Vintage DD turntables. Are we living dangerously?

Totem.
I first posted this on another thread.
Here updated for relevance....

To me it is obvious that a pod will move due to Stylus drag. The question is how much.
In order to calculate this I needed a figure for stylus drag. A search on the web proved inconclusive but then ironically the answer came from the original timeline thread. There, one TT is mentioned where specific data is given on the amount of laser pointer movement per revolution and its distance from the centre spindle.
This TT is a beautifully engineered machine with, from memory a 22 kg platter driven by a fractional horse power motor via a thread. Hereafter I will call this TT. "TD"

The specifics were 2 mm movement on a radius of 400 mm, per revolution.
With this information it is possible to calculate the retardation torque and hence the drag. From this it is possible to calculate how much the pod moves.

So assumptions......
A pod similar to Halcro's is used on TD
Platter 22 kg of uniform section
Pod plus tone arm 11.5 kg
Height to record surface above mounting surface 150 mm
Pod feet 100 mm spacing in a equilateral triangle
Pod/ arm CofG, 75 mm above mounting surface
Pod CofG Central inside the mounting feet
Pod feet are not adhered to the mounting surface. ( no penetration of the cones into the shelf )
The same arm and cartridge used on TD is used on the pod.
TDs motor only provides enough torque to maintain original speed before stylus is lowered, after it is lowered.
Stylus is lowered at a radius of 140mm
Platter has a diameter of 320mm

The first answer is the force applied to the platter to cause this retardation torque. This works out at around 0.0031 newtons. Actually a large number under the circumstances but it is slowing a 22 kg platter!
Using this force and applying it at a height of 150 mm to the pod we get a tilting of 8.2 microns towards the platter, more or less.

Observations.
With the stylus at a radius of 140 mm, the two front feet of the pod do not appear to be a right angles to the arm when viewed from above. This will reduce the tilt a little. It is unknown what happens to the magnitude of the stylus drag as the arm tracks towards the centre, so it is possible that the pod will tilt as calculated once the arm is at right angles to the feet, assuming that this happens before the end of the inner grooves.
As an aside the tilting at a radius of 140 mm produces a yawing effect on the pod such that the arm rotates approximately along its axis. This effect is caused by the configuration of the feet. It is tiny and likely insignificant.
The calculations assume that TDs motor does not sense the slow down and produce restorative torque. Since it is a synchronous motor it will act to try and maintain speed. This will put more energy into the system and increase the tilt.
The pods appear to be slightly crescent shaped. If this is the case the CofG will be biased towards the two feet closest to the platter. This will increase the tilt.

The CofG is probably higher than shown here due to the feet and the weight of the arm. If it is the tilt will be greater.

The calculations assume that TDs platter has a uniform section. If its radius of gyration is larger or smaller than this suggests, the tilt will similarly be larger or smaller.

If TDs arm and cartridge was fitted to a pod and used on Halcro's TT, things would be different again. This because the TT-101 does NOT slow down. It is putting even more energy into the system, so the tilt would be larger.

The calcs assume that the motor assy cannot move. In Halcro's case it can and will.

The 8.2 micron figure is an average. The pod will move more or less depending upon the groove modulation.

Actual dimensions and weight of the pod will materially change these numbers.

I do not know if the amount of movement is of any significance, but yeah baby, it moves with the grooves. This compromises one of the three ideals I mentioned featured in the mythical perfect TT. This was my starting point in these discussions.

Hiho

I have a customer who has both a worked MK3 and MK2, giving him a unique perspective.

In his opinion the performance gap between then is reduced, post upgrade.

IMO, it is not in the design architecture, it is in the build. The two motors are fundamentally the same design but the MK3 has been fed growth steroids.

Cheers.

Hiho
I think that the greyness is produced, amongst other things, by errors in the speed sensing mechanism. These are due to the way they are put together which can be corrected. I don't think that "pulse" is the correct term for this phenomenon.
IMO the technics TTs get into trouble because they have so much correction torque available. They can respond faster than most TTs to error signals. Even if these error signals themselves are erroneous.
My treatment method is applicable to any DD TT.

The term " jittery" you use is spot on and is what I use to describe the effect on my krebsupgrade web site.
Once heard it cannot be unheard!

Hiho
Thank you for your considered response.

Re the contentious subject of cogging.
Coreless motors cog, as accidentally powering up a Goldmund Studio without its platter, clearly demonstrated. ( JVC 4 pole, 2 phase, coreless motor). It was a dumb thing to do, but informative never the less.

The problem that you rightly ascribe to the SP range is not IMO cogging.
It is way too high in frequency to be so. The motors have 15 stator poles, don't know how many mag poles, but some higher number.
Making around the mid 200 HZ of power pulses per revolution. But they are 3 phase motors, each phase separated by 120 degrees. The sinusoid supplied phases slide into each other, improving linearity considerably, so "power pulse" would seem to be the wrong term to use.
The SP10s higher pole and phase count than a number of coreless DD motors, would imply, for a given output torque, lower amplitude but higher frequency cogging.
( The motor only delivers the torque asked of it under the load conditions at that moment in time). If it delivered higher torque than the load demand, the platter would accelerate.
I hear the problem in a standard SP10 in the kHz range. This is a feedback speed sense issue and is not intrinsic in the motors architecture and any cogging that it may produce.
It can be fixed.

Lew.

Yes the SP10s are DC motors. Maybe the confusion arose when I said that they are 3 phase. The supply voltage does not swing across zero.

cheers.

Lew.

You may be onto something with the multi-prong plugs. It is likely that the individual male and female pins are crimped. Over time with dissimilar metals, wire and pin, along with a bit of humidity, electrolysis is possible/probable. I have with some gear pulled these connectors apart, cut the wire and soldered it back on to the pin. It is a real PITA since the new connection has to fit back into the plastic holder.
You could of course simply bypass the plugs altogether and solder the pairs.

Good luck with the Witch hunt.

Schubert.
Ah yes. For me it was the SR-929.
A work of visual art.

Fleib.

I think that we have done this subject to death.
Let's agree to differ and move on to some other topic?

BTW, apologies for misspelling your moniker. Big fingers, small keyboard.

Cheers.

Pryso
My upgrade concentrates on making the motor and speed sensor do what the designer intended them to do. As designed and as built are two entirely different things. All the Kings horses cannot put back together what is broken in the drive. The technique is applicable to any DD. The TT-101 would be a good candidate

Re mats. In earlier TTs that I built, I experiemnted with rubber, lead, chamois and acrylic. I gravitated towards acrylic. Then more than 20 years ago I purchased my SP10 MK3.
Didn't like the original rubber mat and couldn't use the chamois because of the platter lip. I found that the acrylic worked quite well. I didn't think of trying copper or SS.
But I felt that the platter itself was compromising things so I made a new one in a tri-laminate of acrylic, duralium and lead epoxy'd together.
This I liked a lot and is what I still use today with a SME reflex clamp.
Pictures available on my Krebs upgrade web site. The black triangular unit.
All that said, I have heard exceptional performance from MK3's with original platter and both SS and soft mats. These were both take no prisoner systems in the USA that are simply spectacular.
So today I would think twice about discarding the original platter and spend more time on the record / platter interface.
There are so many options and ultimately it seems to come down to personal preference.

The linear temperature coefficient of expansion of Aluminium is 0.000023m/m degrees C and is indeed approx. double that of steel.
Using this figure on a LO7D and assuming a 5 degree C delta, we get a change in distance to spindle of around 0.04mm. (approx 1.6 thou inches) Further the arm is lengthening and shortening relative to the ambient temp. This would somewhat mitigate the distance change.

Of course if one goes the pod way, the shelf's thermal expansion, contraction becomes the decider of the dimensional change.

The thermal expansion "problem" has been grossly overstated.

Halcro.

You accept without question that a stylus can slow a 22kg platter. This, while said platter is being driven by a substantial motor. Why can't you accept that a lighter structure can be rocked on its feet by the SAME applied force, which now has a mechanical advantage?

I am not saying that the movement is a problem. What I'm saying is that it exits. It is clear that many people get brilliant results with this type of setup. I myself have auditioned rigs like this and liked the results. But that does not detract from the facts. The pod moves

Check my math if you like, the numbers are sound.

Fleib.

The TT in question has a very powerful motor, capable of rapidly accelerating the platter.

I agree, 8 microns is tiny and this subject is tired.

cheers.

Halcro.

Lighten up, this is just a friendly discussion.

Cheers.

Fkeib. I agree, 100% torque conversion is not possible.

That is why I listed this assumption..."TDs motor only provides enough torque to maintain original speed before stylus is lowered, after it is lowered."
I have assumed that the motor plays no part in dealing with stylus drag. It is only delivering enough torque to overcome windage and bearing friction.
Any additional friction in the form of stylus drag is resisted by the stored energy in the rotating platter. That is why I needed to calculate its moment of inertia.
This makes my calcs conservative, since there will be some restorative torque from the motor. Since we don't know how much, it was considered best to ignore this parameter.

The info I did find online relating to calculating stylus drag used a free wheeling platter, running at rated speed. Timed to stop without and then with the stylus lowered. My calcs are a derivative of that method.

I don't think that Maynard and I would agree on very much at all!

Dover.
The figure I used for Aluminium is a generic number we use for calcs here. If we want to be more precise we take into account the method of forming the material and of course the specific alloy. I did say "around 0.04mm"
The point I was making, was that the dimensional change is a tiny fraction of the several millimetres per degree c that Halcro sighted.

When I made the decision to build my TT plinth out of acrylic the problem of stiffness was considered. My TT uses 2 x 30mm thick Acrylic sheets, separated and fused to a lead spacer. By separating the two structural plates a form of 'I' beam is produced since shear between the two plates is strongly resisted.
This is a similar idea to that used in the light weight wooden 'I' beam floor joists. Individually the three parts of the beam are quite flexible. Gluing them together in this form however makes a very stiff structure.
The same principle is at play in plywood construction, which my TT copies.

The triangular shape with the motor centrally mounted puts the bulk of the material around the motor, concentrating the strength in that region. It then tapers towards the edges progressively reducing in strength as it approaches the feet. A square plinth, however, will be more sensitive to the problem you sight. If in fact it is actually a problem.

My room is air conditioned 24/7. Temperature is tightly held

It is a 25 year old design and I am still happy with it. That said, if I was to build a ground up TT today, just for fun, I would likely use different materials and architecture.

Jmowbray.

More information on the upgrade is available on my web site. The procedure is now available for the SL-1200 series.


Please PM me or contact me via my Krebsupgrade web site for a detailed description.

cheers.

Dover.

Please read my post again.

I did not say that one should make an I beam with stacked ply.
Ply AND "I" beams use a similar technique to separate the outer layers where any bending causes compression in one layer and tension in the other.

Why are we debating this???

Fleib.
To add a little more to your question. I took a lot of notes when upgrading a customers LO7D. One surprising observation is that the motor stator is rubber mounted via grommets. Such that it is possible for the stator to twist a little backwards in an anticlockwise direction when applying torque. This would seem to be counterintuitive but it is there in the design.

I don't know about the 3% servo thing, but it has been quoted many times in various posts. If this is how it is built, it would be a reasonable assumption that it is effectively open loop once up to speed. In this way it is relying on the synchronous motor's innate speed accuracy and the platters inertia to maintain the correct RPM.

Chris74.
Thanks for the expansion of details on your SP10MK2 platter.

Further, I agree on the robust nature of the bearing assembly. Both it and the MK3 are substantial indeed.

Geoff.
I see two main structural paths in a TT... One from the record upper surface, thru the platter, bearing, plinth (or shelf for those of us who prefer au naturel ), arm, cartridge and stylus. "The loop path"
The other, from the record surface to Mother Earth "The ground path".
For now, ignoring suspended TTs

I have used lead extensively in my TT designs, but it has always been to " laminate" these two paths and has not been inserted into either of them.
I have found this methodology to work well.

Cheers.

Lew
The rubber grommets look original.
I will PM you a pic of the stator to compare with yours.
Inserted in the centre of the grommets is a small metal tube stand off to prevent the grommet from being crushed when the bolt is tightened.
There are also a couple of nylon? stand offs to prevent the rotor from contacting the stator windings when the platter isn't in place ( partial mag lev feature)
Anyway, slight stator rotation is permitted by this arrangement and if original it is an interesting design decision.

Halcro
Can you expand on the comment re VPI cutback in production of their DD TT
I understood that they were selling well. It has certainly received good press.

Thanks

Halcro.
In the manual of the ET2 Bruce Thigpen claims that a pivoted arm contributes to wow and flutter due to its geometry.

The TT-101 has exceptional figures in this area.
Since you have a rig with 3 arms, it would be interesting to run 3 consecutive tests of say 1 minute each with the three arms. This to to see if there is any difference. Each arm has different geometry, so if Bruce is correct, we may expect to see a difference between the three readings.

Purely for academic interest.

Halcro.

I re-read my last post and realise that it could be misinterpreted. Clarification here...

BT suggests that when accessing speed accuracy with a pivoted arm, the measured performance is inferior to the actual performance. This due to the geometry of a pivoted arm.

Since you have a rig with three different arms, it would be of interest to run three sequential tests to see if there are any measured differences.
If there are, maybe we could infer that the TT-101 is actually better than the readings indicate.

Fleib.
Copied here the reference to wow and flutter from the ET2 Dampening trough owners manual.

Food for thought.

"WOW AND FLUTTER
Wow and flutter, FM distortion and surface irregularities in the LP should all be grouped
together because, as we will see, they are all tied together.
When you cut a pure tone (say 1kHz) onto an LP and then play it back on a
turntable/tonearm/cartridge system, you would hopefully want 1kHz to come back. Something
close to 1kHz comes back, but rapidly being shifted up and down around 1kHz. If the frequency
is shifted up to 1001Hz and down to 999Hz within a short period of time, the amount of shift is
.1%. If the shift occurs less than 10 times a second, it is considered as flutter. The two measures
are generally lumped together and called wow and flutter.
“Weighting” is applied to the measurement to reduce the measurement’s sensitivity to very
low and very high rate of frequency shift. The actual amount of frequency shift is much greater
than the number implies. The weighting network is supposed to create a number related to a
subjective ability to hear wow and flutter.
Reviewers have incorrectly attributed wow and flutter to the turntable. Since the advent
of the belt drive turntable, wow and flutter has been purely a function of tonearm geometry, the
phono cartridge compliance with the elastomeric damping, and surface irregularities in the LP. In
our own lab we have measured many high quality turntables using a rotary function generator
directly connected to the platters of the turntables.
The measured results are usually an order of magnitude better than the results using a
tonearm and test record (conventional wow and flutter method). Further proof exists if you take
two tonearms, one straight line and one pivoted and mount them both on the same turntable. The
straight line tonearm will give a wow and flutter reading with the same cartridge/test record of
about 2/3 to ½ that of the pivoted arm (.03% < .07% to .05%). This is because the straight line
tonearm has a geometry advantage and lateral motion does not result in stylus longitudinal motion
along the groove of the record.
Another proof is to take two different cartridges, one high compliance and one low
compliance, and take measurements with both using the same turntable and tonearm. The reading
of wow and flutter will be different. All wow and flutter readings are higher than the rotational
consistency of the turntable
A damping track applied to a tonearm (straight line or pivoted), will reduce the measured
wow and flutter usually 10-30% and sometimes as much as 50%. ET-2 wow and flutter readings
with a typical cartridge and good turntable will usually measure (.02 to .04%) which is extremely
low for an LP system. With the damping track installed flutter readings with the ET drop still
lower and with one test record we measured readings as low as .007%.
Surface irregularities on the vinyl of the LP record are the primary cause of rumble or
random low frequency noise, which causes the tonearm/cartridge spring system to start
oscillating. This oscillation occurs continuously during playback. It is a primary cause of wow
and flutter and FM distortion in phono playback. Surface irregularities occur not as a part of the
record cutting process, but result from the molding process used in making the record
You can see visually small ripples on the surface of an LP as it is turning. These continuously
excite the tonearm resonance"

Halcro.

Thanks for the tests. Do you have the companion numbers for each arm?
I was interested in any consistent differences between the arms/carts, which could go some way to validating BT's ideas. There was nothing else in my request.

As for Dover's and my comments about what is happening "between" each pulse of the time line. I was hoping that this topic was in the 'agree to differ basket', but since it has been raised....you only need look at the traces you have just posted. Sharp spikes on the raw trace... this is a servo in action! Rapid acceleration/deceleration of the platter. Yet the platters average speed is 33 1/3. The smoothed ( green) trace filters these spikes, so it tells only part of the story.

cheers.

Halcro
The Wilson does have speed feedback, don't know about the George.
If you look at the George's raw trace, even though it is all over the place, it is quite smooth in comparison to the others and approximates a sine wave. This would imply a different speed control architecture.

The key point is that that the traces you posted clearly show that within a single revolution, the TT-101 has significant very short duration speed changes, but it's AVERAGE over one revolution is exceptionally stable.

Cheers.

Halcro
Look at the raw trace for the WE8000.
Start at the first lower min freq, just above 3130 hz. Other than the max at around 3164hz, count every sharp change in direction until immediately before the next min of around 3130 hz again.
I count 14. The platter changes speed 14 times during that single revolution.

Cheers.

Halcro.

The TT-101 is a machine, a thing, nothing more. It, like all machines, is far from perfection. My objective comments, made as a result of properly interrogating all of the Feikert test data you provided, are in no way a reflection on its owner.
The same cannot be said for your attempt to censor me by suggesting I "ignore these discussions" because I posted facts that you find uncomfortable.

Fleib.

The error resulting from record eccentricity is surprising.

Take a small 0.5mm eccentricity on a 100mm radius and we get around a 1% error. (The tracking radius makes a difference)

Nakamichi were on to something way back then.

Dover.

Yes "removing the counterweight" would result in a dramatic degradation of the sound!

As to increasing the horizontal mass of my arm by 300%. One needs to specify at what horizontal frequency this measurement is taken.
At 0.55Hz (eccentric record at 33 1/3 rpm) the leaf spring on the ET2 counterweight is stiff. IOW on a standard ET2, at a horizontal excitation of 0.55Hz, the mass of the counterweight assembly must be added to the weight of the spindle, wand and cartridge. This means that when tracing an off centre record, my cartridge sees more or less the same mass as one mounted on a standard ET2. (applies to average weight cartridges and associated counterweights)
BT has confirmed this and I posted his response on the ET thread, maybe you missed this?

Resonance transmissibility theory 101

cheers.

Halcro.

The explanation from Marcus confirms my comments re the 14 speed changes in the first revolution.

Here is the relevant sentence...

"That's what the spikes are coming from: it's a superposition of eccentricity and "real" WOW and flutter."

The raw data tells us much about the nature of the platters rotation. We see the "real" WOW and flutter as distortion of the sine wave.

In other words, perfect speed would show just a sine wave symmetrically centred about the 3150hz line on the raw data graph. This then being filtered by their program to produce a straight line at 3150hz for the low pass graph.

cheers.

Halcro.

Perhaps you misinterpreted my last post.
When I said "perfect speed" I did not mean "perfectly centered".

A real world test record, which is off centre, produces the sine wave. Speed errors are superimposed on this sine wave, as Markus says.

So with a real world test record running at perfect speed, we would get the results I described.

cheers.

Go the ABs!

Hiho et el
I feel that I need to put my two cents worth in on the subject of cored and cordless motors. Some one has to act as a champion for the Technics line. I should also state that I have a commercial interest in the subject, so feel free to discount entirely what I am about to say....
My take on the subject is not between cordless and cored but high torque relative to the platters moment of inertia and low torque relative to same, coupled with the feedback design.
In other words the motors grip on the platter. It's "responsiveness". The Kenwood and JVCs live in the low responsiveness camp. The technics SP 10 range are firmly in the high responsiveness camp.

The SP 10 range in standard from is noisy, not in the conventional sense but noise which is a function of the music being played. There is also a tension, stress if you will, to the presentation and a greyness that over time is downright irritating.
But these faults are not intrinsic in their design, rather it is in their build. These are two completely different things.
I want drive and punch when I am listening to popular music but I want finesse and nuance when I am listening to my favourite genire, baroque.
I listen to music to connect and feel an emotional response. To feel the joy of discovery when the violinist bends the note just..so.
I want it all.

I could not live with a standard SP10 MK 2 or 3. I don't have to, but I very happily live with my worked SP10 Mk 3.

Cheers.

Lew.
Thought that you may find this amusing. Here on the perimeter of the world we spell gary, grey (Gray even comes up with an auto correct spelling error as I type this)

>>>>>>>>Gray vs. grey

Gray and grey are different spellings of the same word, and both are used throughout the English-speaking world. But gray is more common in American English, while grey is more common in all the other main varieties of English. In the U.K., for instance, grey appears about twenty times for every instance of gray. In the U.S. the ratio is reversed.

Both spellings, which have origins in the Old English grǽg, have existed hundreds of years.1 Grey gained ascendancy in all varieties of English in the early 18th century, but its dominance as the preferred form was checked when American writers adopted gray about a century later. As the Ngram below shows, this change in American English came around 1825. Since then, both forms have remained fairly common throughout the English-speaking world, but the favoring of gray in the U.S. and grey everywhere else has remained consistent.

Some people make their own distinctions between gray and grey. You can find some interesting examples in the comments below. There is nothing wrong with these preferences, but they are not borne out in broader usage. For most people, gray and grey are simply different spellings of the same word.

Both spellings are used for the participles, grayed/greyed and graying/greying, as well as for most of the words and phrases involving gray/grey. For instance, grey area/gray area, referring to an area having characteristics of two extremes, is commonly spelled both ways. So is graybeard/greybeard, referring to an older man with a beard, and gray squirrel/grey squirrel (which refer to closely related types of squirrels on opposite sides of the Atlantic). There are at least a couple of exceptions, though: greyhound, for the breed of dog, always has an e, while grayling, which refers to several types of fish, always has an a.<<<<<<<<<<<<<<<<<<

Dover.

The essence of an I beam and ply is the separation of the two outer layers (sheets). Any bending causes compression of one sheet and tension in the other. This increases the stiffness of the structure.
As I said earlier. It is a 25 year old design. Were I to design one today, it would use different materials and architecture.

(Dover "To be clear what was being measured in my post on the Timeline thread – the 2mm lag was generated by setting the TT speed with no stylus playing and then measuring the lag when playing. A 2mm lag at a radius of 400mm is a speed error of 0.08%") and ("In summary, your maths is wrong because your calculations are based on a misinterpretation of my Timeline test results.. When the 26kg platter speed was set with the stylus playing, there was in fact no measurable retardation.")

My post calculates the retardation torque required to cause the slowing, from stylus raised to stylus lowered and then applies this to the pod.
What happens AFTER you adjust the speed with the stylus lowered makes no difference to the calculated pod movement.

Re the CofG of the crescent shaped pod. I didn't make this note clear. It appears that the feet have been moved back towards the rear of the pod, moving the CofG closer to the front feet. The photos are a little ambiguous. If this is not the case then yes the CofG will move away from the front pod feet reducing the tilt.

The platter is heavier than I thought. If its radius of gyration is as I assumed, this extra weight will increase the torque required to slow it and by definition increase the pod tilt.

Gentlepeople.
I'm with Lewm.
"What is missing is an objective way of assessing platter performance that is truly meaningful"

There are so many different approaches out there and a bunch of them perform brilliantly, yet often they contradict each other in design philosophy. One would think that designers would gravitate towards a common theme if there was one path towards the holy grail..... It seems not.

On the topic of adding weight above the manufacturers original design. The thought here that increasing (or decreasing) the moment of inertia, "I",too much would upset
the time constants and gain profile in the feedback loop. I agree that it should be considered. I took great care when discarding the original SP10 MK3 platter and replacing it with my own design, to keep "I" close to the original.

The LO7D is a special case since the outer ring clamp greatly increases the platters moment due to the concentration of its mass towards the circumference. It therefore makes sense to offer a switchable setting for this.

On the other hand Artisan Fidelity offer a, presumably heavy, copper matt for their SP10's. Albert Porter uses a Stainless Steel matt on his simply stunning MK3 rig. A customer of mine, Pass Labs, makes a stainless steel platter to replace the original SP10 MK2. From photos I have seen it is likely that "I" has been increased.

With the SP10 range, at least, there seems to be a high tolerance for different platter moments.

Now that Dover and Halcro seem to have made up. Can we all agree that the spikes in the sine waves from the raw data are actually showing speed changes, W&F?
Secondly, as a family, TTs with tight speed correction tend to show greater aggressiveness in said speed changes?

Aigenga.
re unipivot arms.
You raise an interesting point, different arm bearings may well show a difference in the graphs. (Arm geometry may also show difference, as BT argues.)
That said, have any of the rigs recently listed by Halcro had unipivot arms?
The W&F visible on the raw data for the TT-101 seems to be quite similar from arm to arm. This would imply that the speed changes are localised in the drive unit itself.

Dover.
I heard the Final enough times at Warwick's to get a fix on what it did well and what it did not so well. Hearing it at your place only served to confirm this.
re the amazing comment. Under the right circumstances I find my car stereo to be "amazing" . We adjust our expectations based on where we are.

Re speed stability, Your comment re DD TTs needing servos to work is immaterial. It is just a different technology that's all.

I am on record during a conversation with Lewm stating that I could not live with a standard SP10 MK3. (To my ears they have a high frequency jitter speed problem that is just plain irritating, which took me 15 years to tame.) I will now expand that statement to say that I have not heard any standard DD TT that I could live with. So perhaps Warwick and I are almost on the same page.

Could I live  with the Final? 
While I admire what it does well the, to me, obvious dynamic speed problems disqualify it from my bucket list.

As I said, we each make our choices based on our particular biases. I have no beef with anyone preferring a different TT to what I would choose. This diversity alone makes our hobby fascinating.

Dover.
yeah yeah we have all heard ( read ) the arguments extolling the virtues of BD,TD and ID over DD.
My point was that a synchronous motor in a thread drive has local error correction and under dynamic loads it is in play (correcting) all of the time, just like global error correction in a DD. Error correction, something that you seem to dislike with a passion, is clear and present in your own TT.
Different in nature yes, but it is feedback 

Further the fact is that the Final slows when you lower the tone arm. This you have published. You say that this can be adjusted out, but this adjustment must only be a best guess average. So there will be dynamic speed changes since modulation level and the radius read are not constant on an LP. IOW the retardation torque is far from constant.

This is the dynamic speed problem that I hear on the Final.

While one can debate the efficacy of the time line in showing what goes on within a single revolution, Halcro has quite clearly shown us all that a good DD TT maintains speed under load. The Final does not and high platter inertia isn't enough to resist this slowing 

Yes some cutting lathes use a gear principle and then many are DD.

Dover
I have listed the time line of my upgrade development on another thread. There I outlined why I persisted with the SP10. To repeat. It's "drive " is intoxicating. Halcro has used a different descriptor for his TT-101 but he and I agree on the point.

BTW a sine / cosine generator is necessary to create a rotating field in the two phases of your synchronous motor.
In its simplest form it can be approximated by a capacitor.
There have, for decades been more accurate devices available to generate the necessary phase shift, so what you have in the Final,  while being elegant, is not a big deal.

Many DDs have three phase motors. This requires a precision triple sine wave generator. Each wave shifted in phase by 120 degrees. These are then fed into a 3 channel power amplifier.
Since they are direct drive these three phase signals must be very accurate, if they were not the platters rotation would be very erratic.

Most quality DD TT manufacturers did this 4 decades ago.
Again it is not a big deal, but they at least provided the purchaser with the power amp.

Also a synchronous motor like yours has local feedback. The rotor lags slightly the rotating field.
With a varying load this phase angle changes and then corrects, hence incremental speed changes.

There is no free lunch when we try to drive something at a constant speed.

Raul
Welcome back, we missed you!

Beyond core functional test numbers,  I do not put a lot of stock into equipment specifications. I prefer to use my ears. As above... We do not listen to specifications.
If we did, tube equipment for example, with its typically inferior measured performance, would be off our Christmas shopping list.
My view is that we can hear way deeper into the output of a product than measurements can articulate.

If you are interested in more detail on my upgrade, please feel free to PM me.

cheers 

Halcro
We have been thru the thermal expansion thing already. As I stated some months back, with an aluminium chassis, the change in spindle to arm distance due to delta temp is very small. Dover is correct, it is no where near a "few millimeters"
Of course the shelf on which you place your TT isn't immune to dimensional changes due to temperature. Don't know the material you use, but the change is also likely to be insignificant.

Cheers.

Rwwear
Many years ago I made a leather chamois mat for one of my home brew TTs 
I purchased a car cleaning chamois from an auto parts store.
A few dollars from memory.
I agree with Banquo that the best results were with the suede side up.
Dont know how this option would compare with the real thing, but it's not going to break the bank giving it a try.

cheers 

Fleib
Many years ago I decided to completely replace the platter on my SP10 MK3
The new one consists of a trilaminate of acrylic on the top 15 mm of lead and then duralium.
The  duralium interfaces with spindle and spigots to the lead and acrylic keeping thinks nicely centered. It also houses the rotor magnet and speed sensor/commutation assembly.
I spent a lot of time before hand experimenting with glues. I settled on an epoxy where I adjusted the ratio of the two parts to create a glue that was around the same hardness as acrylic. The idea being to simulate a laminate where the lead was fused to the acrylic. This rather than introducing a lossy layer of glue between the two materials.

The total weight is around 10kg as per the original. Moment of inertia is slightly higher.

I like the result. That said back then I did not experiment with metal or graphite mats. I now know that a SS mat works extremely well on the SP10

cheers 

Fleib
The speed sensor consists two parts..., 
The sensor coils which are stationary and a toothed magnet which is attached to the platter.
Sorry, I wasn't clear in my earlier post.

Cheers