advertisement


Turntable speed analysis part II

What has been interesting to note from Simon's experiments with viscous drag is that a platter such as that on an LP12, even when driven via the filter of a compliant belt, is not substantial enough in & of itself to smooth out fluctuations in the motor's pull if there are any and that it takes the application of drag to ameliorate the effect of the motor. The platter, it would seem, is perfectly capable of responding to the fluctations. That said, Simon's modified Kuzma doesn't have a suspension system acting alongside the belt further confusing things.

The recording I put up is from a deck that is much heavier but still belt driven. I don't know what contribution the particular record used is making or even which test record it is.
 
I dont have a deck to try it on but....

It would be easy to experiment with eddy-current damping on the LP12. The outer platter is thick mazak, so pretty conductive; all you'd need is a magnet on the s/steel top plate, say diametrically opposed to the cart around the 10-11o'clock position (a modest permanent magnet at that distance under the rotating platter will have NO effect on the cart output)

- then increase the radius, or shim-up closer to the platter to increase the damping. Changing radius would be much better controlled result given the sprung subchassis.

NB point for consideration: the damping is exerted by the top plate, not the sprung subchassis - so I think it would take the platter's rotational dynamic behaviour closer to an unsprung, belt drive deck.
 
Would it not affect the isolation afforded by the suspension if it were strong enough to affect platter rotation?
 
Not vertically. Horizontally, yes, as above. Perhaps make a bigger mess of things.

From my scant knowledge of the LP12 top plate I couldn't see quite how to fix the magnet to the subchassis usefully - i.e. the ideal damping reference - since the top plate is s/steel and ferritic if not strongly magnetic.

However - if anyone is using carbon-fibre or aluminium top plate - then a strong magnet fixed to the subchassis would exert a drag on the moving platter above...
 
pure sound said:
What has been interesting to note from Simon's experiments with viscous drag is that a platter such as that on an LP12, even when driven via the filter of a compliant belt, is not substantial enough in & of itself to smooth out fluctuations in the motor's pull if there are any and that it takes the application of drag to ameliorate the effect of the motor. The platter, it would seem, is perfectly capable of responding to the fluctations. That said, Simon's modified Kuzma doesn't have a suspension system acting alongside the belt further confusing things.
Click to expand...

Somewhere buried in the fifty odd pages of this thread is a good explanation of the dynamics provided by Werner. Think of it like this, if you were to hold the motor in a fixed position with your fingers and gently rotate the platter so that the belt stretched, when you let go the platter would rotate back and forth on the elasticity of the belt, a bit like a mass bounces on a spring at its resonant frequency. Because the bearing has such low friction, this oscillation has very little damping and in theory could continue for a long time. By adding viscous damping, you lower the Q of this oscillating system, reducing the impact of the motor vibration.

Additionally, when the motor is faced with a higher resistive load, it needs to generate more torque, produce a larger field and draw more current from the supply. Under load, the motor will still go round at the same drive frequency, but the angular relationship between the rotor and stator will change. At the moment, we are postulating that this produces a "smoother" field and less torque ripple, though I cannot find any theory to support that.


pure sound said:
The recording I put up is from a deck that is much heavier but still belt driven. I don't know what contribution the particular record used is making or even which test record it is.
Click to expand...

Which one? We are all friends here, no need to be secretive. ;-)
 
Neals draggy PL71, compared to what I think is the latest version with updated PSU. I seem to have used a denser FFT this evening. Things line up quite nicely,

PL71_drag_compare_spec.png


And the polar,

PL71_drag_compare_polar.png


I could convince myself that the polar is smoother. There appears to be no obvious downside to the heavier oil in the bearing, spiike at 13 and abit Hz notwithstanding.

Paul
 
Re Johns new sample, I thought listening to it that it was very slightly better than your previous sample, though not the level of improvement we experienced with Ynwans modded bearing or my own. I wonder if it is just suffering from too few poles or if the lack of 'grunt' from the PSU is an issue. Ynwan and myself both use a Geddon style PSU and I found that better than my quite carefully modded Hercules when used with the higher drag bearing, maybe simple is better here? Interested to find out.

Guy, that's a pretty good sample from revolution to revolution. It's not perfectly round by any stretch but the 'oddities' seem to be cyclical, maybe a scratched test tone track? Please do reveal, what is it, some tractor parts built NA deck? ;-)
 
Guy's unknown,

Demodulated spectrum, I've made this much wider than usual because there is so much activity. Obviously I can narrow the graph if there are reasons to look at specific areas.

GSS_demod.png


And the polar,

GSS_polar.png


It will be interesting to find out, but either this turntable is faulty or it's the recording.

Paul
 
It would be easy to experiment with eddy-current damping on the LP12.
Click to expand...
Indeed.

See my plot somewhere on page 53. I tried a needledrop comparison way back when but rather inconclusively.

Paul
 
Thanks Paul, it's interesting that subjectively there is a change in the presentation of the deck. I'm not sure I like it but it is different...bass seems to be more prominent with greater midrange detail but the upper registers seem more recessed...could be setup but all I've done is change the oil. I'm waiting on an even heavier oil to be delivered so it will be interesting to compare again....
 
Interesting that it traces pretty much exactly the same path on each rotation, even though there is a lot of ripple.
Click to expand...
It's a direct drive, so every repeated variation occurs on a multiple of the rotation frequency and lines up in my polar plot. The bumps are related directly to the number of motor poles. Belt drives and idlers produce messier plots that can look superficially 'better' but the various sources of error aren't locked together in the same way. I've tried plotting polars at multiples of the motor speed for a belt drive or idler speed, but so far without a convincing pretty picture.

The PL71 isn't a very sophisticated DD, and it shows. Of course whether it is audibly significant is another question altogether.

Paul
 
Paul R said:
Guy's unknown,

Demodulated spectrum, I've made this much wider than usual because there is so much activity. Obviously I can narrow the graph if there are reasons to look at specific areas.

GSS_demod.png


And the polar,

GSS_polar.png


It will be interesting to find out, but either this turntable is faulty or it's the recording.

Paul
Click to expand...

Looks like clipping to me
 
How would clipping manifest as cyclical speed instability? What you are seeing are 5 almost identical, very rapid, though small, changes in pitch.
 
the deck I uploaded from is a Micro Seiki SX8000 which has a very substantial platter, I think an air bearing and a dc servo controlled motor with a heavyweight flywheel as far as I know. The quality of the test record is an unknown.
 
pure sound said:
What has been interesting to note from Simon's experiments with viscous drag is that a platter such as that on an LP12, even when driven via the filter of a compliant belt, is not substantial enough in & of itself to smooth out fluctuations in the motor's pull if there are any and that it takes the application of drag to ameliorate the effect of the motor. The platter, it would seem, is perfectly capable of responding to the fluctations. That said, Simon's modified Kuzma doesn't have a suspension system acting alongside the belt further confusing things.
Click to expand...

Simon has essentially implemented the same modification I originally made to my own deck - and my deck does have a suspension - similar benefits were realised.
 
John Channing said:
Somewhere buried in the fifty odd pages of this thread is a good explanation of the dynamics provided by Werner. Think of it like this, if you were to hold the motor in a fixed position with your fingers and gently rotate the platter so that the belt stretched, when you let go the platter would rotate back and forth on the elasticity of the belt, a bit like a mass bounces on a spring at its resonant frequency. Because the bearing has such low friction, this oscillation has very little damping and in theory could continue for a long time. By adding viscous damping, you lower the Q of this oscillating system, reducing the impact of the motor vibration.
Click to expand...

To confirm this is correct...on my LP12, I took off the platter and put it back on upside down so that I could access the motor. With a clean finger gently pressing on the motor, I gently turned the platter and let it go. It does indeed oscillate at low frequency (2-3hz i would guess) and made 5 to 6 oscillations before coming to rest. If you turn the platter too far, the belt slips, so this only works for small displacements. For those adding viscose liquids to their bearings, I would be interested to hear about the results from the same test.
 
So well, i'm back here again and have read about the 'spring' bounciness effect of the belt and how the bearing, when it has very low drag, would do nothing to damp it.

Then i think: How about, for a belt drive turntable, a paddle that sticks out to the side of the platter, said paddle 'rowing' over a pool of viscous fluid? Fluid damping... for the turntable itself !!

In any case it wouldn't be extremely difficult to implement if the paddle is close to the tonearm bearing shaft, and the viscous fluid well is a donut around the bearing...
 
You must log in or register to reply here.


advertisement


Back
Top