Tables That Feature Bearing Friction

Hi everyone,
I know I will regret posting in response to some of the above, but before I comment on some of what Dertonarm has postulated above repeatedly, allow me to address the original question:

For any given motor/powersupply combination there is a specific load against which it will work producing the least vibration/smoothest rotation. This is, strictly speaking, only true for a motor which has perfect spindle bearings. They are another source of irregular behaviour, particularly since the load on a belt drive motor pulley when used with a thread is rather high(compared to the tension required if a rubber, neoprene, silicone... belt is used). It is possible to address this problem, at the price of complexity and noise(i.e. counter-pulley for the motor spindle...other options exist).

Off course stylus drag isn't capable of slowing down a platter AND at the same time modulating the deceleration(or the acceleration!) at the exact frequencies that are being played back at that moment. The platter acts as a low pass. Nevertheless, stylus drag is NOT constant and therefore requires lost energy to be fed back into the system at an ever changing rate to maintain absolute speed, requiring constant acceleration/deceleration. If the platter is loosing energy due to additional "drag", be it generated by an eddy current brake(no noise added), a felt brake, a paddle running through a silicone bath or a particularly lossy bearing(i.e. the DPS turntable), the influence/impact of the losses generated by the stylus drag become smaller.
Increase the drag tremendeously and you will get the influence of stylus drag below the threshold of audibility(let's just say for a moment that there is such a thing...)*.
You need to increase motor torque(meaning, all other things being equal, increase motor noise) once the losses become too high, both to be able to maintain target speed and to bring the platter up to speed within a reasonable amount of time.
*The same can be achieved by increasing platter mass/inertia, BUT the remaining speed variations will be lower in frequency and higher in amplitude if no means of damping(same as aditional drag, sorry)is employed. Low pitch variations are more obvious/bothersome to some than they are to others.

Since amplitude and frequency of the platter speed variations can't be reduced to Zero, any turntable will be compromised, unless it's platter would feature infinite inertia, in which case all other factors(motor, idler, belt etc.) become obsolete, since nothing can move an infinitely inert structure.

The insistence on the superiority of Dertonarm's super heavy platter driven by a lossy drive/"slipping" thread is hard to understand, particularly if we apply the laws of physics, as continously demand by the master(couldn't resist that one ;-) himself.

If the thread is allowed to slip while it is driving the circumference of the platter, slippage will (likely)occur at the pulley too. Even if the slippage was constant, and it is NOT(according to physics, keyword:"stiction") , slippage on both the driven and the driving surface will result in chaotic behaviour, COMPROMISING the evenness of energy transfer.

If, as some turntable manufacturers have done, the platter speed is monitored and used to control the motor, hunting and pecking is inescapable and only the chosen feedback time constant, the platter mass, the motor torque and the little bit of stretch left even in Aramid or Dyneema threads will govern the speed variation amplitude and frequency around the nominal target speed. And if there is belt/thread slippage or an out of round platter, the "error correction" will have to work very hard.

Dertonarm stated that neither direct drive nor idler drive was usable to build a turntable with high mass/inertia platter. Oh boy... we are mostly listening to shitty records cut on Neumann lathes, some of which use a Technics SP10 MkIII motor(albeit driving a 40kg, large diameter platter)
The Onkyo PX-100M, an eddy current direct drive tt features a 24lbs platter(without the mat) and is an excellent deck.
Other examples in conflict to the above statement were mentioned by Chris Brady.

What does a super heavy platter buy the designer other than inherent higher frequency speed stability? Problems!
You'd be surprised how much wobble/tumble can be detected on platters even if the bearing tolerances are super tight(the Continuum Audio site once had some indepth analysis graphs and animations)
Dertonarm will likely reply that an airbearing is the solution, but it isn't(it's a compromise too). It minimises bearing noise and friction(a major problem with super heavy platters) but can't restrain platter tumble as well as a pressurised oil or grease bearing. For what it's worth, essentially all industrial axial airbearings for heavy loads are made for much higher rpms.

Dertonarm suggests/requests the side load on the platter to be zero. Very good(unless a spindle bearing à la Bill Firebaugh is used)! But if you do this by putting a counterpulley opposite of your driving pulley and maintain not only the distance but also the diameter of the counterpulley, there is a strong risk of all 4 "free" thread sections acting as strings, resonating at the same frequency. Any such behaviour will wreak havoc on the smooth operation of the motor as it changes the load it sees rapidly. Put a break on the counterpulley or change the pulley diameter and the distance accordingly and the problem is solved.
Off course, once the platter is made heavy enough, that becomes neglectable too... depending upon your neglectability threshold.

My conclusion: Dertonarm's way is one way, but not the only way. It is, like all ways, compromised(I wouldn't dare calling it flawed).

A few last words on how "commercial" designers/engineers are often described here and in other forums.
Yes, there are some(too many) that do not have a deep understanding of physics or electronics(not to mention the growing number of copycats). But just as many do and they are into it because they happen to share other people's(be it customers or colleagues) enthusiasm for music and all the gadgets that allow us to enjoy it in it's preserved form. It is simply not true, that all of us think with a target price tag in front of our eyes first. It is simply not true, that all of us think: "yeah, I guess that'll do..." And that is a FACT.

Finally, it is downright ARROGANT to belittle designers who are capable of coming up with a component that delivers 95% of what is currently possible at 10% of the price. Yes, it is all about compromises and ESPECIALLY an all out assault on the state of the art will eventually run into facing this as well.

Happy Easter,
Frank Schröder

Dear T bone,

>Would not the drive (pull) side have a different tension than the 'lag' side - thereby making same resonance on 4 strings impossible?<

You are correct if the thread has stretch and is under tension. Dertonarm suggests the use of a thread with no stretch and advocates low enough a tension to allow for slippage(if I misunderstood, I apologize). Both longitudinal and transversal waves contribute to the vibration of a connecting medium(be it a rubber belt or a thread), modulating the load "seen" by the motor. The ratio between the two differs though. Since Dertonarm was approaching the discussion from a theoretical point of view and has emphazised individual aspects, I thought it would be fair to point to conceptual flaws as he has asked everyone to do.
But he has built a turntable following his dogma, appearently without running into a problem in this area as he said that achieving and maintaining correct(or what he considers correct) thread tension was not difficult. Complextity and resulting chaotic behaviour of elements working together can come to rescue here, rendering a theoretical problem nonexistant in praxis. Dertonarm was asking for input on how to improve upon existing solutions and my suggestion was merely aimed at that.
As a designer my aim has always been to avoid a source for problems rather than quantitatively minimizing the problem. I've built turntables with thread drive and counterpulley in the early eighties and later commercial implementations of the mirror image positioning of the counterpulley all fixed some major issues - side thrust on the platter bearing, uneven operation of the subchassis(Audiomeca Roma) - and introduced others - noise, belt "flutter", slippage(where none was intended to occur).

BTW, a platter driven by a constant stream of air will be harder to build and not necessarily better than one with a pulsed supply. If the pulsed supply works against, lets call it "teeth" or "wings" of appropriate shape, a very even drive force in agreement with Dertonarm's dogma can be achieved.
If the system losses are small enough and the inertia very high, such a pulse can occur only once per revolution or even less frequently.
One could see an analogy between tts and high precision timekeeping devices. The astronomical regulator with a "free" escapement and heavy(high inertia pedulum) driven with minimal force(keeping the drive system's influence small) stands for Dertonarm's thinking, the low mass high frequency oscillation of a quarz or even Cäsium atomic clock for the quarz-"locked"(it's not locked) instantaneous(it's not instantaneous)correction direct drive with light platter.

The atomic clock produces ultimately less deviation from perfect accuracy. Now why then do the best DD-PLL turntables with light platters(i.e. EMT 948) sound inferior to the best DDs with heavy platter or the best belt, tape, thread or rim driven turntables? Lack of proper execution or fundamentally flawed? This seems like a rethorical question, but it isn't.

If your motor has infinite inertia(god brought it up to speed initially...) and your drive is lossless, then your platter can be infinitely light, will be infinitely inert and therefore rigid as well. No more energy storage or mechanical impedance matching problems(topics for another thread), Yeah! :-)
I'm gonna run, create an avatar to build this theoretically perfect turntable in cyberspace. Not shure I'll be digging the music over there though...

Cheerio,

Frank

Hi again,
Mark Kelly is correct about the tension differential, provided the thread/belt does stretch and that there are frictional losses in the platter/bearing system. I don't think he said that slippage was neccessary, but rather that it was unavoidable in real life(maybe I got that wrong...)
Slippage can be avoided, idler, belt or tape creep can't.
My comment on the counterpulley symmetry relates to Dertonarm asking for a debate in the theoretical domain. I made it the sake of debating idealized concepts rather than compromised implementations(called reality), please take that into account.

A single constant airstream working against "wings" will cause cogging. Two(or more), 180°(x°) perfectly out of phase air vents with proper orientation of the wings will cancel the cogging(don't want to be responsible for the maschining of that one...). But this isn't all that different from a conventional multipole motor.
A single leaf and a one impulse(or 0,5, ...) per revolution is different. One could achieve the same without the complexity of an air supply, the turbulence issues, etc. by using a very heavy platter, direct driven with the motor being switched on only infrequently , be it for a very short duration or ramped up, then down again, but only when the speed drops below a set deviation threshold. Oops, I see the word feedback on the horizon, mmh...
The inertia of the platter will be an intergrator again, the constant motor noise/mechanical jitter will be exchanged for an occasional pulse and the bearing losses ought to be zero ideally.
Low frequency pitch stability will once again be less than perfect :-(
There again, it contains compromises... I'm still tempted to build such a device(and sell it for --- ONE GAZILLION DOLLARS, buahahaha...!).

A pulsed air supply with transfer function matched "wings" won't cause cogging IF you solve the problem of syncronisation(and if there was no such thing as turbulence, a.k.a. CHAOS).

Now back to "real" work :-)

Frank

P.S.: Related to the original post: There are at least 6 different currently manufactered turntables that partially rely on creating additional drag on the platter for maximum speed stability.

Hello Lewm,

No, I don't agree with Dertonarm's contention. If you reread my posts, you'll find that I'm trying to point out that there are several ways of approaching the challenges to run a platter at constant speed, ALL of which are flawed in one or several ways, as soon as we are dealing with reality.

I am very much familiar with Chris's table, in fact I suggested the use of an eddy current brake acting on the platter as soon as he started making platters with brass or aluminum bottom layers and I heard both a prototype and the first production model in comparison to his then top of the line tape driven tt. His direct drive turntable offers exceptional speed stability(significantly better than his tape driven decks) and is proof that it can be done(and, again, look at one of the last generation cutting lathes).

I'd prefer to discuss real world experiences, but Dertonarm made such a big deal out of the supposed mediocrity of most, if not all contemporary turntables that there simply was no point in "getting real".

Strangely, I have not gotten a response from him...

Best wishes,

Frank

Hello Dertonarm,
>The "air-pulse" drive (no pulse would exclude a problem) has some considerable problems while looking close to a theoretical ideal first. It not only will require considerably (really serious...) periphery, but indeed a "assist"-motor to bring the (high mass) platter to requested speed first and than de-coupling.<

Weren't you asking about "all out",no compromise, no commercial consideration approaches? Did I not explain some of the caveats? Yes, the periphery would be expensive, but maybe not as expensive as it appears(remember, this is not about a commercially viable project).
The Onkyo PX-100M eddy current drive turntable does feature an idler drive just to bring it up to speed within half a revolution, then the eddy current drive takes over(and it has a break too). It can be done...
Platter speed needs to be monitored, which in turn can govern the force of the air(or any other driving) "pulse". It will have a VERY slow recovery time, but(that is where my chronometer analogy comes in) the disturbing influence of the drive system could be minimized.

To assess which principle is superior, we'd have to build two(or more) otherwise identical turntables that can be driven by more than one means. I have done this several times many years ago, Chris Brady has done this more recently, so his assesment holds merit, even if some may not agree with all of his peripheral design choices.

I suggest opening another thread to discuss that there is VASTLY more to building an excellent turntable then the drive principle, even though nothing else matters if the record doesn't spin at (as close to)constant speed to begin with.

Dertonarm, I didn't ask you any questions, I just showed that your approach is just as far from or as close to being perfect as some other approaches, at least from a conceptual/theoretical point of view. But that part of the discussion was doomed from the start... and yes, it's not neccessary to repeat yourself yet again.

Nevertheless, have an enjoyable and relaxing Easter weekend!

Frank