Random 33.3 knowledge

[FreaK]

So I was nerding out the other night about high precision actuators and stuff, as one does.
I discovered that the current state of the art for precision movement devices is in piezoelectric actuators. Really neat little things. So in the cold light of the next morning I decide to look into some specs for these things, like what if they're the endgame for TT drives right? So the short answer is maybe not as they are noisy. It's like a lot of little ceramic feet tapping on your drive wheel. But maybe with some experimentation that can get worked out.
But here's the interesting part:
The companies who make this super high precision stuff give specs in terms of micro radians. Absurd levels of resolution in positional accuracy here. But a funny thing happened when I decided to discover how many radians are interesting to us when talking about turntables. Nominally speaking at "33.3" RPM we're talking about nominally 3.5 radians per second.
But that's only nominal, if you wanted a good level of turntable performance you're probably going to want to add a 3 to your RPM figure right? There's a really surprising difference in the significant figures in terms of radians per second when you keep adding 3s to your RPM.
33.3rpm = 3.487168 rad/s
33.33rpm = 3.490309 rad/s
33.333rpm = 3.490624 rad/s

try it here:

Revolutions per Minute (RPM) Converter - Inch Calculator

Convert revolutions per minute (RPM) to another unit of frequency such as hertz, kilohertz, or megahertz, and see the conversion formulas.

www.inchcalculator.com

I also discovered that our weird "33.3"rpm number appears to come from a nominal 200 Degrees per second, a case where going from 33.3 to 33.333333333333333 gets you ever closer and closer to exactly 200deg/s, from 199.8 to 199.99999999999998

I think the takeaway here is that the way you do your math in your motor controller does have a direct impact on how your TT will end up sounding. Divide any of those numbers by say 25khz, and you can see that more significant digits does lead directly to more accurate HF reproduction.

 

[kirk57]

I wonder how accurate is the turntable used when cutting records?

 

[egellings]

I wonder how accurate is the turntable used when cutting records?

See if there are sites for cutters such as the Scully models. They may list specs for the machines.

 

[FreaK]

That is the follow on question. But also, what is their target, deg/s, rpm? How does the governing system measure whether those are met?
If we take 200deg/s as the gold standard you need a fair bit of precision if you're measuring or correcting that in rad/s as it's 3.4906585 rad/s which also doesn't match any of the 33.3~rpm to rad/s conversions.

Curious what the BIG direct drive designs do, as they don't afaik, use analog correction, the big OMA and the Wilson Benesch, come to mind.
In practice this is kinda like splitting the hairs between Baerwald, Stevenson and Löfgren. but it is interesting. I genuinely thought this would be a lot less so and only require like one or two significant digits to get everything aligned.

Just to muddy the waters a bit more, there's information in the setup of the piezo actuators that says their control signals are on the order of ~80khz
And, 200deg/s is 12000deg/m, comes out to 209.43950999999998 rad/m, which finally does bring our rad measurement down to some fewer significant digits, because we can cut that down to 209.43951 rad/m without significant error. Damn π is messy!

 

[flavio81]

I wonder how accurate is the turntable used when cutting records?

Many lathes used synchronous motors. This means that the actual speed will depend on the line frequency at that moment of the day.

For example most Neumann lathes used synchronous motors by Lyrec. As far as I know, only the latest Neumann lathe, the VMS 82, used a Technics Direct Drive.

VMS 82 are rare and the pride and joy of any cutting studio.

 

[RVT1K]

The company I retired form made electron-beam lithography tools.
We had a special application that required movement in the Z axis -a.k.a. up and down- on top of the typical X/Y motion.
The stage was what provided the precision X/Y motion and piezo actuators were used to drive two wedges to create motion in the Z axis.

 

[Sleiven]

Am I missing the point here?
If you state the revolutions as 33 1/3 or in Radians/sec it is just a matter of units and does nothing to your turntables performance.

33 1/3 is 33.33333333333333 with a infinite numbers of 3 behind .

People call it 33.3 out if lazyness, the correct term is 33 1/3 rpm ,that is the definition. Which turns out to be 360/( 60/33 1/3)=200 degrees per sec

If you want your rpm in radians use the Phyphox app. But it will not make your turntable run better….

If your variation is less than 0.03rpm you have a good speed stability . There is no point in worrying about 0.003 rpm. You cannot buy a turntables with such accuracy, and even cutting lathes cannot be that accurate.

 

[KentTeffeteller]

I wonder how accurate is the turntable used when cutting records?

As accurate as a well maintained EMT broadcast turntable, or the Technics SP-10 Mk II turntables are. (A note, the latter is what many Neumann SX-68 or later model lathes have been upgraded to, from the original Lyrec Direct Drive motors those Neumann lathes were factory fitted with. That Technics motor by itself is called the SP-02 motor).

 

[kirk57]

Divide any of those numbers by say 25khz, and you can see that more significant digits does lead directly to more accurate HF reproduction.

How does a table's speed affect high frequency reproduction?

 

[Sleiven]

Playback Frequency=actual rpm/33.33*recorded frequency

All frequencies are shifted by the same amount, so high frequencies are are not reproduced different from others. Higher RPM just shifts the pitch ov every tone

So @FreaK statements makes no sense to me, I do not understand his reasoning. Maybe something is lost in translation?

 

[audiotemp]

 

[FreaK]

How does a table's speed affect high frequency reproduction?

I'm thinking of it in terms of "sampling error" so to speak. A single 25k waveform exists inside of one degree of rotation. You're right, in steady state, it's just a slight pitch shift.
Like, obviously we're talking about, for our purposes infinitesimally small differences, but it does remain that numerical representaion when converting between units does create a small error bar.

It's just sort of a thought experiment about whether one did have a motor that was capable of that level of speed resolution, and you're using a digital controller/encoder, what's the approximation of "exactly 200º/sec" you're going to use?


I did know there were motors other than the Lyrec available for Neumans, but I didn't actually know about that Technics option.
Always learning.

 

[Sleiven]

You have to explain better what you mean. It does not seem like your statements or terminology is related to vinyl playback