Lathe/Turntable speed stability: Additive or Cancelling Errors?

[microstrip]

I would like to add a comment to this thread about speed stability.

Speed stability has two components - average velocity and instantaneous variations of velocity, usually quantified as wow and flutter.

As Gary stated the first one makes impossible to carry direct analysis of the sampled signals of two successive plays of the same LP. It could be done only using very powerful software to deconvolve the effect of the average variation of speed from. But the more important aspect for turntable sound are the instantaneous variations.

In order to measure wow and flutter usually a special LP having a track recorded at 3000 or 3150 Hz is used. An instrument analyzes the frequency variation and gives a single number to quantify the weighted variation of frequencies. However as the cutter system also introduces wow and flutter, some high quality turntables have lower wow and flutter than the existing test LPs and the measurement is carried using laser techniques.

Wow and flutter, however has an weighted average and a peak value. Manufacturers give the average value, but some people say that the peak value is also meaningful. As far as I know only a few peak values are known for some direct drive turntables.

You can see some information about wow and flutter measurement at wikipedia:

http://en.wikipedia.org/wiki/Wow_and_flutter_measurement

 

[naturephoto1]

One obvious thing I forgot to mention, mainly because it is not random, is the variation due to the record itself. Mass density variations in the vinyl can again cause a bit, but the largest contributor in my experience is the hole in the middle. It is rarely perfectly centered (though is usually awfully darn close!) and if it is loose the record can shift a bit during play. The former is clearly correlated, but the latter is trickier... If the hole is large enough so the record can shift, then it can be affected by other things both random and not (platter/motor wobble, plater/mat friction, etc.)

Of course, I left off the impact the tonearm might have due to friction in its bearings, mass/momentum, counterweight, anti-skate, position on the record, and all that jazz. The cartridge could also vary a bit depending upon the track (position) and all the other factors.

It's a wicked complicated playback system.

BTW, the measurement system I used (actually, helped use) was optical, with a special thin metal platter of carefully control density and thickness, with dark "slots" to allow the optoelectronics to gauge speed. Our first trials involved a Hall effect sensor, but you need magnets and there was always a question of magnet placement and density.


Edit: I see Frank mentioned record hole centering -- should have read ahead a page!

Don,

Would the record clamp help to keep the record in position to keep it from moving even if it is not perfectly centered? Also, would any warp and vinyl undulations also have any affect on the speed consistency as the platter is spinning?

Rich

 

[JackD201]

How about we simplify things a bit? Here is what formed my opinion on speed stability. I used to DJ and I am still a bedroom DJ. I use 2 identically set up Technics SL-1210 M5Gs which share AC from the same power strip. I have never witnessed two decks keep in synch for a "mere" seven minutes. If you want to "hear" speed drift this is the best way to do it. Mild Wow and Flutter causes mild flanging in the early part and as speed changes towards the latter part of the track things just fall apart and the 4/4 beat turns into the final stretch of the Kentucky Derby.

My view is that it is always additive. In the example above I attempted to demonstrate that it is, for all intents and purposes, impossible to even get two songs pressed from the same batch to synch up from the exact same starting position with everything else down to tracking force and VTA being the same on crystal locked DD tables. If this is the case I can't see how we can do anything but accept that whatever instability of the Lathe put on the LP is all we can live with. We can't fix it but by getting as stable as we can without reference to anything else but consistent 33 1/3, 45 we add less.

 

[DonH50]

Don,

Would the record clamp help to keep the record in position to keep it from moving even if it is not perfectly centered? Also, would any warp and vinyl undulations also have any affect on the speed consistency as the platter is spinning?

Rich

Hi Rich,

The clamp will keep the record from moving, yes, but of course whether it really helps depends upon where the record ends up after clamping down on it. If the LP is not centered, the clamp will reduce any variance from shifting, but you'll still have the "fixed" steady-state wobble. Should get rid of, or at least reduce, the random element.

Warps do indeed introduce periodic speed/pitch changes, because the cartridge's speed and tracking force/deflection varies with the warp. A clamp can make warped records better or worse in my experience; sometimes a clamp makes the warps even worse -- think of a warped record that you flatten in the middle so the outer edges go up even higher. Again, warps tend to be steady-state, not random, though of course since the warps often vary along the radius their impact can change at different spots (tracks) on the record.

FWIWFM, a random wobble is less audible than a steady-state drone to me, and I'll bet also to you.

IIRC, the study I helped with so many years ago was partially funded by a company developing a clamping system (pretty sure Discwasher contributed as well).

Good questions! - Don

 

[DonH50]

Jack -- What sort of drive system does that Technics TT have? I ask because I have observed that speed-controlled motors often exhibit some sample-to-sample speed variation, often larger than what we might expect. I know absolute speed was not spec'd too tightly, though it (as anything) became a marketing ploy for a while, with +/-0.1 % accuracy being touted. Not sure how well that would hold up over time (depends upon the synchro circuit). Most of us won't notice if the absolute pitch is a few cents off, but if you set up two tables with the same LP you may notice the effective time-shift you describe after 5 or 10 minutes.

One of the other tests we used to do was use a reference standard LP with accurate test tones on it and measure the actual frequency and frequency variation over the track time. The results were often interesting...

 

[JackD201]

Direct Drive Don.

 

[DonH50]

With speed control, or a motor power supply, i.e. not line-referenced?

Hmmm... Direct Drive Don, sounds like something I could have on a business card...

 

[Bruce B]

Well you can a least get 2 analog tape machines locked to SMPTE code!

 

[microstrip]

How about we simplify things a bit? Here is what formed my opinion on speed stability. I used to DJ and I am still a bedroom DJ. I use 2 identically set up Technics SL-1210 M5Gs which share AC from the same power strip. I have never witnessed two decks keep in synch for a "mere" seven minutes. If you want to "hear" speed drift this is the best way to do it. Mild Wow and Flutter causes mild flanging in the early part and as speed changes towards the latter part of the track things just fall apart and the 4/4 beat turns into the final stretch of the Kentucky Derby.
(...) .

Jack,

Since the speed of the SL-1210 is synchronized using a PLL that compares the phase of a pickup signal coming from the platter with an internal oscillator you can easily link them to have exactly the same speed +/- 0%. As far as I remember Technics turntables used to have a magnetic strip on the inside of the edge of the platter that was read by a pickup head .

All you need is the turntable schematics, a soldering iron and some cable.

 

[JackD201]

@Don - DC motors and not line referenced Don Direct Drive See Fernando's post above for more details.

@Bruce - Oh yes you can!

@Fernando - We know that Wow and Flutter isn't solely the fault of the drive mechanism and that off center spindles, warpage and resulting differences of stylus drag (particularly with belts that are too elastic or slack) cause it too. Cogging in motors do it too. I think that we can assume that the Q-lock is dead on. The strobes on the lower left hand corners show it to be so but even if it isn't and we recalibrate two SLs to have the same exact average 33 1/3 and 45, the .01 or .1 W and F deviations (whichever it is they claim) don't necessarily synch up. Add all of these factors together and you get substantial drift over time. Add to that that pitch increases slightly the closer the stylus gets to the spindle because of the way the records are cut. With two identical records this shouldn't be a problem but one can't help but wonder just how identical they are.

I know it is an extreme example, I mean how many guys here will try to synch two TTs playing the same song aside from me right? One drunken night when I still had my HRX, Keith (nephilim) and I tried to synch it up with the SL using the SDS in realtime. It was hilarious and totally hopeless. It's just meant to illustrate that if two identical decks playing the same record can't synch up, the probability of a home deck with the lathe that cut the master synching up is, well, pretty close to nil.

So this brings us to Lee's original questions and the answers, in my mind, would be theoretically YES and YES but with a gigantic BUT

Unlike time coded material where one machine can chase the other we're talking about two separate systems with zero communication between them. The deviations are most likely to come in totally different cycles. Even with decks that should have the same cycles, a DJ has to ride the pitch control faders on an SL manually for long crossfades. This is because there is no way to set a start to a cycle, you just have to react to the deviations after they've made their presence known just as a slaved Tape machine does.

I guess what I'm saying is that we shouldn't even sweat it. We just have to accept that the lathe itself was not perfect and the imperfections will show in the recording albeit minute enough not to be even close to bothersome. We can calibrate our decks as best we can and hope it isn't doing too bad a job of being "true" by having variations greater than the lathe. If variations are less than that of the lathe I would consider it a five star effort. Will extra decimal points make us enjoy more? I don't know but probably not. It might satisfy by way of being reassuring to the stickler types. The types that might go for a $650,000 Turntable. I'm no where near that kind of a stickler. It's not like I have perfect pitch or anything so unless a viola starts sounding like a violin or vice versa I'm fine.

 

[DonH50]

My limited experience with that end of the recording business coupled with more experience with consumer products lead me to believe that the vast majority of consumer TTs' speed accuracy and stability is far worse than the lathe.

I agree we shouldn't sweat it either way, just enjoy the music.

 

[microstrip]

My limited experience with that end of the recording business coupled with more experience with consumer products lead me to believe that the vast majority of consumer TTs' speed accuracy and stability is far worse than the lathe. (...) )

Surely. The vast majority of consumer turntables is fixed speed and uses cheap synchronous motors. They are synchronized to the mains frequency, that has limited stability, and belt and platter tolerances also do not help to get the proper speed. But I was addressing turntables that want to be on the podium of WBF .

But any decent turntable having a quartz oscillator similar to the one existing in a cheap, but very accurate, Casio clock can be more accurate than a fantastic high quality cutter built forty years ago, specially considering that speed accuracy beyond a certain value is not important in sound reproduction.

BTW, the Casio clock can be more accurate than a mechanical Omega Speedmaster, but I prefer the later.

 

[tony ky ma]

Thinking of TT's speed stability, in what basic count ? turns in minute ? one turn ? 1/2 turn ? 1/4, 1/8 etc, adjustment by out side force can't be quick enough in small counts, IMO the best and easy and cheap way is a thin elastic thread as the belt in a belt drive TT, auto adjust by strain as a spring. of cause it will depend on the material and length and tension etc., no way to measure out but can detect by the feeling of sound quality change from speakers, you won't believe it till you try
tony ma

 

[FrantzM]

Hi

Interesting that my reference to the cutting lathe became a whole subject of debate. For my own edification. aren't most direct drive TT PLL and can someone tell me why Belt Drive became all the rage in high end circles? Now we even see return of idle pulleys drive which can't be great for noise, wow and flutter or speed stability but these are only objective criteria after all ...

 

[mep]

Frantz-I think the *rage* for belt drive tables in the high-end stemmed from removing the motor from the plinth and now having a degree of isolation from the motor to the rest of the table. Interesting that VPI's Classic table has the motor back inside the plinth. And yes, people are buying up the old idler wheel tables like the Garrard 301, 401, and the Thorens TD-124.

 

[Bruce B]

Well from what I've been searching on the net, the Neumann and Scully lathes are:

FR: 50 - 15k
Speed acuracy: within 0.5% at both 33 1/3 and 45. Regulation within a single revolution is accurate to within 0.25%

I don't think we have anything to worry about !!!

 

[garylkoh]

In 1984, when I was in college in the UK, we visited Linn and met Ivor Teifenbrun. He had just released Linn's first LP (Carol Kidd - actually, it was the third but there's a story behind that), and I recall distinctly that him telling all of us that he had to build a lathe to cut the lacquer because the available lathes did not have sufficient frequency response and speed accuracy. He thought that the Linn turntables at that time already were far ahead of the best available lathes. In order to improve the LP12, he had to first improve the record. Good sales guy! I gave him my week's allowance to buy that album and lived on bread and water for the rest of the week.

I still have that album, and it's still one of the best sounding LPs I own.

 

[microstrip]

Well from what I've been searching on the net, the Neumann and Scully lathes are:

FR: 50 - 15k
Speed acuracy: within 0.5% at both 33 1/3 and 45. Regulation within a single revolution is accurate to within 0.25%

I don't think we have anything to worry about !!!

Specially as the Kenwood L-07 (one the direct drive turntables that should climb the WBF podium, in company with Mep Technics SP10 ) has these specifications:

Drive System - Quartz Pll Direct Drive
Motor - Coreless and slotless DC servo motor
Platter - 13 inch diameter (330mm) aluminum alloy diecast laminated with Duralumin with stainless turntable sheet
Speeds - 33-1/3 and 45 rpm
Wow and flutter - less than 0.02% (WRMS)
Rumble - better than -94dB DIN weighted
Load fluctuation - 0% (within 120g of tracking force0
Transient Load fluctuation - less than 0.00015% (at 33-1/3 rpm, 400Hz, 20g.cm load)
Transient Load fluctuation - less than 0.00008% (at 33-1/3 rpm, 1000Hz, 20g.cm load)

 

[Bruce B]

Specially as the Kenwood L-07 (one the direct drive turntables that should climb the WBF podium, in company with Mep Technics SP10 ) has these specifications:

I'll look up the specs of my direct-drive EMT 948 and report later.

TT diameter: 33cm
Speeds: 33 1/3, 45 and 78
Wow/Flutter at 33 1/3: < 0.075%
Rumble: -70dB DIN weighted

Cart: TSD-15 w/ custom vdH tip
FR: 20 - 30k

 

[fas42]

Add to that that pitch increases slightly the closer the stylus gets to the spindle because of the way the records are cut

Jack, that's a new one on me. Would you care to explain exactly what's happening here, how it comes about?

Thanks,
Frank