The first thing to deal with is the torque issue. The thing about high torque idler motors versus low torque belt motors is largely a furphy.
The output torque at the shaft of the Garrard motor is a little under 10 mNm. The output torque at the shaft of the 3W Hurst motor used by VPI is 26 mNm.
The reflected torque at the platter is the shaft torque times the gearing ratio. For the Garrard the gearing ratio is about 48 to 1 so the final torque at the platter is about 440 mNm. For the Hurst motor the gearing ratio is 18 to 1 so the final torque at the platter is 470 mNm.
The "low torque" Hurst thus has more torque than the "high torque" Garrard.
To add insult to injury, the Garrard motor slows by 10% when delivering that torque where the Hurst motor does not slow at all.
Mark Kelly
|
Oops
I looked at the wrong motor on the Hurst table, the actual motor used by VPI is the lower torque model so the figures are 16mNm at the shaft and 290mNm at the platter respectively.
This would seem to favour the Garrard but only if you find a 10% speed reduction acceptable. If you want the Garrard to slow by a lesser amount (say 1%) the useable torque drops almost proportionally - at 1% the torque is about 1mNm at the shaft, say 50 mNm at the platter.
Mark Kelly
|
I was thinking about the SL1200 series - the SP10s used push pull pairs on a dual rail power supply. Since there is no coupling capacitor it can't be the problem but it is still the case that the amplifier will be sensitive to the quality of its power supply. |
Albert
I measured those results in my lab on a Garrard motor I know is in good working order. I have also measured several other motors used in idlers and found very similar results. Since the results accord with the theoretical results expected from these designs I have no reason to suspect that they are anything but typical.
You are correct in saying that the torque numbers by themselves prove nothing, with one exception: they prove that the argument over motor torque does not provide a key to the sound of an idler table vs belt drive.
That variable being eliminated, we can now ask "what other characteristics of these tables may result in the sound we hear?"
I nominate two for discussion: the very small degree of mechanical creep in the idler transmission and the very high reflected inertia of the typical idler motor. Which is more important? I don't know. Yet.
I have designed a belt drive which has a very, very low level of creep. It is in the process of being built and will be on show at RMAF in the Galibier room if we get it finished in time. I do not expect it to sound like an idler table, I expect it to sound like a belt drive with the belt creep problems removed.
The drive design allows for the addition of a high speed inertial system. I expect that if this system were added the sound would change; it remains to be seen exactly what this change is and whether it is seen as a benefit. The inertial system is still on the drawing board, it will not be at RMAF.
One of the problems is that the "donor table" has inevitably been designed to perform with a different drive system. Accordingly I've asked Thom to audition the drive on his lowest model table on the grounds that this has had the least attention given to optimising the synergy between the drive and the other mechanicals.
Mark Kelly
|
Johnny
The net effect of the belt stretch is that it causes creep over the drive pulley, so we are talking about the same thing.
The direct effect which you postulate is reduced by the second order low pass filter formed by the belt / platter combination. The maximal velocity variation for a given length change is the product of the radial displacement produced by the length change and the corner frequency of the filter system expressed in radians per second. The numbers come out in the parts per million range. |
Albert that's not strange at all when you look at the way Technics implemented their servo drives.
They drive the motor coils with single ended amplifiers using capacitive coupling to block the inevitable DC offset. Any such amplifier is very sensitive to the quality of the coupling cap used because it is in series with the motor coil. If the impedance of the cap increases, as it inevitably will with age, the drive available to the motor drops which will in turn reduce the forward slew rate.
I agree that there are a lot of things going on with the various drive mechanisms. Where we might disagree is with your implication that this means they cannot be adequately analysed. The reason I was dismissive of the parts per million difference due to belt stretch is that the belt creep is around 1000 times larger, so there's no point in worrying about belt stretch - by the time you fix the creep problem the stretch problem is gone as well.
Mark Kelly
|
Johnny B
You don't seem to understand the function of a low pass filter. It is not possible for the belt drive flex per se to affect anything by more than a few parts per million for the reasons given.
A change in belt tension will however create belt creep and this effect will be around 1000 times larger. That was my point.
Mark Kelly
|
From the Websters dictionary 1905 edition: "A barbarous term used in school philosopy for essence" - gotta love a dictionary with that degree of vitriol.
My intended career was as an academic in philosophy. My chosen field was logic and the structure of consciousness, especially with reference to linguistics and mathematics.
One of my heroes is Willard van Orman Quine, who was a logician at Harvard.
http://en.wikipedia.org/wiki/Willard_Van_Orman_Quine
One of Quine's books is titled "Quiddities". I rather liked the term so when I established my consulting company I called it Quiddity Technical Services.
I don't like monikers but Audiogon insisted that I register as a business and that I take a moniker which referred to my business name, so there you have it.
Mark Kelly
|
And to take that theme and refer back to the OPs question, I might ask: how much of the total inertia of the system should the platter represent? What influence does splitting some or most of the inertia away from the platter have on the perceived "drive" of the system?
It seems to me that many of the designs which are known for having lots of what Albert calls "drive" have moderate mass / moderate inertia platters tightly coupled to a high inertia drive system. If I'm not mistaken Win had a similar concept in mind when he desgned his TT.
Mark Kelly
|
Dan
It would be giving something away, but you are on the right track. It's the design of the filter that takes all the time and effort in the Spice model.
Matt:
Q 1 Asked and answered.
Q 2 No.
Q 3 No.
Mark Kelly
|
Axelwahl,
On your first point: That one's easy. I take it as a truth universally acknowledged that if the motor were not running, frictional forces (of whatever source) would cause the system to slow down and in doing so it would lose kinetic energy.
It therefore must be the case that energy is being put into the system for it to remain at constant speed. The only possible source of such energy is the motor, therefore the motor must have the primary role in maintaining constant speed. All that platter inertia does is to reduce the slope of the decelerations (and equally the accelerations).
Your second point falls foul of a fundamental property of feedback loops, that the speed of the feedback loop must take account of the speed of the "forward" portion of the loop. If you interpose a low pass filter such as the belt and platter in the loop then you must slow the loop down to prevent oscillation. If you leave the belt and platter outside the loop then the loop cannot possibly compensate for losses in the belt / platter system such as torque dependent belt creep.
No free lunch.
Mark Kelly
|
Dan
For the kit designs I had two overriding criteria: they had to be as cheap as possible whilst still providing acceptable performance and they had to be almost universally applicable (hence your comment about platters / bearings). Unfortunately this means that they are a long way from optimised for any specific application.
For the bespoke designs I gather as much information as possible, to the extent of getting specific numbers for the rotational moments of inertia of the motors used from the motor manufacturers or in the case of one drive having to measure the numbers myself. I then build a model in a Spice program using some translational analogies and spend a lot of hours doing dynamic modelling.
Depending on the sophistication of the drive, the specific platter and bearing numbers can have some influence on dynamic performance but the most important parameters are the motor and its electrical control. When I am happy that I understand what's going on, we go to prototype.
The results? Well, I think you'll be surprised at what can be achieved with standard AC motors and belts even less compliant than your Mylar when the drive mechanics are understood. Similarly if a manufacturer sends me several samples of a high cost three phase motor and says "do your best then bill me" the results can be pretty good.
The downside is that the controllers end up being quite expensive. I don't know if there's a viable model for producing an aftermarket controller using any of these techniques.
Mark Kelly
|
T-Bone
Maybe you have me confused with someone else.
I have never advocated the use of any form of slip in a turntable transmission. I understand the theory of it being used in an attempt to reduce cogging but I think it's the kind of half baked idea propounded by people who really don't understand the mechanics of TTs. You will have noticed that there are a lot of them about.
I have pointed out many timnes that belt creep (and idler creep) is an inevitable consequence of using compliant coupling materials but that could not be construed as advocating its deliberate introduction.
Mark Kelly
|
HiHo
Here are a few things which I can prove to be true:
1. Newton's Third Law of motion holds for turntable motors so the reaction torque reflected into the chassis will be the mirror of the forward torque applied to the platter.
2. The variation in reluctance of a "coreless" motor is much smaller than that in a motor using an iron cored stator.
3. As an iron cored DD motor rotates, the servo loop compensates for the variation in reluctance by decreasing torque as the rotor pulls towards the lowest reluctance position and increaes torque as the rotor pulls away from that position. This happens many times per revolution, depending on the slot and pole numbers of the stator and rotor respectively. The exact quantum is the least common multiple of the slot and pole numbers.
4. It follows from 1 that the torque variation reflected into and propagated through the chassis is much smaller with an ironless stator than with an iron cored.
It is my conjecture that this phenomenon explains what you are hearing. Naturally I cannot prove this so I won't say that it is *definitely* the case.
Some support for this idea comes from some engineering work done at Sansui towards the end of the analogue era where they designed a DD with two counter-rotating platters to obviate the problem (called X-99 I believe).
Mark Kelly
|
Lew
The DP80 uses an iron core stator.
The distinction between Denon's "3 phase AC outer rotor motor" and Technics "Brushless DC with integral magnet platter" is largely semantic.
The Technics motor is a 3 phase outer rotor motor which includes a circuit in the motor which generates 3 AC waveforms due to the motion of the rotor. These waveforms are necessarily synchronous with the rotor. The waveforms are then amplified to a level determined by the PLL controlled servo loop and fed back to the motor drive coils. The PLL is fed by Technics famous frequency generator circuit.
Without access to the Denons circuit details I cannot say exactly how the Denon generates the frequency required to run its motor but I can say that it also employs a PLL controlled servo loop to slave the coil drive to a motion dependent signal, this time generated by a magnetic signal recorded on the platter (a primitive version of a rotary encoder). The loop presumably also controls the voltage of the drive amplifiers - if it did not the level of cogging would render the motor useless.
From a practical point of view the only difference would be in the fidelity of the drive waveform. The forward drive voltage in the Techics motors I've seen is fairly ugly, the engineers relied on the high speed of the FG servo to smooth the rotation. Denon's encoder is a lot slower so they would have to have a cleaner waveform to start with. They are both neat solutions to the central problem, neither appears to me to be inherently inferior to the other.
Mark Kelly
|
After chiding someone on VA over miunderstanding servo loops I should correct an error in my description of the Denon.
Where I said "it also employs a PLL controlled servo loop to slave the coil drive to a motion dependent signal...." I should have said "it also employs a servo loop which adjusts the coil drive, using a motion dependent signal...."
The master is the quartz reference.
Mark Kelly
|
Axelwahl
I don't have enough information to make an informed judgement about the SME's controller implementation.
Mark Kelly
|
Clarkie
I can't see where the distinction lies. In both cases you have a passive inertial element (the platter), an active source of energy (the motor, which also has inertia of its own) and a transmission (belt or idler) which links them.
The motor's function is to replace the energy lost from the system. The transmission's function is to adapt the speed of the motor to that of the platter.
Most of the distinctions made betwen belt and idler can be viewed in terms of how lossy the transmission is, it historically having been the case that belt TTs were made with a much lossier form than were idlers.
I think the inertia of the motor is also important but misunderstood.
Mark Kelly
|
Clarkie
You appear to be making two statements:
1: High mass platters improve the speed stability of belt drives
2. The transmission in an idler is stiff, so the motor is tightly coupled to the platter.
and following them with a conclusion:
3. Therefore "the two systems are completely different".
The two statements are reasonably uncontroversial, but I cannot see how they are supposed to lead to your conclusion.
The obvious implication from your statements is that belt drives are necessarily less tightly coupled than idlers and that the platter in a belt drive is somehow free to rotate in an uncontrolled manner. This is supported by a statement in your previous post where you said that a belt drive has "a motor that only really pushes when the platter slows".
This is completely wrong. If the two systems are designed to have the same drive compliance then they are by definition equally tightly coupled. There is nothing to prevent this being achieved in practice; that it has not been seen as a desireable goal by the designers is an historical artefact, not a matter of physical necessity.
Axelwahl
you have confused the terms slip and creep in my post. They are not the same thing.
Mark Kelly
|
Sorry that last part was measnt to be for Hiho not Axelwahl. |
Hiho
Consider the speed of three things: firstly the belt on the drive side of the motor pulley, this being the section which is pulling on the platter; secondly the belt on the non-drive side of the pulley, this being the section where the belt is feeding off the pulley onto the platter; and thirdly the surface speed of the pulley itself.
Belt slip is where the speed of both parts of the belt is slower than the surface speed of the pulley.
Belt creep is where the belt speed on the non-drive side of the pulley is slower than the drive side, so the belt creeps over the pulley to make up the speed difference. This is because the belt is stretched by the torque it is transmitting and that amount of stretch must relax over the pulley. If the belt were equally stretched on both sides there would be no tension difference and thus no torque transmitted.
Mark Kelly
|
Hiho
The creep is proportional to the strain in the belt divided by the parallel sum of the wrap lengths of the pulley and platter.
The compliance of the mylar belt is quite low, so there's not much strain present. The wrap length of your "pulley" is about 500mm rather than the 10 - 50 mm of conventional drives.
Your system will give about the lowest creep available (this side of direct drive of course).
Mark Kelly
|
Hiho
When I said "about the lowest creep of any system outside DD" I was thinking that the drive I designed for Thom Mackris was probably slightly lower. I was wrong.
I've run the numbers, assuming you are running a 1/2" x 3mil Mylar belt around two 300mm platters placed 400mm apart, and you win. We have slightly lower compliance (0.063 mm/N vs 0.117 mm/N) but you have a longer effective pulley length (236 mm vs 68 mm) so your creep is lower: 0.003 vs our 0.005, referred to a radius of 150mm.
As a reference, a typical belt drive system might have a creep number around 5 - 10 and a good idler system (like Win's) will have a number around 0.05, over 100 times better than the belt drive.
Mark Kelly
|
Lewm
There is effectively no damping effect available from the two opposed magnets. Eddy currents are proportional to field strength times relative velocity divided by electrical resistivity. The configuration used results in low relative velocity. The electrical conductivity of available magnetic materials is quite high - NdFeB magnets have about 100 times the resistivity of copper.
For all: A distinction needs to be made between velocity dependent drag, which will tend to stabilise the system against speed variation and classical friction (velocity independent) which will not. Hydrodynamic drag and eddy current effects are in the former camp, standard bearing friction and stylus drag are in the latter.
Mark Kelly
|
Lewm
Win correctly described the influence of the Verdier's deliberate use of a trapped layer of lubricant to contribute hydrodynamic drag which lowers the Q of the platter / belt system.
That he did not use exactly those terms is neither here nor there - I have chosen my words to highlight the difference between two types of resistance which are usually lumped together as "friction".
Mark Kelly
BTW (on) usually means momentary.
|
Lewm
The easiest solution I can see to your problem is to have the switch control a relay. If you have lowish DC voltages available (preferably either 5V or 12V) this is easy to do.
You'll need some logic to convert the momentary action to a latching action but it's not hard to do. Since I have B+ delay on all my valve gear I use mains switching by relay as a matter of course anyway and I like the safety aspects.
Mark Kelly
|
