advertisement


Turntable speed analysis part II

Many thanks Paul, was the file from my post 683 or 687?

The motor layout is this:

PL71_Motor.jpg
 
Thanks Paul, I must say I'm very pleased with the graphs, it really would appear that I'm at the limit of the test pressing. Now to reclaim that .4%. ;-)

Flavio, it's a Kuzma Stabi/s 12" version. But I have replaced the bearing, subplatter,platter and motor with Lp12 parts and built a PSU based on a large isolating transformer and simple phase shift capacitor network. The real improvement comes from modifications to the bearing which has increased drag.
 
Flavio, it's a Kuzma Stabi/s 12" version. But I have replaced the bearing, subplatter,platter and motor with Lp12 parts and built a PSU based on a large isolating transformer and simple phase shift capacitor network. The real improvement comes from modifications to the bearing which has increased drag.

Thanks! I'm taking note, now attention on this: On the last 5 or 10 years some motor oil additives have appeared that use 'nanolubrication'. This is not quack science but real: These additives consist of extremely small (<0.1 micron) particles which 'stick' to the metal surfaces, evening them out so the finish is smoother (even if the metal was originally mirror polished!), significantly decreasing friction.

I have seen tests of those additives firsthand and i can vouch that they do reduce metal-to-metal friction in an impressive way. As a car nerd, i wouldn't pour any of those substances in my engine so easily since motor oil has much more functions than just preventing metal-to-metal contact, but i feel that this should be an excellent additive to turntable bearings !!!

Example of nanolubricant substances:
http://www.lowerfriction.com/product-page.php?categoryID=19

One of such products is "CERATEC" by Liqui-Moly. CERATEC is made of Hexagonal-Boron Nitride (hBN). "Liqui-moly" is known as "Lubro-moly" on the US, by the way.

I shall be testing one of these days the addition of a similar additive to my turntables' bearings.
 
(My assumption is that the increased but SMOOTH drag of your more viscous lubricant is overcoming the ROUGH drag of the turntable bearing... if we could smooth the bearing surfaces even more, then we can have a similar effect without having to increase the lubricant's viscosity)
 
Flavio, you can polish your bearing all your like, it might make a tiny difference to the rotation of your deck. I wouldn't waste my time trying it.

The issue seems to be that because of the low number of motor poles, and the large variation in attraction between rotor and poles across the course of one motor rotation that the motor will 'peak' as elements pass come towards other, increasing its ability to pull the belt, and 'trough' as the motor elements pass each other. The PL71 polar graph is a perfect example of this effect.

Adding viscous damping to the bearing adds a constant, almost linear load to the bearing. The idea being that the increased load from drag swamps the variance in attractor between motor elements as it rotates. This would seem to be proved true by the results achieved.

So you could polish your bearing to a mirror finish and that might even out bearing drag, but it will do nothing to damp the motor variance and it really is the effects of this 'cogging' in the motor that ruins speed stability.

If you had a huge heavy platter and very weak motor, then maybe bearing smoothness would be important enough to worry about. In most decks it isn't.
 
Flavio, you can polish your bearing all your like, it might make a tiny difference to the rotation of your deck. I wouldn't waste my time trying it.

The issue seems to be that because of the low number of motor poles, and the large variation in attraction between rotor and poles across the course of one motor rotation that the motor will 'peak' as elements pass come towards other, increasing its ability to pull the belt, and 'trough' as the motor elements pass each other. The PL71 polar graph is a perfect example of this effect.

Adding viscous damping to the bearing adds a constant, almost linear load to the bearing. The idea being that the increased load from drag swamps the variance in attractor between motor elements as it rotates. This would seem to be proved true by the results achieved.

So you could polish your bearing to a mirror finish and that might even out bearing drag, but it will do nothing to damp the motor variance and it really is the effects of this 'cogging' in the motor that ruins speed stability.

If you had a huge heavy platter and very weak motor, then maybe bearing smoothness would be important enough to worry about. In most decks it isn't.


Understood and i think you guys are 100% right.

But still i want to try this nanolubricant thing, it has the potential to reduce the noise floor IMO.
 
Ah well, that's quite another thing Flavio ;).

When looking at the lubrication of a TT bearing, I would encourage you to consider the following aspects:

What loads are applied ion the horizontal and vertical and how do these differ? What mechanism actually applies these loads?

What the surface velocities of the moving surfaces are and what forces actually encourage oil to exist between those moving surfaces?

What element may specifically be generating noise?

I think that one of the major problems with turntable main bearings is that analogies are often drawn between them and motor car engines. In reality, there is very little in common between these bearing types - other than the fact that one surface rotates against another. Consider the environment within an engine, the velocity of rotating surface, the load applied - consider why oil is actively pressurised - what is pressuring the oil in a turntable bearing?
 
I am not sure you are still interested.

In the process of building my diy turntable I became interested in measuring its performance. Are you willing to share the software you are using?

This is the only recording i have to share right now:

http://www.mediafire.com/?xz2us79ggicm9oz

Its from a yamaha yp-d71 direct drive. Analogue productions 3150hz signal.

I forgot to keep the recordings from my diy turntable, I hope to have more soon.
 
FM demodulating the above with a 20hz bandwidth in Spectrum Lab, then plotting the FFT in rightmark I get this:

yamahawf.png
 
The issue seems to be that because of the low number of motor poles, and the large variation in attraction between rotor and poles across the course of one motor rotation that the motor will 'peak' as elements pass come towards other, increasing its ability to pull the belt, and 'trough' as the motor elements pass each other. The PL71 polar graph is a perfect example of this effect.

That is interesting, I had always wondered why multi-motor designs like the Voyd worked and that could be the answer - if you align the peaks and troughs from different motors, you get a smoother total torque. It would be interesting to be able to directly measure these torque ripples, I note from a quick internet search this data is not on any motor spec sheets.


Adding viscous damping to the bearing adds a constant, almost linear load to the bearing. The idea being that the increased load from drag swamps the variance in attractor between motor elements as it rotates. This would seem to be proved true by the results achieved.

This idea has been used on the Bauer DPS
 
That is interesting, I had always wondered why multi-motor designs like the Voyd worked and that could be the answer - if you align the peaks and troughs from different motors, you get a smoother total torque. It would be interesting to be able to directly measure these torque ripples, I note from a quick internet search this data is not on any motor spec sheets.




This idea has been used on the Bauer DPS

yeah my lovely loth x othello had a large oil capacity giving a resistance to the load.

until a linn dealer farked the bearing when putting a cartridge on....heaven only knows why they even fiddled with the bearing.

it did used to need a good 15 minutes for the pitch to stabilise though.
 
I am not sure you are still interested.
This is still an ongoing project.

In the process of building my diy turntable I became interested in measuring its performance. Are you willing to share the software you are using?
At the moment I'm not willing to support it, so it seems best to keep it in house.

I'll process your file and see if it matches your result.

Paul
 
No problem on the software. I re-read the thread and had previously missed the discusion on it. I will keep playing with gnuplot and see if i can find something that will generate enough data points for a meaningfull fft with it.
 
So would that be multi motor, or viscous damping ?...or both ?

I have a Transcriptors reference which has the latter.
 


advertisement


Back
Top