntable.

The vinyl record groove contains an analog of the vertical amplitude axis variations of the music waveform. These amplitude variations are laid out along the length of a long linear groove (that happens to spiral). These amplitude variations are laid out along this groove length with the assumption (repeat: ASSUMPTION) that the horizontal time axis for the music waveform will later be supplied by a turntable rotating the vinyl disc at a constant speed. But the time axis for the musical waveform is not (repeat: NOT) encoded into the vinyl disc itself. The disc groove contains only the amplitude axis variations, so it only contains half the information needed to recreate or plot the original music waveform on a graph. The disc groove contains not the slightest clue about what the horizontal time axis of the music waveform is supposed to be (as you know if you've ever played back a 45 rpm LP at 33 rpm). There are no guiding ticks imprinted into the groove at a constant timing rate (say 1 per second). There are no discrete CD pits to guide the turntable platter to spin at the correct rpm.

It's like a gentleman's agreement. The record manufacturer actually gives you only half the music waveform in the groove, for your cartridge to read, namely the vertical amplitude axis. You agree to supply the other required half of the music waveform, the horizontal time axis, by agreeing to employ an accurate turntable to play back the record manufacturer's disc. The turntable that you chose to employ literally supplies the time axis half of the music waveform, while your cartridge reads the amplitude variation half (and only that) furnished by the record manufacturer, as the groove is passed underfoot by your turntable recreating the time axis half on the fly.

Consider the following analogy. Imagine first that you want to draw a music waveform, like the ones you've seen in previous IAR articles. Draw it on a square piece of graph paper. Note that you can freely move your hand in two dimensions on the graph paper, so you can simultaneously draw both the varying amplitude (height) and progressing time (horizontal axis) of the waveform on the graph paper. Next, imagine that you're doing the same thing, but you've turned the piece of graph paper sideways, so that your wrist moves from side to side (instead of up and down) as you're charting the waveform's amplitude variations.

Now, imagine that you can only move your wrist from side to side, and can't move your hand up and down at all. Your hand holding the drawing stylus has now become just like a phono cartridge holding a stylus that can only read the side to side variations in a record groove. Your hand holding the waveform drawing stylus is mounted on your arm, the same way that a cartridge holding the waveform reading stylus is mounted on the pickup arm of the record player.

If you were to try drawing a music waveform, while limiting your hand to only this side to side motion, you couldn't do it. There would have to be a further mechanism for moving the drawing stylus in your hand along the time axis of the graph paper where you want to draw the complete music waveform. You could for example rely on a strip chart recorder, which could dispense the graph paper in strip form at a fixed time rate (you've probably seen strip chart recorders in the form of earthquake recorders, where the side to side needle motion indicates earthquake amplitude, on a steadily unrolling strip of graph paper; if you're unacquainted with this, imagine a roll of toilet paper unrolling at a steady rate). The strip chart recorder makes the graph paper move along under your hand at a constant speed, thus creating a steady time axis for the waveform you wish to draw. And the strip chart literally creates this time axis. Your hand is limited to reproducing (accurately we hope) only the amplitude axis of the music waveform, since you are now limited to side to side motion.

That's exactly what a turntable does. It literally creates the time axis half of your music waveform, while the cartridge, which is restricted to side to side motion, reproduces (accurately we hope) the amplitude information that the grove contains.

What happens if the turntable speed is inaccurate in any way (momentarily or over the long run)? Let's go back to the analogy of your drawing the music waveform on a strip chart recorder. If the strip chart recorder fails to move the paper under your drawing hand at a precisely constant speed (and the correct constant speed), what would happen? The waveform you drew of the music would be distorted! For example, your hand might move side to side to the correct amplitude position for a musical peak, but if the strip chart recorder moved the chart paper a bit too fast, then your hand would draw that musical peak at the correct amplitude but in the wrong location on the time axis. The peak would come too soon on the resulting graph you drew on paper of the music waveform. How would this be a distortion? The portion of the music waveform preceding the peak would be compressed or squashed, into a too short time period. When you took your graph paper out and looked at the music waveform as a whole that you had just drawn, it would not be correct. It would be squashed horizontally. It would be a distorted, inaccurate version of the music waveform, as surely as if you were to have drawn the incorrect amplitude by moving your hand side to side in an inaccurate manner.

The key point here is that an error in the time axis would produce a distorted error in the correct shape of the music waveform, as surely as if your hand were to have erred in a distorted way in reproducing the correct amplitude swings of the music waveform. We give a great deal of attention to making sure that cartridges are accurate reproducers, so that they correctly read the side to side swings of the record groove that furnish the amplitude information about the music waveform, and thereby do not distort the music waveform themselves. But that's literally only half the story. We should also devote equal attention to making sure that the turntable accurately furnishes the time axis half of the music waveform. If we don't, then the final resulting music waveform will be distorted, as surely as if the cartridge were contributing unwanted distortion by inaccurately reading the amplitude axis half in its side to side swings.

The lesson is clear. You could buy the world's most expensive, most perfect cartridge, that exhibited perfect accuracy in reproducing the amplitude half of the music waveform from its side to side swings in tracing the record groove. But, unless your turntable is perfect in creating the time axis half of the music waveform, the final music waveform you hear will be distorted. The right amplitude played at the wrong time will distort the music waveform as surely as the wrong amplitude played at the right time.
We discussed this same lesson in IAR Hotline, in conjunction with digital. We saw there that jitter (also a simple timing error) distorted the final music signal from a digital system, as surely as if some bits of resolution had been lost in decoding the amplitude of the music waveform. We saw that jitter, though a simple timing error, did not merely produce timing errors (such as frequency or pitch variations) in the final music signal, but instead actually introduced distortion of the music waveform, especially modulation distortion, and in particular frequency modulation distortion. This distortion would be apt to make music sound fuzzy, defocused, smeared, dirty, grundgy, frazzled, etc., while degrading transparency, stereo imaging, clean purity, etc.

In the past, most people have assumed that any speed errors in a turntable would be audible only as pitch errors, making the music sound off pitch or at worst slightly wobbly in pitch. But turntable speed errors also have other sonic consequences, which are far more pernicious. By playing the right amplitude at the wrong time, turntable speed errors create a distorted music waveform, even if the rest of your system were to be perfect. Indeed, as cartridges get better and better, becoming far cleaner and more accurate in tracking the amplitude half of the music waveform from the groove, and as the rest of our system chain continues to become at once cleaner and more revealing of everything (including not only the music but also distortions from our program sources), the waveform distortions due to turntable speed errors become more noticeable, and become the last remaining hurdle of the state of the art.
Moreover, the distortions due to turntable speed errors could actually be more pernicious than those due to cartridge imperfections. Cartridge distortions tend to be amplitude distortions, thus producing harmonic distortion (some of which is actually psychoacoustically benign) and amplitude modulation distortion (due to the fact that music, being complex, contains many frequencies at once). But turntable speed errors predominantly produce a different kind of modulation distortion, called frequency modulation distortion. Pioneering work by Paul Klipsch 50 years ago already found that this FM distortion is more pernicious than AM distortion, being more audible and more objectionable in smaller amounts.

If FM distortion is present in very large amounts, and with certain slow modulating rates, it might be heard predominantly as pitch variations, or wow and flutter. But FM distortion in much lesser amounts can still create sidebands, distortion byproducts, around each musical note. These unwanted sidebands could make the music sound fuzzy, defocused, smeared, dirty, grundgy, frazzled, etc., while degrading transparency, stereo imaging, clean purity, etc.

FM distortion from a turntable takes what should be a purely precise time line ordering and spacing of musical events, and instead applies a different time line, one which variously contracts and expands like the pleats of an accordion in motion. The waveform is distorted so that some musical peaks are too close together, while others are spaced too far apart; some waveform valleys are too wide, while others are too narrow; some peaks themselves are too wide, while others are too narrow. You can imagine what this might for example do to the sound of a pure, clean, simple trumpet note. A trumpet note basically looks like a series of identical sharp spikes, spaced apart from each other by the same exact time intervals. However, turntable FM distortion would make these spikes have different widths or thicknesses. The too thick spikes might sound darkly grundgy and garbled, while the too thin spikes might sound brightly frizzy or edgy. Thus, you'd already have a trumpet sound that is at once too grundgy and also too frizzy, with both dark and bright distortions, instead of a pure, clean, consistent musical note. Furthermore, turntable FM distortion  would place some of these spikes too close together while others would be too far apart. This would probably make the trumpet note sound smeared, defocused, and closed in, with degraded intertransient silence, instead of sounding pure, clean, articulate, and open, with a black silent background. Incidentally, the same effect at lower frequencies, where a bass note depends on a repeating pattern for its sonic quality, could explain why some turntables (and CD transports with close in jitter sidebands) have problems playing bass notes with steady pace, timing, foot tapping rhythm, tight slam, etc.

A turntable with truly accurate speed could allow your music to sound more transparent, clean, pure, articulate, individualized, pristine, delicate, open, and airy, with better intertransient silence and better stereo imaging.

How can one make a turntable have more accurate speed? What's wrong with the speed accuracy of all other turntables to date?

We have basically had just two kinds of turntable drive systems over the years, rigid drive from a multipole motor and elastic drive.

Rigid drive has the disadvantage that it directly couples speed variations in the motor to the turntable platter. All multipole motors have speed variations. A typical turntable motor might have 24 poles. This configuration gives the motor 24 discrete power kicks per revolution of the motor. Between power kicks the motor just coasts, and thus slows down. Therefore, if the average speed of the motor is to be correct for a whole revolution, it has to speed up to faster than correct speed during each kick, to compensate for the speed it loses during each coast between kicks.

Remember the foot driven merry go round you played on as a kid? Every once in a while you'd reach down with your foot to give the merry go round another kick, to keep it going. The average speed of your merry go round was (by definition) between the higher speed you injected with each kick and the slower speed it coasted to that made you realize it needed another kick of speed added. That merry go round had significant speed variation, and so does a multipole motor, which works on the same principle. Incidentally, note that if you try to maintain the merry go round at some given average speed as a correct target speed (say 33 rpm), then the kick method is hopelessly inadequate. That's because over 99% of the time the merry go round will be going at some wrong speed, too fast or too slow, having either just been kicked to some higher than correct speed, or having coasted past the correct speed down to some slower speed before it gets the next kick. Indeed, the merry go round will be going at the correct speed only momentarily in passing, during each kick and coast cycle.

Likewise, turntables with rigid drive from multipole motors are almost never at the correct speed. They kick up above the correct speed and then coast down to below the correct speed. This kicking, cogging problem applies regardless of the particular rigid drive system, and regardless of the number of commonly used poles in the motor (2 to 24). For example, rigid drives with rubber puck reduction coupling have commonly used 2 or 4 pole motors rotating at 360 rpm, which works out to yielding about 22 or 43 kicks per revolution of the platter, while direct drive motors have commonly used 24 poles, which of course yields 24 kicks per revolution of the platter.

It's disconcerting to know that your turntable is actually almost never at the correct speed if it uses this kick and coast drive mechanism. Who wants the pitch of his music to be constantly shifting? But as discussed above, the sonic degradations from this mechanism reach far beyond shifting pitch. The kick and coast mechanism is operating at a basic repetition rate of about 13 Hz in a direct drive turntable with a 24 pole motor (about 24 Hz in a puck system with a 4 pole motor). This means that the kick and coast mechanism distorts your music, creating spurious FM distortion byproducts as sidebands, spaced 13 Hz (or 24 Hz) away from each and every musical note, from every musical nuance, that your record player is trying to reproduce from the disc. What's worse, there are also harmonics of these sidebands, since the kick and coast mechanism is not perfectly sinusoidal.

This is a very nasty picture of what a turntable can actively do wrong to mess up your sound. The turntable, far from being a passive carrier of the disc containing all the musical information, is totally responsible for creating the time axis for your musical waveform. And, if it fails to do this perfectly (by exhibiting speed irregularities), then it literally creates distortion byproducts that spring up like weeds amidst all your music. These distortion byproducts themselves sound dirty, and they also smear your musical information, filling in what should be a black background of intertransient silence between musical details, with the garbage of their added energy, and thereby degrading the individuation of each musical nuance. How can you enjoy the pure beauty of individual roses against a background of black earth when there are weeds sprouting up all around the roses, cluttering the view and obscuring the black earth background? That's why most direct drive turntables and puck drive turntables make the music sound grundgy and veiled.

It's worth noting that this kick and coast problem exists even before we turn to further problems in some turntables of servo error, servo correction, and servo drift. These further problems can make the speed irregularities even worse. But servo correction systems cannot cure the kick and coast problem of multipole motors, for the simple reason that the correcting servo cannot deliver the speed correction by any mechanism other than a kick of a motor pole; even the fastest, most perfect servo correction system has to wait until the next motor pole kick in order to actually deliver any speed correction.
It's also worth noting that turntable rumble, even if spec'd to be very low or at an inaudibly low frequency, can contribute to FM distortion of your music by a turntable. Some vector component of rumble (and indeed of any vibration) is bound to be along the groove line direction, which defines the time axis of the music waveform being recreated by your turntable and cartridge. Thus, this rumble or vibration vector component modulates the time axis of your music waveform, creating FM distortion sideband byproducts around all musical notes, as surely as if the speed of the motor were changing in sympathy with these vibrations. That's why some direct drive turntables with very low claimed rumble "within the audible spectrum above 20 Hz" might still make music sound grundgy and veiled, since they are literally distorting music, creating modulation distortion sidebands spaced less than 20 Hz from every musical note (plus harmonics spaced further away).

The speed variations (and so also the distortions) introduced by this kick and coast problem can be ameliorated somewhat by making the platter rim heavier (yielding a higher moment of inertia), but only up to a point, and unfortunately that point is still a compromise. The speed variations can be reduced by making the platter's moment of inertia high relative to the strength of the motor pole kicks, so that each motor kick doesn't disturb the platter speed as much. But, if the turntable designer makes the platter massive enough so that each motor kick can't significantly disturb the platter speed, then the turntable will take forever to get up to 33 rpm from a standing start -- precisely because each motor kick can't significantly disturb the platter speed (i.e. it can't disturb it enough to get it up to 33 rpm from a standing start). This puts the turntable designer in a hopeless compromise dilemma. Look at his dilemma from another angle. As soon as he makes the platter light enough, relative to the strength of the motor kicks, to be able to get the speed disturbed all the way up to 33 rpm from zero by the motor kicks in a reasonably short amount of time, then the platter is already more than light enough for the same strength motor kicks to disturb its speed a much smaller amount (say from 33 to 34 rpm) in a much shorter amount of time.


Automodulation Distortion

The turntable designer might instead try to deliberately use a very weak motor, with very weak kicks. But that still doesn't solve the above compromise dilemma (he'd still have to move to a lighter platter, if he wants reasonable startup times). And it also creates a new problem, relating to automodulation distortion.

What's this new distortion? Turntables get slowed down by outside forces, such as the drag of the stylus in the groove. This drag changes as the groove angle changes with large amplitude music signals. The more that the stylus is yanked side to side by a larger groove excursion representing a louder music signal or transient, the more its passage is impeded in the direction along the linear vector of the groove travel representing the time axis.

What's doing the impeding? The turntable is moving the groove under the stylus in a linear direction along the time axis. If there's zero music signal amplitude, the groove is relatively straight, and the groove walls are parallel to this direction of travel, and so these groove walls offer minimal resistance to the stylus gliding past them. But when there's large signal amplitude, then the groove swings wildly from side to side, so the groove walls are more nearly perpendicular to the ultimate direction of the stylus' travel, which is the time axis direction. Thus, the stylus slams into the nearly perpendicular groove wall, which naturally offers more impediment to its travel in the time axis direction. Think of it as the difference between gliding along on an ice rink, parallel to its surface, versus slamming into an ice wall perpendicular to your direction of travel. Pretty dramatic difference, right?

What happens when the stylus slams into the groove wall? Who gives way? Well, the stylus might be smaller than the groove wall, but he's a sturdy little bugger, being made of diamond and being anchored firmly to the pickup arm fixed on the plinth. So he hardly budges longitudinally, in the time axis direction (although he's free to swing from side to side, to track groove modulations). Instead, it's the groove wall that yields. Its soft vinyl gets momentarily deformed. As the compressed vinyl reacts to and springs back from this deformation, some of the energy gets transformed into heat, while another part causes the groove wall to partially shudder to a stop, i.e. to slow down. Thus, the groove wall slamming into the immovable stylus, at a nearly perpendicular angle to the groove's time axis motion, causes the groove wall itself to recoil from this collision and slow down in its motion along the time axis direction. When the groove slows down, of course the whole record slows down and the whole turntable platter slows down.

In sum, when the music gets loud, the larger side to side groove modulations become more nearly perpendicular to the stylus' steady travel along the time axis direction of the groove. The groove wall has acquired a vector component that effectively collides with the stylus, instead of gliding along parallel to it. Along this time axis direction, the stylus position is fixed, so it is the groove wall that loses in this collision, and the groove (hence turntable) slows down in the time axis direction.

This means, quite simply, that loud music slows down your turntable.

But any variation in the speed of the turntable is a distortion of the time dimension itself, which is one of the two axes defining the music waveform (the other being the instantaneous amplitude, furnished by the groove's side to side modulations). Distort the time dimension and you distort the music waveform, as surely as if you had distorted the amplitude dimension.

Therefore, loud music makes your turntable distort the music. Loud music distorts itself, when played on most turntables. The type of distortion is still FM distortion, but it is a particularly cruel joke because the music is causing its own degradation. Thus this is automodulation distortion. Incidentally, there is a similar phenomenon in some digital systems, where the amount of and nature of jitter (which ultimately causes distortion) is correlated with the amplitude and nature of the music itself -- again, music causing its own distortion.

With a turntable, an increase in the amplitude dimension of the music waveform causes a distorted stretching of the time dimension of that same waveform. The time axis of the music waveform is modulated by what is happening on the other (amplitude) axis, instead of ticking away steadily, constantly, and independently on its own.

Note that this automodulation distortion is different from the more ordinary type of nonlinear distortion, wherein cartridges, amplifiers, speakers simply become more nonlinear as signal amplitude gets larger. In that ordinary type of nonlinear distortion (AM distortion), the amplitude axis directly becomes distorted at various signal levels, due to the nonlinear amplitude response of some system. But this automodulation distortion is more sneaky and indirect. Here, it is the time axis rather than the amplitude axis which gets distorted at various signal levels. And it is only indirectly that distortion of the time axis leads to distortion (via FM distortion) of the music signal. Even though it is an indirect distortion, it is just as real, and in fact is even more pernicious, since FM distortion sounds uglier and is more detectable and objectionable in smaller amounts than AM distortion (some of which can even sound benign or actually even euphonic, such as second harmonic distortion).

How can the turntable designer keep the platter from slowing down every time the music gets louder, thus causing FM automodulation distortion? The only available tactic is to make the platter heavy, with a large moment of inertia, so that its large angular momentum tends to keep the groove going at a constant speed along the time axis. But this then implies that the turntable must have a relatively powerful motor with relatively fast response, in order to adequately power a heavy platter, and in order to restore its speed losses due to stylus drag within the same fast time frame that music's louder transients occur.

And this in turn means that one whole turntable design approach, weak motor with light platter, is ruled out. This design approach might be one way to address the multipole kick and coast problem, but it cannot adequately address this new problem of automodulation distortion.

So that puts us back with a heavy platter, a powerful motor, and a relatively fast motor response. But this also puts us squarely back into the dilemma of the multipole kick and coast problem. As soon as we make the platter heavy enough to avoid automodulation distortion, then we also have to make the motor powerful enough so that the periodic pulses of its kicks will be felt.

As if this weren't bad enough, there's also a problem of how fast to make the servo correction circuitry that drives the motor. After learning about automodulation distortion, we now see that we have to make the motor response fast enough to respond promptly to heavy music modulation within the time frame of music. But if a powerful motor is fast enough to respond to music, then it is also fast enough to directly intermodulate with the frequencies of music. Thus, we are now faced with yet another new dilemma in designing a turntable. We can make the servo correction fast to avoid automodulation distortion, but then we run the risk of directly intermodulating with musical frequencies. Or we can make the servo response slow, to avoid directly intermodulating with musical frequencies, but then the turntable will always be slowly drifting and hunting in speed, hardly ever being right on target (as is indeed the case for many direct drive turntables).

And, in either case, regardless of whether we make the servo response fast or slow, we still must face the basic problem that any multipole motor can only deliver its power via discrete kick pulses. So we're still stuck with the basic kick and coast problem.

Where does that put us? It appears that multipole motors are the devil's own instruments, distorting music from analog turntables because of their kick and coast problem, so long as the motor is rigidly coupled to the platter (via direct drive or indirect but rigid puck or idler drive). We've explored a number of possible turntable design approaches (light vs. heavy platters, weak vs. strong motors, fast vs. slow servo loop correction), and nothing really cures the basic problem of distortion and smearing that the multipole motor brings to the half of the music waveform for which it is responsible, the time axis dimension.

Clearly, we need help, big time. And, just in the nick of time, here comes our purported savior out of left field, riding a white horse. It's the elastic belt to the rescue! The elastic belt furnishes an elastic, non-rigid coupling between the motor and the turntable platter. The elastic stretchiness of the belt acts as a filter, filtering out higher frequency vibrations, so they are not transmitted from the motor to the platter.

Different belt materials and constructions have differing amounts of elastic stretchiness along their length direction. Some are relatively stiff and unyielding when you try to stretch them along their length, while others stretch easily. Generally, the easier a belt stretches elastically, the better it is at reaching to lower frequencies to filter out unwanted vibrations. By making a belt sufficiently elastic, we can make it reach down to the frequencies of a multipole motor's kicks, and filter them out. Incidentally, once a belt has been made elastic enough to filter out these low frequencies, it usually pays to also make it elastic enough to also filter out the motor's rumble vibrations, which start at the motor's primary rotational frequency. Belt drive turntables still use multipole motors that put out the dreaded kick and coast problem, but the belts are generally tuned to filter out these multipole problems.

The kicks from the multipole motor may be regarded as unwanted vibrations, which we don't want the platter to feel. A belt of sufficient elasticity filters out vibrations above a certain frequency, which could include all frequencies from the multipole kicks (including the fundamentals of say 22 Hz or 43 Hz as above, plus higher frequency overtones). If the platter can't feel the vibrations of the kicks, then it also can't feel the temporary speed increases that those kicks would tend to induce. Thus, the platter wouldn't feel the kick and coast problem from the multipole motor, if the energy from the motor is transmitted through an elastic belt. If (thanks to the elastic belt) the platter doesn't feel the kick and coast speed variations that the multipole motor is still putting out, then its speed will tend to remain more constant. And, if the platter speed remains more constant, there will be less distortion and smearing of your music signal, because a full half of the waveform will have been created more accurately by the turntable. That, in a nutshell, is why most belt drive turntables sound better than most rigid coupled drive turntables. The turntable is responsible for recreating half of your music waveform, and the belt drive can recreate this half more accurately than a rigid drive, because it can filter out the dreaded kick and coast problem inherent in every multipole motor.

However, on closer inspection, the elastic belt is not quite the savior on a white horse that it purports to be. He's more like Don Quixote riding an old grey mare one step away from the glue factory. In short, the elastic belt brings with it a whole fresh set of problems and engineering compromises.
Past articles published on belt drive have oversimplified the benefits, and have underplayed the problems. If you ask a typical belt drive turntable designer or technical proponent what the elastic belt achieves, chances are he'll say that the filtering action of the belt averages out the speed irregularities of the multipole motor's kick and coast problem, thereby sending a constant speed drive to the platter. That's a true statement, generically speaking, as far as it goes. But it doesn't go nearly far enough. It only scratches the surface of the way that the elastic belt really works, and thus it fails to reveal the many problems and tradeoff compromises. It gives a rosy generic oversimplified summary description of what a belt is ideally supposed to do, but it ignores the details of what a belt actually does.
We'll have to dig deeper, if we are to understand why belt drive turntables are compromised. What does a turntable belt do, actually, not just generically?

The motor of a belt drive turntable has a pulley, with a certain circumference. The platter of the turntable has a rim (perhaps tucked underneath the platter), with a certain circumference. The ratio of the pulley circumference to the platter rim circumference will determine the rotational speed of the platter, as a fraction of the rotational speed of the motor. For example, if the multipole motor had a rotational speed of 333 rpm and the ratio of the circumferences were 1:10 (the platter rim being 10 times larger than the motor pulley), then the rotational speed of the platter would be 33.3 rpm.

Thus, the turntable platter speed depends upon the speed of the motor (naturally), and upon the ratio of these two circumferences (also known as the reduction ratio). And, as we now know, the turntable speed is fully responsible for recreating half of your music waveform, the time axis dimension. So, the time axis half of your music waveform depends on the turntable platter speed, which in turn depends upon the ratio of these two circumferences. Therefore it follows that the time axis half of your music waveform depends directly on the ratio of these two circumferences.

In particular, if we want the time axis half of your music to remain stable, unwavering, and undistorted, then we also want the ratio of these two circumferences to remain fixed, stable, and unwavering. So far, so good. That seems clear, straightforward, and obvious enough.

Now, where does the belt figure into all this? The belt moves around the circumferences of both the motor pulley and the platter rim. It transmits the rotational motion of the motor pulley to the platter. Specifically, the rotating motor pulley circumference engages and moves a certain length of belt, say 1 inch worth, and then the moving belt in turn engages and moves 1 inch worth of the platter rim over the same period of time. The belt's primary responsibility is to faithfully communicate 1 inch's worth of circumferential travel by the motor pulley into exactly 1 inch's worth of circumferential travel at the platter rim. This is how the ratio of the two circumferences is communicated between the motor and the platter. For example, if 1 inch represents the complete circumference of the motor pulley, then (assuming our 1:10 ratio) 1 inch would represent 1/10 of the circumference of the platter rim, so a complete rotation of the motor pulley would produce exactly 1/10 of a complete rotation of the platter, and thus the platter would rotate at exactly 1/10 the rotational speed of the motor. Note that the belt is responsible for accurately maintaining a constant, unwavering ratio of motion between the pulley and platter rim.

As a visualization exercise, imagine that a 1 inch portion of the belt's length is painted a different color, say red. Ideally, that red 1 inch portion of the belt's length will be driven by a matching 1 inch portion of the motor pulley circumference, for a given period of time we'll call t. As the belt moves, this red 1 inch portion will gradually travel over toward the platter rim, travelling at whatever speed the belt happens to be moving. Now, the belt's primary responsibility is to ensure that this red portion engages and moves exactly 1 inch worth of the circumference of the platter rim, and accomplishes this in exactly the same duration of time t. If the belt succeeds in doing this, then it will faithfully communicate the correct ratio of circumferences and the correct ratio of speeds, and then the belt drive turntable will have a chance to run at the correct speed, and accurately recreate the time axis half of your music waveform.
But, if the belt cannot fulfill this primary responsibility, then the belt drive turntable is doomed to run at incorrect speeds, and is doomed to distort your music by incorrectly recreating the time axis half of your music waveform. These music distortions will be different than the distortions brought on by the multipole motor's kick and coast problem, discussed above for turntables with rigid drive coupling. These music distortions will be brought on by the belt itself, failing to act ideally. So already we can see that a belt, introduced to solve one turntable problem (kick and coast), might bring with it a fresh new set of problems, if it fails to act ideally.

Why might a belt fail to act ideally? There are many reasons. Most designers of belt drive turntables aren't even aware of all these problem areas for belts, so naturally their turntable designs haven't even addressed these problems, and thus produce haphazard performance results. Even the best designers of belt drive turntables, fully aware of all these problems, have to tiptoe through a minefield of engineering tradeoffs and compromises. They have come to know, all too well, that belt drive brings a whole new area of problems, and is not simply a panacea for the ills of rigid drive coupling.
To start with, let's just scratch the surface by merely mentioning two problem areas where things might go wrong with belt drive. First, the belt might slip slightly in going around the smaller diameter, faster spinning motor pulley. Then the 1 inch red portion of belt would not correspond to exactly 1 inch of the motor pulley circumference. And then, when this red 1 inch portion of belt arrived at the platter rim and drove exactly 1 inch of the platter rim circumference, it would not be faithfully communicating an accurate speed ratio between the motor and the platter, so the platter would rotate at the wrong speed.
Second, the belt might physically distort along in its length dimension, so that the red portion is not precisely 1 inch long at all times. How could a belt possibly distort along its length dimension? Simple. Remember that the whole point of using a belt in the first place was to exploit its elasticity along its length dimension, to absorb and filter out the kick and coast vibrations and speed irregularities of the multipole motor. An elastic belt is elastic, by definition, precisely because it is willing to distort its length dimension in response to applied stretching forces.

If the belt distorts its length dimension at all, then we can't count on any given 1 inch red painted portion staying precisely 1 inch long at all times. Suppose, for example, that the belt stretches for some reason as it goes around the motor pulley, so that the red painted portion stretches to become 3/2 inch long. Then this red painted portion leaves the motor pulley and starts on its journey toward the platter rim. Suppose that the stretching forces are no longer present for this part of the journey, so that the red painted portion resumes (shrinks back to) its normal length of 1 inch. In this scenario, the belt would incorrectly, unfaithfully communicate what should be equal circumferential travel between the motor pulley and the platter rim; it would fail in its primary responsibility of communicating and translating 1 inch of circumferential rotation by the motor pulley into 1 inch of circumferential rotation at the platter rim. In this scenario, it would have required 3/2 inch of circumferential rotation at the motor pulley, to produce 1 inch of circumferential rotation at the platter rim; or, to put it another way, 1 inch of circumferential rotation by the motor pulley will produce only 2/3 inch of circumferential rotation at the platter rim. Thus, the platter rim would be driven at only 2/3 of the correct speed, thanks to the mere fact that the elastic belt stretched its length dimension for a little while, at some point, along its journey.

The belt is wholly responsible for translating motion from the motor pulley to the platter rim, and thus becomes wholly responsible for the speed (including speed constancy) of the platter. If you look at the belt in action, you'll see it moving along its length dimension, as a continuing parade of 1 inch portions (one we painted red, the rest not) marches by. The speed of this motion gets imparted to the platter, to become the speed of the platter and the speed of the record groove, hence the time axis dimension of your musical waveform.

Thus, the belt can be viewed as a literal metaphor for the time axis dimension of the music signal. As you watch the belt length move by, so are you also watching the progressive unfolding of the time axis dimension of the music you are listening to. For music to be rendered clearly and cleanly (distortion free) by the turntable, we know that the time axis dimension must unfold at a perfectly constant and steady speed. This essentially requires that, as we watch the belt moving by, we can see that it is moving at a constant speed along its length.

So far, so good. Most turntable belts probably seem to move at a pretty constant speed. But that's not all that is required. As we saw just above, the belt length as such must also remain perfectly constant, without elastic stretching. If it stretches, then 1 inch of circumferential motion at the motor pulley no longer becomes exactly 1 inch of circumferential motion at the platter rim, which messes up the reduction ratio, which messes up the turntable platter speed.

The solution to this problem might seem easy: simply employ a belt that doesn't have elastic stretch along its length dimension. Indeed, some string drive turntables approach this design extreme. However, there's a big fly in the ointment. Remember that the whole point of introducing the belt drive concept was to filter out the kick and coast problem of multipole motors. A belt that doesn't have elastic stretch can't do this. So then what's the point of introducing belt drive in the first place?

The only way that a belt can filter a multipole motor's kick and coast problem is by distorting its length dimension, by elastically stretching and then collapsing. In other words, the only way that a belt can cure the speed inaccuracies of a multipole motor's kick and coast problem is by introducing its own new inaccuracies, which arise from its distortions of its length dimension. From the frying pan into the fire?

Let's take a moment to take a micro-look at the way in which belt stretching speed inaccuracies develop, so we can appreciate the many dimensions of these gremlins and their intractable nature.
Imagine you're holding a rubber band between your two hands, so that it is fully extended but not yet stretched. The rubber band is now at (or just a hair longer than) its natural rest length. Now imagine that you give this rubber band a sudden pulling jerk. This rubber band will stretch. Its length will increase.
Each pole of a multipole motor supplies the same kind of sudden pulling jerk. This jerk, if rigidly coupled directly to the platter, would produce a spike of temporary speed increase, resulting in the kick and coast speed irregularity problem.

But what happens if, instead of rigidly coupling this jerk directly to the platter, we couple this jerk to an elastic belt? The elastic belt will stretch, just as the rubber band did. Its length will increase.
Now, what happens when a belt's length increases? As we saw above, length changes in a belt will distort the faithful transmission of motion from motor pulley to platter rim, distorting the ratio between the two, thus distorting the speed of the turntable, and therefore distorting your music. In particular, if a belt becomes longer after leaving the motor pulley, this can produce a slower speed for the turntable.
Thus, in response to the sudden jerk from each pole of a multipole motor, an elastic belt distorts your turntable speed. But then how can a belt be helping anything? Who needs two distortions instead of one? The answer is that two distortions can be advantageous if they offset or cancel each other out (remember the pre-distorted groove of Dynagroove?).

How do these two distortions offset each other? Each sudden jerk from a multipole motor tends to increase the turntable speed, while each distorted stretch in belt length tends to decrease turntable speed. In other words, a belt's length-changing, speed-distorting response to each sudden motor jerk actually works to distort the turntable speed in the opposite direction from the motor jerk itself, thus helping to cancel out the speed increase variations that would be caused by each kick from a multipole motor.
If a second distortion is to offset and cancel out a first distortion successfully, then it must match that first distortion exactly, being as it were a mirror image of it. Since speed is a time based phenomenon evaluated instant by instant, and since our goal here is to maintain constant speed instant by instant, it follows that the two distortions should match and offset each other instant by instant. But that would be a very difficult, perhaps impossible design engineering job for the turntable designer.

The first distortion, the jerk from the motor pole, has a complex waveform shape that changes instant by instant. The waveform of this jerk depends on the physical contour of the motor's poles as they come together and then drift apart, on the magnetic strength of the poles, etc. The design engineering ideal might be to have the belt's elastic characteristics perfectly matched to the complex characteristics of a particular motor's jerks, such that, instant by instant, as the motor's jerk waveform increased speed, the elastic belt would stretch the precise amount of length needed to slow down the speed by precisely the same amount. Then the two offsetting distortions could track each other perfectly, instant by instant. To accomplish this, the belt's dynamic acceptance characteristics, showing how it responds in real time, instant by instant, to changes in jerk waveforms, would have to be evaluated, and would have to be perfectly matched to the jerk waveform of a particular motor.

However, this kind of belt evaluation and belt matching information is not generally available to the turntable designer. And we don't see turntable designers being fussy enough to demand such information. Indeed, most belt turntable designers haven't taken nearly enough care to even try matching their belt to their motor (or to their platter). The operative design principle has instead been: if the belt has the right diameter and a little stretch, then it's good enough to spin my platter round and round. That cavalier ignorance of (or avoidance of) the issue of exact speed control has given rise to a plethora of belt drive turntables which introduce as many new speed regulation distortions from the belt as they eliminated by filtering the multipole motor's kick and coast problems.

Moreover, even if the detailed dynamic characteristics, of the motor's changing kick jerk waveform and of the belt's response instant by instant, were to be studied, it's possible (indeed probable) that no physically real belt could be built that would truly offset each kick jerk's speed error, instant by instant. In part that's because a belt's length stretching (furnishing the offsetting correction which slows down speed) is triggered by and proportional to force changes from the motor's kick jerks (though the speed of its response to force changes is unknown); however, the motor's own speed error changes (furnishing the error which needs offsetting) are not proportional, instant by instant, to the force changes, but instead are proportional to the integral of the force changes from the motor's kick jerks (velocity is proportional to the integral over time of force)..

If a turntable design engineer can't get a belt's dynamic acceptance characteristics to exactly match and offset the instant by instant waveform of a multipole motor's jerks, are there at least some compromise design approaches by which he can at least get a belt to approximately match the overall motor jerks, averaged over time? Yes, but they're pretty lame. One commonly available spec for a belt, its static compliance, indicates how much it stretches after a fixed, unchanging force has been applied for some time (the belt's steady state position after all dynamic phenomena have died away). But this static compliance says nothing about the belt's dynamic response, instant by instant, which is what we care about since speed constancy instant by instant is our goal. Static compliance also says nothing about a belt's response to an input jerk waveform that is changing over time, instead of the fixed, unchanging force used to test static compliance. So static compliance is a pretty irrelevant spec to guide the turntable designer.

If static compliance is the only belt spec available, can the turntable designer utilize it, perhaps for a much cruder, lamer approach to designing a belt drive turntable? One tactic would be to calculate the average over time, within each kick jerk, of the extra force supplied by each kick jerk. The static belt compliance could then be selected to respond with the appropriate amount of stretch to this average extra force, the appropriate amount of length stretch being that amount which would slow down the turntable by just the right amount to offset the average speed increase from each kick. But there's a problem with this tactic. If the belt stretches just enough to accommodate the average extra force of each kick, then it might not stretch enough, quickly enough to accommodate the peak extra force of each kick. Thus, some of the peak extra kinetic energy of each kick jerk might not be converted to belt stretch (i.e. converted to potential energy), but instead will be transmitted by the belt to the platter. In other words, the belt won't completely succeed at filtering out the kick and coast problem of the multipole motor.

Another tactic would be to select a more compliant belt, which would stretch enough at the peak extra force of each kick to slow down the turntable the appropriate amount to offset the speed increase from the kick. But there's a converse problem with this tactic. The belt might stretch too much, and thus slow down the turntable too much, for those parts of the kick jerk waveform that are at less than peak force level.

A third tactic would be to look at the total kinetic energy put out by the multipole motor's jerk, summed over time (not evaluated instant by instant), and then try to find an elastic belt with exactly the right static compliance to convert all of this kinetic energy into longitudinal stretch (potential energy). If the belt is not compliant enough, then some of the extra kinetic energy in each kick jerk will not be converted to belt stretch, to potential energy, and thus some of the kinetic kick jerk and coast problem will be passed on by the belt to the platter. On the other hand, if the belt is too compliant to match the particular motor, then it will stretch too much, thereby slowing down the turntable too much.

It's doubtful that most turntable designers even go to the trouble of trying the above three design tactics, for matching the belt to the motor. It's doubtful that most even bother to try matching the belt's acceptance characteristics in any way to the motor's kick jerk output characteristics.

In any case, none of these three tactics is truly satisfactory. All give up hope of matching and offsetting the two distortions instant by instant. Thus they give up all hope of attaining truly constant turntable speed instant by instant. Instead, they settle for achieving more nearly constant turntable speed on the average, over time.

Unfortunately, that's not the way you listen to music. You listen to music instant by instant, as it flows by, not as some average over time. You also hear distortion instant by instant, if the turntable speed varies instant by instant. What do we mean by "instant"? Assuming a 20 kHz bandwidth, even a flicker of speed irregularity that lasts merely 1/10,000 second could distort your music. Some turntable might have perfect overall average speed over time, but that doesn't help. Its speed has to stay constant each and every 1/10,000 second.

As a reductio ad absurdum, imagine the worst turntable in the world, with horrible wow and flutter, and thus horrible distortion of your music. It might still have perfectly accurate speed taken on the average over time (for example, all the peaks of horrible flutter might match all the valleys, so that a 30 minute record side is over in exactly 30 minutes). Yet it would sound awful. Obviously, any claim or engineering design goal of achieving speed constancy on the average over time is meaningless. Obviously, we have to insist on the higher standard of achieving speed constancy instant by instant.
In sum, it's virtually impossible to design a belt drive so that the input acceptance characteristics of the belt truly offset or filter the kick jerks output by the motor, or truly achieve the required goal of speed constancy, instant by instant. As noted, most belt drive turntable designers don't even try.
But there's another tool at the disposal of the belt turntable designer. Unfortunately, this tool also brings some problematic baggage with it.

We've concentrated on the input acceptance characteristics of the elastic belt, as it stretches its length in response to each kick jerk from the multipole motor. But the elastic belt also has some output release characteristics.

After you stretch a rubber band with your hands, it then wants to shrink, and it exerts a force on your hands, trying to get the rubber band back to its normal length. Likewise, an elastic turntable belt stretches in response to each kick jerk from a multipole motor, and then it wants to shrink back to its normal length. In so doing it exerts a force on the other objects in the system, the motor pulley and the platter rim.

In other words, after the elastic turntable belt accepts the input of being stretched, it then provides an output of trying to shrink back to normal. This process could be viewed as an approximate conservation of energy, a sacred principle of physics. During the stretching phase, the extra kinetic energy of each kick jerk from the motor is converted to potential energy in the form of belt stretch (some energy being lost to heat). Then, when the belt tries to shrink back to normal length, the extra potential energy stored in its temporary stretch is reconverted back to kinetic energy, thus affecting the motion of the platter and the motor pulley.

To see how this works, it's useful to consider the continuously moving belt as comprising four sections. First, there's the section of belt that wraps perhaps 3/4 around the circumference of the platter rim. Second, there's the section of belt that wraps perhaps 1/2 around the circumference of the motor pulley. Third, there's the section of belt in free air that's travelling from the motor pulley to the platter rim. And fourth, there's the section of belt in free air that's travelling from the platter rim back to the motor pulley. Let's concentrate our attention on what's happening just in this fourth section of continuously moving belt.

Each pole of the multipole motor supplies a kick jerk which yanks this fourth section of belt, thereby supplying inputs which the belt accepts by stretching. The poles of a rotating multipole motor occur frequently enough so that there are at least several such yanking inputs on this fourth section of moving belt, before that section fully moves onto and beyond the motor pulley. Thus, any given fourth section of belt we might look at has already been yanked and stretched several times.

As discussed above, when the belt is stretched, the speed of the turntable platter is slowed, compared to what it would have been if the peak speed of the motor pole's kick were to have been faithfully transmitted by a non-elastic belt.

Now, what happens when the fourth section of belt releases the potential energy of its stretched length and shrinks back to normal length? If you stretch a rubber band between your two hands, it pulls on the constraints at both ends, namely your two hands, as it tries to shrink back to normal length. Likewise, the fourth section of turntable belt pulls on the two constraints at both ends, namely the motor pulley and the turntable platter. Since both these constraints are rotatable, the shrinking fourth section of belt tries to rotate them. The shrinking fourth section of belt actually tries to rotate the motor pulley backwards. But since the motor pulley is being actively driven forward by the motor (of hopefully adequate power), this might not be a serious concern during the active pole drive portions of the motor's rotation (but, between poles, while the motor is coasting, note that the valleys of the kick and coast problem are actually made deeper and worse by the belt shrinking back to its normal length after it is stretched).

On the other hand, the platter is not being actively driven by anything other than the belt, and the platter speed is what matters for the time axis of your music, so what the fourth section of belt does to the platter speed when it shrinks back to normal length is of crucial concern.

When the fourth section of belt shrinks, it causes the platter to speed up. It's effectively yanking at the rotating platter, saying come on, let's go faster (much as the multipole motor's kicking jerks had yanked at this same fourth section of belt). The platter can't stretch in response to these yanks, but it can rotate faster, so that's what it does. And, if these yanks from belt shrinking are in any way irregular over time, then the platter's speed rotation in response to these giddyap yanks from the shrinking belt will be irregular. But speed irregularity is what we want to avoid, and we came to belt drive turntables with the hope of getting regular and accurate speed. Can we get it from belt drives or not?

As we discussed above, it is virtually if not actually impossible to get a belt's stretching behavior, in accepting the kick jerk input from a multipole motor, to perfectly offset the kick and coast problem. Now we see that the belt stretching, when it accepts the input from a kick jerk, and acts to slow down the platter, could perhaps be offset against the opposite tendency, namely that the belt acts to speed up the platter when it releases the potential energy of the stretch and shrinks back to normal length. Might these two opposing tendencies perhaps be precisely balanced against one another, to produce exact and constant speed control from a belt drive turntable?

Indeed, we could envision that the fourth section of belt, having been fed several stretching kick jerks within its length, might even contain within itself the mechanism for potentially averaging out the kick and coast problem, introduced at the motor pulley end, before it even reaches the platter rim at the other end. If the belt material is constructed so that the stretches from each kick are quickly converted from potential to kinetic energy, then the irregularities of stretching and shrinkage, which affect the platter speed, might average themselves out over the length of this fourth section of belt, and thus effectively never reach the platter rim, thus never causing speed irregularities at the platter. Can we count on the fourth section of belt to simply average out, within its length, the kick and coast problem of every multipole motor?

Nearly every belt drive turntable designer has thought that the answers to the above questions is yes, yes, yes, and yes. But, if you want truly accurate and constant speed, if you want truly clean, distortion free music, the answer is no, no, no, and no.

Most belt drive turntable designers have blithely assumed that you can slap together any old reasonably compliant (stretchy) belt, any old reasonably heavy platter, any reasonably smooth bearing, any old motor, and presto, you have a belt drive turntable with enough speed accuracy to bring to market. There's other design work to do, such as vibration isolation and packaging aesthetics, but they assume that the speed accuracy and constancy problems have been solved by the simple use of a belt. These designers have been ignorant of the weaknesses, compromises, and limitations of belt drive, and they have also been ignorant of the engineering required to at least get the best speed accuracy out of a belt drive design. There are more belt drive turntable designs than there are engineers who know how to design a belt drive turntable for the best possible speed accuracy. That explains in part why belt drive turntables sound so different from one another, in a diverse range of sonic aspects including pitch (speed being slow or fast), pitch wandering, rhythm and pace, solidity of bass impact, etc.

Turntable designers who blithely assume that the answers to the above questions are all yes are in for a rude shock when their belt drive turntable doesn't work as well as they hoped. They face a number of shortcomings in speed accuracy performance which they can't explain, and perhaps don't even know how to adjust for.

First, there are issues of absolute speed accuracy. For example, if a designer calculates the theoretically correct ratio for his motor pulley to his turntable platter rim, he'll probably find that his belt drive turntable actually runs at the wrong speed (usually too slow, giving a poor sense of pace and rhythm). One problem is that, when he counts on the belt to average the peaks and valleys of the multipole motor's kick and coast phenomenon, he doesn't have a profile map of the motor's speed and energy output over time (this depends on the actual geometry of the poles in the particular motor he selected), so he doesn't really know what the average input to the belt is. A second, related problem is that there is some energy loss in the belt itself (to heat), as it first stretches in response to the motor's kick jerk input yanks and then converts most (but not all) of this potential energy to kinetic energy as it shrinks in length and in turn yanks on the platter rim. This energy loss will make the turntable speed slower than predicted.

Designers of belt drive turntables usually respond by machining their pulleys to an ad hoc diameter, rather than a scientifically predictable diameter, in order to get the correct platter speed with a given belt. Such ad hoc designing is already a warning flag, alerting us that the belt drive design approach is fundamentally compromised in terms of providing truly accurate speed.

Then there's a third, related problem. What if the belt manufacturer slightly changes the exact way he makes the belt, without telling the turntable manufacturer? The speed of the turntable will change, because the amount of this energy loss to heat depends on the precise material composition of the elastic belt, and how its long chain molecules intertwine. A fourth, related problem is that this energy loss, and so the turntable speed, will be affected by the aging of the belt and by changes in the ambient temperature of your room.

Second, there a re issues of speed irregularity. The belt releases the pent up potential energy of its stretch as output, converting it to the kinetic energy of shrinking back to normal length. But it releases this output in a completely different time frame than it accepts the input from the motor's kick jerks. In other words, the energy output of the belt (which tends to speed up the platter) has no temporal relation to the energy input to the belt (which tends to slow down the platter). If these two effects are temporally unrelated, then they can't truly offset each other in real time. It's as if one boxer with his hands freely swnging rapidly through air were opposed by a second boxer whose hands had to slowly work their way through molasses; the second boxer couldn't keep up with the first in real time, and thus he couldn't truly offset the efforts of the first boxer, even if he were just as strong.

In this case, the belt shrinkage back to normal length (which tends to speed up the turntable) operates in a much slower time frame than the belt stretching (which tends to slow down the turntable). Why is that? On the input side, the motor kick jerks are usually relatively powerful, and the belt is usually quite compliant (easy to stretch), so the belt responds quickly, stretching quickly in response to m motor kick jerks. It's like the quick boxer. On the other hand, when that fourth section of belt tries to shrink back to its normal length, it has an enormous load on its back that it must tug along, this load being a relatively heavy turntable platter (with a high moment of inertia). This fourth section of belt is trying to do the 100 yard dash in shrinking back to its normal length, but it must tug along a giant river barge, and that slows down its response dramatically. It reacts like the boxer with his hands stuck in molasses.

The belt's shrinking response is so much slower than the belt's original stretching that there's no hope of the two truly balancing or offsetting each other in real time. Thus, the belt drive turntable designer cannot rely upon any such offsetting to obtain constant speed. The only turntable to even try this design approach was the Weathers turntable of the early 50s, in which Paul deliberately mated a weak motor (thus slowing down the response of the belt stretching) with a very light platter (thus speeding up the response of the offsetting belt shrinking). But it had other problems, including poor immunity to automodulation distortion from loud music passages (see discussion above).

We saw earlier that it was probably impossible to engineer a true offsetting match at the input to the belt, where the belt accepts the kick jerks from a multipole motor, and we now see that it is also probably also impossible to engineer a true, real time offsetting match at the output from the belt, where the belt releases its stored up potential energy to the platter.

Grim news indeed, for belt drive designers who hope to achieve regular speed constancy. So what do they do? What can they do?

They throw up their hands, throw a Hail Mary pass off into space, and pray for the best landing. They fall back upon a saving grace of the gross temporal mismatch we discussed above, between the belt stretching and the belt then shrinking back to normal.

How does this saving grace work? Recall that the quick response occurred at the input to the belt, as it was being stretched by the kick jerks of the motor poles. And the slow response occurred at the output from the belt to the platter, as the belt tried to shrink back to normal length, tugging the heavy platter along behind it. It so happens that, with the slow response being at the output and the quick response being at the input, the output simply won't see the quick things happening at the input. It's as if you had a power amplifier with a quick input stage and a very slow output stage; the output stage simply wouldn't respond fast enough to let you see, at the output of the amplifier, the quick things happening at the input. In other words, the amplifier as a whole would respond as slowly as its output stage (or indeed as slowly as the slowest, weakest link in its internal chain).

This means that the belt drive turntable designer can cheat by using this saving grace. He can get away without understanding any of the other intricacies and tradeoffs of turntable design discussed above. A Hail Mary pass can win a game even if a team doesn't understand how to design football plays and is inept at actually executing them; likewise anyone can "design" a belt drive turntable by using this saving grace, even if he knows little about turntable design.

When this saving grace is relied upon to "design" a belt drive turntable, the kick jerks from the multipole motor are not truly offset in the belt itself, or in the fourth section of belt that we have focussed our recent attention on. These undesirable kick jerks actually reach the platter. But, thanks to the saving grace fact that the belt's output is so slow in tugging at the heavy platter, these quick kick jerks don't affect the platter speed much.

In the power amplifier analogy, the slow output stage could also be likened to a filter introduced in the amplifier circuit, to filter out most higher frequencies (i.e. most faster changes in the signal). Likewise, the slow response of the belt, trying to shrink back to normal length while being forced to tug at a heavy platter, can be likened to a filter introduced in the turntable drive circuit. This filter would filter out most of the rapid speed changes from the multipole motor, such as the dreaded kick and coast problem from each pole. This filter effectively integrates or averages the rapidly changing speed input from the motor, to produce a more constant average speed output. So this averaging process can indeed be effective in reducing the kick and coast problem coming from a multipole motor.

But notice that neither the belt nor the motor actually control the accurate speed of the turntable any longer, especially not in real time. Instead, the motor and the belt merely make small repeated contributions, which are deliberately slowed down so as to minimally affect platter speed in real time. The designer settles for an averaging process as a creator of speed constancy, and settles for an average as a measure of speed constancy.

To make this averaging process effective, the designer needs a powerful averaging tool. The motor and the belt have been discarded as useless in this quest, so that leaves the platter as the only other physical element remaining that could be used as an averaging tool. This explains why turntables platters are made heavy (actually, heavy at their rim, so they have a high moment of inertia). The higher the platter's moment of inertia, the more powerful it is as an averaging tool (and it also has more value in being immune to automodulation distortion from stylus drag during loud music passages, as discussed above).

To say the same thing another way, the platter as an averaging tool works like a slow, low pass, integrating filter, which lets very low frequency variations (including speed variations) through, but which works to filter out higher frequency, more rapidly changing variations (including speed variations, such as the kick and coast variations from multipole motors). Like every other filter, this platter filter has a certain fixed slope, so it doesn't totally eliminate unwanted higher frequency, more rapidly changing speed variations such as the kick and coast variations. Instead, it merely reduces these unwanted higher frequency speed variations by a certain amount, in proportion to how low in frequency the filter starts operating. The lower in frequency such a filter starts operating, the more effective it can be at reducing the unwanted higher frequency speed variations by an even greater amount. To be maximally effective at reducing the higher frequency unwanted speed variations from the multipole motor's kick and coast problem, the filter should start operating at the lowest possible frequency. This means that the platter's moment of inertia should be as high as possible, consistent with product cost and some other technical factors such as load on the main thrust bearing.

Indeed, the sum total of design work going into may belt drive turntables, with respect to speed accuracy and constancy, has been simply to combine a belt of some indefinite and unknown compliance with as heavy a platter as the product budget can afford. This is the Hail Mary pass that most designers throw into the air, both praying and knowing that they'll get halfway decent speed performance, thanks to the saving grace of any old heavy platter acting as an averaging filter with any old compliant belt. But is halfway decent good enough for you?

Furthermore, there are also some problems and penalties with this Hail Mary design approach. Firstly, by making the platter heavier in order to improve its filtering performance, the designer is actually worsening the disparity we discussed above, wherein the energy output from the belt is much slower than the energy input to the belt, so one cannot offset the other in real time. Making the platter heavier gives the belt a heavier load (higher moment of inertia) to tug on when it tries to shrink back to normal length and thereby feed kinetic energy into the platter speed. Thus, making the platter heavier actually works to worsen the belt's ability to correct the platter speed expeditiously (and there a number of reasons why it might need correcting, some of which are discussed below, others of which relate to correcting for stylus drag slowdowns as discussed above). Making the platter heavier further gives up control of the platter speed by the other two elements of the system, the motor and belt.

Making the platter heavier also further reduces the belt's ability to play any significant role in being an internally self correcting mechanism for speed constancy (especially within that fourth section). There might have been some hope, with a belt carefully engineered and matched to both motor and platter, that the belt stretching from the motor's kick jerks (which tend to reduce platter speed) might have been at least crudely or partially offset by the belt shrinking back to normal length (which tend to increase platter speed), thus providing at least some measure of speed constancy benefit or improvement within the belt itself and by the belt itself. But, with the platter being made much heavier, the heavy platter dominates so there is no matching among the three drive elements, and the temporal disparity between belt output and input is worsened to the point where the belt itself cannot contribute significantly to speed constancy.

Designers typically throw away all hope of matching the belt to both the motor and the platter (or they don't even know enough to know what such matching might mean). Instead, they fall back upon the simplistic Hail Mary game plan, relying on the saving grace filtering feature of any old compliant belt with a very heavy platter. In so doing, they effectively surrender control of speed constancy to the platter.

But what's so bad about that? Why not surrender speed constancy to the platter? It turns out that there's a very nasty penalty for doing this. The compliant belt and the heavy platter form a reactive tank (LC) circuit, in which energy is exchanged back and forth between the belt and the platter. Another name for such back and forth exchanges of energy is oscillation. These energy exchanges occur in their own time frame, thereby producing speed oscillations that occur in their own time frame.

Speed oscillations are of course speed irregularities, just the opposite of the speed constancy we are seeking here. These speed oscillations typically occur at a low frequency, sounding like wow or pitch uncertainty (unlike higher frequency speed irregularities, which produce flutter and the grundgy or fuzzy sound of FM distortion). Making the platter heavier places these oscillations at a lower frequency, perhaps even to the point where a belt drive turntable would not have audible wow but could still make you seasick. Making the platter heavier also can make these speed oscillations larger (worse), since the tank circuit is now even more reactive.

And furthermore, in a cruel twist of irony, making the platter very heavy, which was the simple saving grace, Hail Mary filter tactic designers used above to reduce the multipole motor's kick and coast problem, now turns around to bite them in the butt. By making the platter very heavy, they have reduced the motor's influence, and have made the platter's own tendencies dominate the situation. Now they suddenly discover that the platter, rather than being the prefect passive partner that tends to run at a constant speed, is instead engaging in active hanky panky of its own, oscillating with the belt in its own dance to its own rhythm. But now it's too late. For they have weakened the motor's influence to the point where it cannot effectively correct the oscillations of the heavy, domineering platter.
That's why belt drive turntables tend to exhibit their speed irregularity problems as wow rather than flutter. They are still severely compromised in speed constancy, but simply at a lower frequency of variation, of modulation distortion, of sidebands, than rigidly coupled drive turntables. Belts don't solve the speed irregularity problems we saw in rigidly couple turntables; they merely shift the problems to a new domain.

Belt drive turntables differ from one other in belt compliance and platter moment of inertia, so their oscillation wow patterns differ. That explains in part why belt drive turntables sound so different from one another in pace and rhythm, in steadiness of pitch, and in solidity of bass (bass notes last longer, so they require a longer sustain of turntable pitch accuracy to sound straight and massively impactive, rather than warbly, wobbly, and weak).

Can the belt drive designer reduce or damp the unwanted oscillations between platter and belt? In principle yes, but in practice it's tricky to execute. In principle, what's required is simply the addition of some resistive damping. This would damp the reactive LC tank circuit of belt and platter so it would no longer oscillate.

A common form of resistive damping is friction. Thus, if a knowledgeable turntable designer wants better speed constancy, he might well consider intentionally adding some friction to his rotating platter. It's worth noting that some Swiss and German engineers are so justifiably proud of their ability to produce nearly frictionless bearings that they cannot bring themselves to make turntables with high friction. As a result, the Thorens turntables exhibit some of the most spectacularly low friction bearings on the planet, and will spin seemingly forever (with the belt removed); but, at the same time they also exhibit some of the worst audible wow, in part because there is, as a matter of engineering pride, almost no resistive damping for the oscillating reactive tank circuit.

What might be useful ways to introduce friction? The fit and/or finish between platter spindle and well could be made poor, instead of smooth and polished. But this would be causing friction via crude irregularities, like two meshing mountain ranges, rubbing each other. The crude irregularities of these two mountain ranges rubbing together would cause unwanted speed irregularities (snags and letting gos), as well as unwanted vibrational rumble (the earthquake rumble of each letting go after each snag). So that tactic is out.

One useable tactic is to introduce a viscous fluid in the bearing, which provides friction in a liquid hence smooth form. This can be especially effective if the spindle is made in a larger than normal diameter, so that the viscous fluid has a larger moment arm (more leverage) with which to work its resistive magic (as in the Linn Sondek). The use of viscous fluid for resistively damping platter rotation can also be enhanced by various helical screw kinds of arrangements, which force the fluid to do extra work in opposing the rotation of the platter (as in the turntables from Max Townshend). It's no accident that these two brands have the best reputation among belt drive turntables for pace and rhythm, solid bass, and master-tape-like clarity. It's because both these designs recognize that belt drive, far from being a simple Hail Mary solution, brings with it new problems that must be addressed, and that overcoming the problem of speed constancy requires at least the addition of a fourth element, resistive damping, to the three usual elements of a belt drive turntable.

Does this mean that the belt drive concept could work perfectly, if only designers would wake up and add some resistive damping? Alas and alack, no. There are still further problems with the belt drive concept, problems still related to our fundamental theme of speed accuracy and constancy.

The belt can isolate the platter from all vibrations of the motor, including simple rumble vibrations (perhaps once per revolution) as well as kick jerk vibrations from each pole (perhaps 24 per revolution). It seemingly makes sense, then, to mount the motor on some rubber bushings, so that its vibrations don't get transmitted directly to the platter via the spindle bearing. In other words, it seemingly makes no sense to mount the motor solidly to the same substructure as also holds the spindle bearing, since vibrational rumble would be transmitted directly via this path to the platter and thence to your record groove and cartridge. The belt isolates the motor vibrations from the platter via the drive path (unlike a direct drive or rigidly coupled drive), so why not also isolate the motor vibrations via the motor mounting path?

This idea is good in theory, as far as it goes. But it doesn't go far enough, because in practice this tactic introduces yet another penalty, yet another speed constancy problem into the belt drive can of worms. If the motor is mounted on a flexible, vibration absorbing mounting (such as rubber bushings), then its pulley can move with respect to the platter rim. This means that the belt length will change every time that the motor vibrates within its flexible mounting. As we saw above, every belt length change produces a speed irregularity for the platter (perhaps delayed in time, if the platter is very heavy). Thus, what might seem like a clever tactic for reducing rumble turns out to be a very bad thing for achieving speed regularity in a belt drive turntable.

A nasty feature of mounting the motor flexibly is that there are so many agents which might contribute to wobbling the motor in its flexible mounting, hence might contribute to speed irregularity. The music coming from your speakers could make the motor wobble, hence cause the belt length to change at musical frequencies, thus causing a new kind of automodulation distortion, tempered only by the heavy platter (but the average speed energy input to the platter by the belt would still be skewed over time in the long run as the music changes, thus causing wavering speed irregularities governed by the music's average content over time). Footfalls would have a similar effect; since they have a very low frequency content, the reactive belt-platter filter would be only partially successful in reducing their adverse influence upon platter speed constancy. Additionally, the vibrations of the motor itself, such as once per revolution eccentricity, which would normally contribute to rumble, now directly contribute to speed irregularity; their energy content is quite low in frequency, and thus also would be only partially filtered by many belt-platter combinations.

Perhaps the worst can of worms for the belt drive concept is introduced when not only the motor is flexibly mounted, but also the platter. The designers at Acoustic Research probably deserve the credit (or infamy) for inventing the concept of the floating subchassis, which suspends the heavy platter, spindle, and spindle bearing on a flexible mounting (usually via a separate subchassis). This design idea was then adopted by Linn and other makers of belt drive turntables. The primary intent of this design tactic is to isolate the platter (hence disc and cartridge) from external vibrations, such as footfalls and the music from your speakers. A further intent is to reduce rumble from the motor, since once again the motor and platter spindle are not directly coupled via any rigid mounting (except that this time it is the turn of the platter spindle rather than the motor to be flexibly mounted). This design tactic succeeds admirably at meeting its intended goals. However, it can have horrendously adverse impacts upon speed regularity in belt drive turntables.

The nature of the chief problem is essentially the same: the belt changes length, and this causes speed irregularities. But this time the problem is far worse in degree. Why? The suspended subchassis, including the heavy platter (made very heavy in the Hail Mary design move above), weighs much more than the motor. And, with a goal of reducing feedback from music and footfalls to the record groove, not just reducing once per revolution rumble, the subchassis suspension has to be effective at filtering down to lower frequencies than a motor suspension. Combine these two factors, and you typically wind up with a suspended subchassis that will move much farther than a suspended motor, and will move (indeed, will continue oscillating back and forth in its suspension) at much lower frequencies than a suspended motor. This in turn means that the speed irregularities will be much worse from a suspended subchassis (since the belt length changes are much greater), and that the irregularities will occur (and even continue oscillating) at a much lower frequency. Since the speed irregularities occur at a much lower frequency, the saving grace filter of the platter's rotating inertia is much less effective in reducing the speed irregularities.

Thus, the typical suspended subchassis in a belt drive turntable hits us with a double whammy of speed irregularity: larger amplitude irregularities, and lower frequency irregularities (which the platter inertia is less effective at reducing).

The speed irregularities caused by a suspended subchassis can be much greater than those caused by the multipole motor's kick and coast problem. In both cases there are forces stretching the belt and then allowing it to shrink back, thus causing speed irregularities. But a suspended subchassis is so much heavier than the small internal rotor of a motor, and can move so much farther, that it is capable of imposing far higher energy to stretch a belt than the kick and coast jerks of a motor, thus causing far greater speed irregularities (though usually at a lower frequency).

Thus, the ultimate irony of the belt drive concept, as usually implemented, is that the cure may be worse than the disease. The first turntable design concept discussed above, where the motor is rigidly coupled to the platter, inherently requires that there be no distance play between platter and motor, so the speed irregularity problem we're discussing here does not occur. Of course, the directly coupled turntable design concept has the problem that the multipole motor's kick jerks are transmitted directly to the platter. The belt drive concept solves this problem, but introduces new problems. This latest problem relates to the suspended subchassis often included with a belt drive turntable, and it can cause the worst speed irregularities of all.

To make matters worse, there's usually a third problem (hinted at above), to constitute a triple whammy of speed irregularity for belt drive turntables with suspended subchassis. The mass of the suspended subchassis and the springs of its suspension form a reactive tank circuit, which tends to keep oscillating after any transient vibrational disturbance, often long after the transient vibrational disturbance has disappeared. Thus, such a suspended subchassis would continue to produce speed irregularities on its own, long after the  external disturbance causing the initial movement of the subchassis (and hence the original fleeting moment of speed irregularity) had ceased. Speed irregularity that drones on and on is naturally much more noticeable and objectionable than a fleeting moment of speed irregularity, so this oscillating subchassis problem would naturally make a belt drive turntable sound much worse.

This oscillating subchassis problem itself has the worst influence on speed constancy in those turntables where the oscillations contain or allow horizontal movement (also called yaw) of the subchassis. That's because side-to-side movement of the subchassis obviously affects belt length (hence speed constancy) the most, while vertical movement of the subchassis affects belt length the least. Also, the vertical movement should be perfectly straight up and down, in order to have the least effect upon belt length. Unless the suspension is perfectly tuned to the subchassis (which is difficult to do, since the subchassis cum platter always has an asymmetrical shape and mass distribution, both static and dynamic), the vertical movement will include a rocking component (like a rocking horse), and then there will be more horizontal movement and worse speed irregularities. This explains why certain turntables (e.g. the Linn Sondek) are so sonically sensitive to the slightest changes in suspension tuning, including even lead dress of the tonearm wiring, in order to get the most perfectly vertical movement (anything short of this goal produces worse speed irregularities and also worse frequency modulation distortion of your music).

This also explains why some carefully engineered subchassis suspension turntables (e.g. Oracle) can sound better than others, because their suspensions quickly convert external jarring (even horizontal) into purely vertical motion; thus they exhibit merely a temporary blip of speed irregularity (corresponding to the original external disturbance), and any ongoing oscillation produces merely minimal ongoing speed irregularity. Indeed, in such turntables the temporary blip of external disturbance might even result in minimal speed irregularity (even temporary), thanks to the saving grace filtering of the heavy platter working with the compliant belt.

On the other hand, some belt drive turntables (e.g. the SOTA) allow considerable horizontal movement of their suspended subchassis, which also continues to oscillate long after the initial disturbance. These belt turntables have some of the worst audible speed regularity, out in the real world where music and footfalls disturb a suspended subchassis.

It's worth noting that the two reactive tank circuits discussed above can compound each other, creating huge peaks and valleys of speed irregularities. A belt drive turntable with a very heavy platter (e.g. the SOTA again) creates an energy tank circuit which is very reactive, very poorly damped, and capable of cycling and recycling (oscillating) very large amounts of energy, all at a very low frequency. If the mass of this suspended subchassis is very large (which it necessarily will be, since the platter constituent thereof is so heavy), and if significant ongoing, oscillating horizontal movement of that suspended subchassis is allowed (as it is in the SOTA), then this energy tank circuit will likewise cycle and recycle very large amounts of energy at a very low frequency. Both tank circuits recycle large amounts of reactive energy that produces speed irregularities, both continue their ongoing oscillations for a long time, and both oscillate very slowly. Thus, they will add and subtract to giant speed errors as they slowly wander into phase and out of phase with each other.

To make matters yet worse, there's also usually a fourth problem of speed irregularity with belt drive turntables having a suspended subchassis. Again, the problem is worst with those subchassis suspensions that allow horizontal movement, especially ongoing oscillations of horizontal movement, in response to external disturbances. What is the nature of this fourth problem? It's a kind of Doppler speed irregularity. If the subchassis moves horizontally, then a component vector of that motion is almost surely to be in the direction of linear groove travel under the stylus. You'll recall that this direction constitutes the time axis of your music signal. So a motion of the subchassis along the time axis would naturally add to or subtract from the speed of the turntable at that moment. If the subchassis not only moves horizontally but also oscillates horizontally, then there would be a cyclical addition to and subtraction from the nominally correct speed, i.e. a cyclical speed irregularity.

What about the platter's large moment of inertia? Wouldn't this help to keep the platter speed constant in spite of the subchassis' cyclical horizontal movements? No. You see, the platter's inertia is based upon the frame of reference of the surrounding universe, not of the subchassis. The heavy platter tends to keep rotating at the same speed relative to the universe, not relative to the subchassis with its oscillating horizontal motions. And so that tangential vector component of the platter's inertia that lies along the time axis tends to keep the linear groove velocity constant with respect to the outside universe, not with respect to the subchassis with its oscillating horizontal motions.

The platter's well intentioned inertia, which would tend to keep the platter rotating a constant speed, is keyed to a different reference frame than the platter itself, and the subchasiss cum platter keeps cyclically moving with respect to that reference frame. The platter, subchassis, and pickup arm keep cyclically moving within the reference frame of the universe, so they keep cyclically modulating the well intentioned effect of the platter's inertia. Thus, the oscillating horizontal motions of the subchassis are superimposed upon the platter's inertial velocity with respect to the universe, and tend to cause speed irregularities. And therefore the platter's inertia is of less help than it should be, in counteracting the various speed irregularities associated with belt stretch and shrinkage.

Indeed, these two unwanted effects can also compound each other. With a suspended subchassis turntable, the belt is forced to stretch and shrink if the subchassis allows horizontal movement, directly causing speed irregularities. These speed irregularities are ameliorated somewhat by the platter's inertia, except that the benefits of these amelioration attempts are compromised by the platter being horizontally moved back and forth within the reference frame for its inertia. Since both bad effects are due to the suspended subchassis moving (perhaps even oscillating) back and forth, they both occur at the same rate, and so they compound each other.

It's almost like Doppler speed irregularity. In a classic Doppler shift, the frequency of a cyclical phenomenon (say a train whistle sound) is superimposed upon a fixed velocity (like the speed of the train), and the cyclical phenomenon (whistle pitch) is shifted as the train changes position (goes by). Likewise, in this turntable example, the horizontal cyclical motion of the subchassis is superimposed upon the relatively fixed inertial velocity of the platter with respect to the universe, and shifts the effective speed up and down by shifting the subchassis' position, so that its speed (like the pitch of the train whistle) tends to go up or down each time the subchassis goes one way or the other in its cyclical horizontal motion along the vector of the linear grove direction that is the time axis. In effect, the oscillating swaying of the subchassis modulates the time axis of your music.

It's worth noting that there's a cute technical, Einsteinian difference between the classic train example and this turntable example. In the train example, we think of the observer as being fixed and the train as going by. On the other hand, in the case of a turntable, the heavy platter tends to keep a constant fixed velocity with respect to the universe, while it is the observer, in the form of the cartridge mounted to the tonearm mounted to the subchassis, who is oscillating back and forth. From a relativistic standpoint, it is irrelevant which, of the observer or the observed object, is fixed, and which is moving. In both cases, the observer will perceive the speed, pitch, or frequency of both the train whistle and the music as changing rather than staying constant.