Saturday, January 25, 2014

Turntable Drive Systems

Here's one on which currently my friend Tim disagrees.  He likes a short Rega belt drive system as being optimal.  He claims the platter mass is the best defense against wow and flutter.

I'm not so sure.  I remember reading about grove drag speed modulation in Peter Moncrief's International Audio Review (IAR) in the early 1980's.  The Japanese had realized the importance of that issue in the mid 1970's, if not before, and designs like the Sony TTS 8000, Technics SP 10 Mk II, and Denon DP 6000 were taking it very much into account with massive torque finely delivered, and proclaimed in specs.  For example one turntable claimed 0% deviation in speed with up to 150g of resistance.

Peter was worried about belt drive, IIRC.

(Update: Peter wrote again (?) on this very same topic in his latest online IAR.  Unfortunately this review hasn't been completed in a long time.)

I think looking just at platter mass is the wrong thing.  One has to see the drive system first as a complete system.  In that system, the platter mass serves as part of a low pass filter, as a parallel capacitor.  A belt or idler serves as a combination of resistance and inductance.  But the belt or idler has dynamic properties which may further help or hinder.

Now belts and idlers have different properties.  I believe it is probably easier to get better motor noise reduction with a belt.  That's because it's easy to get arbitrary amounts of "inductance" (stretchiness).  And one can use a weak motor (high resistance).

But I believe there is also more lag with regards to motor correction stability.  Any errors take some time to work themselves completely in a belt drive system.  An idler wheel doesn't vary as much.  There isn't much stretchiness, and the motor is high torque.  Also the idler rubber recovers instantaneously, it doesn't have to work itself out, re-stabilizing the entire length of the belt.

Belt drives where the motor isn't fixed to same plane as the platter also introduce the possibility that timing may be slightly thrown off by suspension movements.  Idler drive motors can move, but the range of motion is very limited, usually restricted to vertical only, and this has little effect on timing.

So why did people hear better timing with belt drives?  I argue they didn't.  As Arthur Salvatore says, the Linn simplifies the music.  In less complex music, the rhythm and timing becomes a more significant, ultimately the only remaining part.

Small degree of flutter caused by belt drive is exactly this…a simplifying effect.  Small timing errors are another way of raising the sound floor, of covering up micro details.

Update: I've now won a Lenco L75.  I will shortly have two working late generation direct drive tables (Mitsubishi LT-30) and one non-working one I hope to get repaired (Sony PS-X800).  I still look at listings for the top direct drive models from Technics, Denon, and Sony.  I got myself the two Mitsubishi's (one for backup) and the Lenco for my birthday instead of getting the Rega RP6 or Michell Gyrodec.  I have placed a bid on a high end tonearm for the Lenco.  After I win a tonearm, I'll order an appropriate plinth for the Lenco.  My initial hesitation in getting the Rega had mostly to do with their anti-record-clamping philosophy and lack of VTA adjustability (which I see as a conspiracy to deprive us of the right to use other phono cartridges than theirs).  My hesitation, if you can call it that, in getting the Gyrodec had mostly to do with the total cost of a loaded and new Gyrodec system, $6000 or more as I had envisioned it, including upgrades and SME 312 arm.  But as a result of my new distrust of belt drive systems, as described above, which only a few weeks ago I was making fun of  ("Idler drives rumbling into the future"), I'm not particularly interested in getting a belt drive now.  I only have some regrets in not getting a new and likely perfect for 10 years table as I think the Rega would have been.  Even the one Mitsubishi I have received so far has a minor but worrisome issue--the moving base of the arm makes a slight bumping sound when the servo moves it for linear tracking.  This likely is not audible.  If you're not into belt drive any more, there aren't as many new choices as there were in 1980, at the height of the Japanese direct drive table phenomenon.  The Rockport which Peter Moncrieff started to review is a custom built direct drive for $100,000.

Rigid arm to turntable coupling unnecessary

Another faulty part of the Linn LP12 religion is the claim that it is necessary to have a firm coupling between headshell back through the arm and the chassis up through the bearing of the turntable and the surface of the platter.  At some superficial level this may be true.  You don't want the arm and platter wandering away from one another.

But the argument as applied to making all the coupling very rigid is completely false, as that creates a connection that transmits not just position…keeping arm and platter from wandering away from one another…but vibration.  Vibration needs to be stopped from every source of vibration as close as possible from that source, and the last thing you want is vibration from the stylus traveling down the arm tube to the pivot, the chassis, the bearing, and the platter, or the vibration from the platter going the other way.

Pretty much everyone else understands this now, though some of the rigid coupling thing rears its head in the drive to eliminate non-commercial user conveniences, like removable headshells.  Imagine being able to change a cartridge without bringing your table back to the Linn dealer so he can use his special jig on the Ittok to prevent damaging the gimbals.

If you skip the rigid coupling thing, which is mostly wrong anyway (the arm must, should, and does move freely with its range of motion anyway, a slight variation in the surface of that motion is of little consequence, so long as it is not transmitted in audio band, and all the more that a pivoted arm doesn't even move in the correct surface anyway, to duplicate the cutting head, as a linear tracker at least approximates) there is little wrong even with old 1/4 cartridge headshells, which I might add were on some of the finest arms and tables (and still are!).

A standard headshell provides sufficient coupling…and it actually reduces resonances, and permits the easy addition of addition damping to the headshell…and even the cartridge body itself, where it is ultimately most useful.  It was the coupling argument which mainly lead to the disparaging of the removable headshell.  It also creates more dependency on the dealer, and saves money in arm manufacturing.

Like the arm mounting itself: damping here is a good thing.

Damping: Essential

My key problem with Linn products is that their eschewing of damping ruins their potential.  I believe the Linn Sondek LP 12 is a fair turntable, but it could be a good one with additional, even user applied (if done correctly) damping.  If one thinks of the LP 12 as a Scottish made and somewhat ineptly copied AR turntable (Linn got the suspension objectively wrong in at least one way, according to my friend Tim, they did not put the center point in the trajectory of the stylus, as the AR correctly did), then, AND ONLY THEN, does one have the right ideas of how to use it.  Not following the Linn anti-damping religion, which virtually everyone else in audio realizes is wrong.  Everywhere else you go, eliminating resonances through clever use of damping is being done more and more.  Everyone else brags about it.

Now I suspect the AR was better damped as well…but generally with less costly products, and likely even the venerable AR turntable, less damping (or most likely, none at all) is used than would be desirable.  (I think the AR may have been made with more fundamentally damping materials than the Linn, but probably additional damping would be useful even for the AR.)

So generally one expects with a lower cost product to have to do some rework, such as applying damping.  But in the case of the Linn, it is an expensive product, already perfected as much as it can be, so you are told by the dealer.  There is no damping not because Linn is cheap, you are lead to believe, but because damping ruins the sound.  (No doubt it does remove the "Linn" sound, added brightness and other resonances.  But that's exactly what I want to remove to have a product I could enjoy!)

Now the Linn has a practical problem with such rework.  It is a highly tuned device.  Even a small amount of user added damping is likely to throw off the suspension, thereby requiring a suspension adjustment.  The suspension adjustments have no readily accessible controls, and in fact they may be best leaved to a trained and experienced Linn dealer anyway.

I may (or not) keep my LP 12.  I'd keep it because they really aren't worth much anymore.  I have a product that may have cost the original owner $3000 or more if they started from an earlier version and added upgrades.  I'd be lucky to get $800 out of it (arm and table) I think now, even if it were working. It's almost worth keeping just as a museum piece.  It's actually minty beautiful.  It's finely made in many ways even if wrongly designed.  It's a pleasure to look at and fairly nice to use.

Now for that $800 I might buy a better sounding player, but maybe not.  And if I could overcome it's weaknesses a bit by damping resonances…all the better.  I'm sure with user applied damping, and perhaps other imaginable modifications, it could exceed the sound quality of a stock AR, Thorens, etc., many fine and highly regarded tables.

Though chassis damping is a bit unimaginable, as described, I might get away with a bit of arm damping.  I see Arthur Salvatore says the Ittok arm is quite good, better than the table actually.  Well I blame much of the added brightness I hear to the arm, which is about as light and undamped as it could be made (since those are Linn's goals, which are not inconsistent with making a cheaper product either).  I think with a bit of removable hockey tape, my friend Tim's favorite suggestion, this arm could be made into a contender.

I think even that small change might make this table satisfactory to me sonically, and otherwise (though I will note in a future post how many ways I dislike the LP 12).  Still, I think it could be reasonably good sounding.  It's basically an AR copy, with high quality manufacturing.  It's universally agreed that the AR was a landmark product, and partly because it was a good sounding product, though it WAS compromised to achieve a far lower price than the Linn.

It's the religion that ruins the LP 12 (though one can also think about the cheapness…and profitability to them…which all the religion is consistent with).  But a used one could be good spare parts, more than it's worth trading for.

OTOH, there are a lot of other decks below $800 that I'd find interesting, and just above that.

Tuesday, January 21, 2014

Deciding on my next Turntable

I have a large collection of LP records, but both of my turntables have been broken for the past two years.  Though my two existing turntables are well regarded as being ultimate perfectionist good sounding turntables, they have not been reliable.  Both of my turntables have broken twice.  They each got fixed the last time they broke, though after much delay.  They are, in order of my preference, a Sony PS-X800 with Fulton mat and Michell clamp, and a Linn Sondek LP12 with Valhalla power supply and Linn Ittok arm.  I've alternated among them, finally settling on the Sony as my favorite until it broke, then using the Linn until it broke.  For reasons of historical accident, the Linn which has non-removable headshell (I originally wrote non-removable cartridge--that's what it seems like) has my Panasonic Strain Gauge cartridge, and the Sony has my Dynavector Karat 17D2 cartridge, both also highly regarded but as different as the two tables are.

My feeling about LP records is that they sometimes offer a superior sound to standard resolution digital sources, and that this is partly because of "information" (any analog storage system has in some sense infinite resolution only partly obscured by noise, and standard digital sources are somewhat limited in this same "information") and partly because records are mastered differently--and better from the standpoint of a serious listener because they are limited in how far producers can play the loudness wars.  Playing LP's can also be a somewhat relaxing experience (or it can be tense…) akin to an I Ching divination.  In my opinion it is worth having an LP playback system if you have LP records, both for regular use and for copying LP records to digital formats for use in my music server.  I have copied about 10 LP records to digital format, and in several cases I like my transcriptions better than commercial CD's.  Two such LP's are among my favorites, Hope from Klatuu and Taking Tiger Mountain By Strategy by Eno.  Though the truth is I really didn't listen to records much after I had a sizable collection of CD's in the mid 1990's.  The truth is I don't listen that much to CD's either.  I often live in silence, and FM Radio is likely my most-used source of music, but it's pretty much only for background music.  Serious listening, which I believe in but rarely do, is best with LP's, CD's, or advanced resolution digital sources.

For quite awhile my plan had been to buy a nice but not super high end Rega turntable.  My friend T thinks they are pretty good, and has defended some of Rega's design choices, such as the small inner platter driven by a small belt.  In fact, this goes way back, he almost walked me through buying a Rega turntable in the early 1980's.  I did end up getting a Dynavector Karat Ruby at his advice, but then I broke it before using it.  I only got the Dynavector Karat 17DII as a gift from a friend, the late Professor Bill Banks, in the late 1980's.

I looked at the Rega models and decided that the P6 was about the level of unit I wanted.  I hate noisy motors and it has a 24 pole motor with custom tuned power supply in a separate box.  The usually light glass platter in the less expensive Rega's is enhanced with an additional ring of glass to bring up the rotational inertia.  Meanwhile, the top end Rega's look nice but have even more impressive looking competition.  I had wanted to buy a similar level Rega P25 back 15 years ago when the dealer convinced me to buy a second hand Linn instead.

So the plan was to buy a P6 for my own selfish birthday present.  But instead around the time of my birthday this month I was preoccupied with my broken dvd recorder, something which I have played with a lot more than any turntable ever (especially the hard drive video recording feature of my RDR-HX900) so I bought myself two replacement video recorders instead, a Pioneer DVR LX70 said to be the best of the genre, and a Magnavox which is one of the few available new in the USA.  The LX70 was a big roll of the dice, but has turned out well so far.

So my birthday passed without a turntable self present, and I had more time to think.  I was thinking of a turntable that looks more like a statement than a P6.  Soon I had discovered the Michell Gyrodec, which now seems very much like the kind of thing I like.  Of course it's cool looking, perhaps the coolest looking turntable ever because of the spinning brass weights.  But it also fits my preferences, being designed for use with a clamp, and offering a screw down record clamp as an option (an option which I'd definitely go for).

I had never thought much about Gyrodec because it was introduced about 5 years after I left my job at an audio salon.  From that earlier period I had known about the Transcriptors Skeleton turntable, and it set a high bar for what I'd like a turntable to look like (a high bar now met by many so-called statement turntables), but to me it seemed a wrongheaded design.  The record is only supported by a few points, with the rest of the record left to vibrate in the breeze.  I knew that was bad because I experienced a similar thing with my Dual 1209 turntable, my very first.  The Dual never sounded quite right, I figured, because it only held any particular record with one or two concentric rings of plastic.  That was a good design for keeping motor rumble out of the record, but not good for establishing groove quietness for good tracking.

Well J.A. Michell had been hired to engineer and build the Transcriptors so I unfairly connected the faulty design it has with the name Michell.  But it seemed either that Michell himself had other ideas, or he learned fast, because the Gyrodec was very different, and exactly as I would have wanted it.  The Gyrodec was built for record clamping, with a platter deliberately made out of a plastic composite material to couple well to a vinyl record.   It also has other very good ideas, such as an inverted oil pumping bearing, and a 3 point suspension with tensioned springs.

So for a week after my birthday I was imagining buying myself a new Gyrodec.  I imagined getting the following setup:

Gyrodec SE
Orbe Clamp upgrade (screw down clamp)
HS never-connected power supply upgrade
Techno Arm

Then I saw a review comparing the Techno arm with an SME 309.  The 309 had far better high frequency characteristics, according to the reviewer.  No more tizz.  So then I put 309 down on my list, replacing the Techno arm.  The 309 also has removable headshells, a feature I very much like (and might even begin considering mandatory).

Well the package total would have come to about $6000 for all this, $2400 for the turntable and $2400 for the tonearm, for starters.  That's as much as I was planning to spend soon to upgrade my broken bathtub to a luxury Kohler Tea for Two bathtub.  I had been figuring the difference between plain vanilla bathtub replacement and Tea for Two bathtub to be about $5000, and I really wanted the nicer one.

Actually, even though I haven't used turntables much, I still might get more use out of a good turntable that a bathtub, which I've used even less.  But a difficulty with the bathtub is that there's only once chance to get it right.  If I put in the cheap bathtub this year, if I changed my mind and decided I wanted the bigger bathtub next year, I'd have to re-do all the construction work, which is a large part of the total cost.  So if I need to skimp this year, it might be better to skimp on the turntable then get a better turntable later.

So it was as I was thinking about all this that I decided to take another look on eBay for turntables very much like my old Sony PS-X800.  Mind you, I've been browsing mainly Sony turntables on eBay for the past two months.  But I really hadn't gone much past Sony with regards to linear tracking automatic turntables, and there aren't many Sony's listed (actual PS-X800's in working condition are extremely rare).

As fate would have it, I came across a newly listed Mitsubishi LT-30.  I quickly checked the online story about the LT-30.  It is very much praised on AudioKarma and other audiophile websites.  One reviewer said it was the best he had heard so far, and another performed a shootout with very expensive Clearaudio and other tables, and the Mitsubishi won as the best sounding table.  That second review caught my eye because the reviewer was using the very same stylus as me, a Dynavector Karat 17D3.  So that satisfied my concerns about compatibility with my cartridge.  I wasn't long after that I did the buy-it-now on the LT-30.

I hope this works out, and if it does I think this is the right table for my bedroom.  I think it's very nice to have at least a semi-automatic table for the automatic lift off feature, in which the arm lifts off the record at the end of play.  Otherwise you have to get up and take care of it, which may not be good if you are making love.

It seems like the LT-30 has an excellent design in many ways, and should be both more reliable than the Sony PS-X800 and easier to fix.  One guy has posted all the details of his LT-30 repair on Audio Karma.  The LT-30 I am buying was just serviced and had its belts replaced, and I paid SquareTrade for a 2 year warranty.

It's a nice looking turntable, and has nearly a full length tonearm with respectable rigidity (stainless steel) and mass for a low compliance moving coil cartridge.

Even my friend Tim thinks this looks good.  He believes I can add damping to the arm, the chassis, and perhaps the arm carriage to make it all less resonant.  And he has argued that the "information loss" argument that claims the stylus needs to be held firmly in position with respect to the turntable platter is all wrong.  In fact, in order to play records, the tonearm must move freely with respect to the platter.  What the tonearm should not do is transmit and reflect audible vibrations.  And there's nothing wrong with the base of the tonearm being on a horizontal carriage that moves to eliminate tracking error, so long as the carriage movement doesn't generate noise (and the LT-30's tracking assembly is said to be very quiet).  What is wrong is vibration in the tonearm itself, which quickly leads to reflections and intermodulations.  It is key that those vibrations be absorbed as quickly as possible and as close to the stylus which generates them as possible.

I think now it may have been obsession with the bogus information loss argument that led to these linear tracking turntables being less popular than they might have been.  Obviously a linear arm is in a more complex geometric system than a pivoted arm.  But there is nothing wrong with that.  And especially the pivoted arm is not "correct" in any sense at all.  The linear arm is mirroring the cutter arm much better.

Tim doesn't even have a problem with the removable headshell.  He says that's fine the rubber washer can help absorb some vibration.  I think removable headshells have disappeared from respectable arms primarily because of the bogus information loss argument.  And I also have conspiracy theory involving high end audio dealers, especially Linn dealers, who want you to come to them for everything you need.

I love removable headshells.  Just as I expected, once my strain gauge cartridge was mounted by a Linn dealer to my Linn tonearm 16 years ago, it has stayed there.  I've never wanted to mess up the cartridge installation done by a professional.  There was a two year gap after my Sony died that I could have moved the Dynavector cartridge to the Linn to play records since the strain gauge demodulators I had were not working.  But I couldn't even get myself to do that, instead I waited until I got another strain gauge demodulator working.   Cartridges are very finicky and once you have one installed properly you don't, you just don't want to mess with it.

Removeable headshells are like getting your freedom back.  Not only do they make it possible to swap cartridges around, they make it easier to install and uninstall cartridges in the first place.  Linn dealers have a special jig which allows them to install cartridges on Linn arms easier, possibly with the table upside down.  That is not the kind of thing I want to do at home, and I'm not on first name basis with a Linn dealer.  I didn't like the sound of my Linn, it hasn't been reliable, and the manual arm and non-removable cartridge are highly inconvenient.  So I'm just not the right kind of person to be a Linnie.