advertisement


Wow and flutter 'meter', etc

davidsrsb said:
Goes against the ultra-low torque motor theory then.
Click to expand...

If you are forced to use a motor design that is physically noisy when delivering torque then a low torque version may reduce noise,. Better to use a capable motor with no physical noise and a psu that drives it correctly.

No reason why a tiny motor, huge platter and modest psu won't achieve the same end result, I'm just not a fan of massive platters vis a vis bearing noise and lifespan. I even use a Teflon thrust pad in my lp12 bearing.
 
I dug out an old recording of, I think, my LP12 playing a W&F track that included the 'silence' before the storm. I also created a perfect equivalent in Audacity. The raw output of my demod tool across the transition looks like this,

HittingTheTone.png


x axis is time in samples (at 44100Hz), y axis is frequency. This doesn't come out at 3150 because the calculation I'm using for 'rate of change' has a frequency response that isn't being corrected for. There is no filtering. The little wobble at the top of the 'p3150' jump is reflecting the improbable transition from flat to tone. 500 on the x axis is about 0.01s.

Anyway I see no obvious reaction to the addition of the drag from the tone over the base groove drag.

There's definitely some mileage in resurrecting this. I'll start with sq's file from YNWOAN's TT.

Paul
 
FWIW Nick G. has now done a GGC/Linux version of the program I'm working on. This seems to work as per the RO version. It is still a 'rough draft' version which I plan to tidy up and document. I'll then make it available. But if anyone is interested in seeing and experimenting with the current version, please let me know. I added a part where it calculates a 'polar' plot and prints the values out in CSV format. Should then be readable with suitable plotting or spreadsheet programs.

Jim
 
davidsrsb said:
Goes against the ultra-low torque motor theory then.
Click to expand...

Yes, it does; but there are more ways than one to skin the turntable cat (one an unpleasant mixed metaphor :)).

As sq, I've not really tried the super low torque super heavy platter option but I have tried low torque power supplies with super low friction bearing and reasonably high inertia platter - the result was that the high bearing drag, high motor torque measued and sounded better. Ultimately the problem with using the platter as a flywheel is that it's just not spinning very quickly (33.3 rpm) - in other words, it's not a very good flywheel. You can overcome this to a point by making it heavier but then you increase bearing point load, potential side forces from imperfect platter balance and you still need the motor to produce sufficient torque to overcome the platters inertia and start it spinning (all of these can be overcome but my point is that there is no free lunch).

Like sq I now use a very heavily damped bearing that causes the motor to continuously run well up its torque curve. Whilst this works for the both of us I wouldn't say it is a golden bullet than just be transferred to any situation. Sq uses what is effectively a suspension less turntable and the suspension on mine has a very different rotational moment than something like an LP12 (for example). In addition, you need a power supply that can accommodate such a load and it may even require remachining the drive pulley or platter. The DPS turntable uses this teqnique. One other advantage of the very thick silicone oil used is that it couples the bearing shaft and housing extremely effectively.
 
YNWOAN said:
...the result was that the high bearing drag, high motor torque measued and sounded better. Ultimately the problem with using the platter as a flywheel is that it's just not spinning very quickly (33.3 rpm) - in other words, it's not a very good flywheel.
Click to expand...

Or to put it another way; those old idler classics with very fast-running high-torque motors meeting drag from an eddy-current brake and then coupled to the platter got it right after all! ;-)
 
Yes, possibly with regard to the eddy brake.... Not so much with regard to the motor coupling :).
 
YNWOAN said:
Yes, possibly with regard to the eddy brake.... Not so much with regard to the motor coupling :).
Click to expand...

Bear with me...

23620750244_33c9299953_o.jpg


Whilst this is an insanely complex and over-engineered drive system I suspect logically it actually makes a lot of sense. The motor is high-torque and very fast-running (1400 rpm IIRC). It drives, via a belt to lose noise, a step pulley with an eddy-brake within. This is stepped-down again to drive an idler and then the platter. When it all works as intended you get a smooth, high-torque yet quiet drive system. Any cogging effects of the motor are way back in the chain and at such a high motor speed to have been lost. It would be interesting to see how such a system could be implemented today with modern motors, bearings etc, though it would likely cost more than a house!
 
Jim Audiomisc said:
FWIW Nick G. has now done a GGC/Linux version of the program I'm working on. This seems to work as per the RO version. It is still a 'rough draft' version which I plan to tidy up and document. I'll then make it available. But if anyone is interested in seeing and experimenting with the current version, please let me know. I added a part where it calculates a 'polar' plot and prints the values out in CSV format. Should then be readable with suitable plotting or spreadsheet programs.

Jim
Click to expand...
Jim,

I'd be interested to take a look. I can probably do a Windows build for the rest of us...

Paul
 
FWIW I think 'high torque' is a confusing term. The actual torque is the same for all turntables, the sum of transmission loss and stylus drag.

The brake is needed on the Thorens and Garrard to set the speed, I don't know if it intrinsically increases 'smoothness' like the addition of some drag to a system with a synchronous motor.

I tried an eddy current brake on the LP12, essentially some neodymium magnets mounted on a piece of folded card and attached to the armboard in the close proximity of the platter rim. This gives the theoretical double whammy, the drag smooths the drive and moves the belt further into asymmetry, and by pulling gently on the arm board the suspension is loaded horizontally. Which I think might be a good thing.

Mostly speculation.

Good of Jim to host this nonsense in his thread.

Paul
 
I am concerned about the smoothness of eddy current braking. You are making assumptions about the electrical uniformity of the braking disk.
 
YNWOAN said:
Yes, sorry Jim.
Click to expand...

No worries. I'm finding this quite interesting.

A decade ago I assumed almost no-one would ever really be interested again in the engineering and measurement of turntables, etc. Quite nice to find I was wrong. :)

Been reading mid-1960s issues of HFN and Walton's book on 'pickups' in recent months. That also seems to me to be something which people will become more interested in. He, and others, did a lot on things like distortion mechanisms, wear, etc, that may have slipped off people's radar during the decades, but may be of interest again now.
 
Paul R said:
Jim,

I'd be interested to take a look. I can probably do a Windows build for the rest of us...

Paul
Click to expand...

OK, I'll put some files up later today. I need to write a short description of what's in the output CSV files generated so people can then know how to make sense of the results if they want to plot them. As it stands, the program has no documentation or comments at all.

Nick's version is probably the best start-point for anyone wanting to port to Windows or Macs. (He added branches for it to compile OK with GCC/Linux.)

As explained, it is currently a very limited program. In particular, it currently can *only* read 96k 24 bit stereo wave files with a plain 44-byte header. That sufficed for me to test it with the recordings I've made. Next on my list is to add the ability to cope with other rates and 16bit. But at present anyone could use 'sox' or some other suitable program to generate a 96k/24 file from some other format if they wish.
 
Jim Audiomisc said:
As explained, it is currently a very limited program. In particular, it currently can *only* read 96k 24 bit stereo wave files with a plain 44-byte header. That sufficed for me to test it with the recordings I've made. Next on my list is to add the ability to cope with other rates and 16bit. But at present anyone could use 'sox' or some other suitable program to generate a 96k/24 file from some other format if they wish.
Click to expand...
96kbps is a bizarre choice for analysing wow and flutter, also excluding many 48k max PCM290x based ADCs like the old NADs
 
Tony that's certainly one way of skinning that cat. Id be interested to see measures for noise with needle-on none turning platter with and without the belt on, to see the noise contribution from motor only and motor brake and idler combined. I wonder which rotating member has the lowest noise. Platter bearing, stepped pulley, idler then motor I would imagine.
 
davidsrsb said:
I am concerned about the smoothness of eddy current braking. You are making assumptions about the electrical uniformity of the braking disk.
Click to expand...

Agreed, though when you think of where it lies in the drive chain it is probably insignificant. In the 124 above given the motor is rotating at something slightly below 1400 rpm and then via a belt drives the larger step-pulley that will still be rotating at over 1000 rpm, this is then stepped down again to drive the idler and finally the heavy platter. I suspect the effects of the eddy-brake will be at such a high frequency to be lost in the rest of the drive chain and/or damped by the platter mass. Conceptually it seems to work well, it's failings are in its complexity and any worn or even slightly dirty part will add noise.

I'd be interested to see what could be achieved by taking the idea of a very fast running high-torque motor driving a pulley with an eddy-brake just like the TD-124, but then using that mechanism to drive a high-mass deck such as a Verdier or Nottingham Dias via a second much longer platter periphery belt. I have a feeling that may get close to the best of both worlds.
 
davidsrsb said:
96kbps is a bizarre choice for analysing wow and flutter, also excluding many 48k max PCM290x based ADCs like the old NADs
Click to expand...

Since - so far - I've mainly been testing and developing the measurement computation, 96k/24 was convenient for me.

Firstly, I do all my LP recordings 96k/24 anyway.

Secondly, since I started off doing plain 'fringe counting' the higher the sample rate, the more accurate I can get the frequency measurements.

This is because the program counts the number of zero-crossings it can find, and counts how many samples apart the first and last example are. As a result, the accuracy of the mean frequency determined is related to the sample rate. That's what lets me get a few parts-per million *provided* the signal/crap ratio is high enough.

I've sinced added in interpolating for the sectional counts. So the sample rate should now matter less than it did. However I've not yet added the code to cope with other rates or 16bit.

As I've explained, all I've done thus far is an *experimental* version to see if it works. It seems to, so I can now do a 'better' version to allow other input rates, etc...

...and now so can others. I've just put up copies of the source code and a 'help' text file which anyone who is interested can get from

http://jcgl.orpheusweb.co.uk/temp/wandfprog.zip

:)

Anyone who can program in 'C's welcome to help themselves. Adapt, improve, use, as you fancy. Good programmers may find the source code amusing, or may weep at my rubbish programming skills, as they prefer. :)

Jim
 
sq225917 said:
Tony that's certainly one way of skinning that cat. Id be interested to see measures for noise with needle-on none turning platter with and without the belt on, to see the noise contribution from motor only and motor brake and idler combined. I wonder which rotating member has the lowest noise. Platter bearing, stepped pulley, idler then motor I would imagine.
Click to expand...

From my experience the belt and step pulley are the main sources of noise in the TD-124, primarily due to the belt as there are no good modern replacements. The motor and idler have caused me very minimal issue. See my thread in the classic room for excessive detail!
 
OK, I've now modified the "!RunImage" version so it can accept various sample rates and 16bit data. If interested, the source code is at
http://jcgl.orpheusweb.co.uk/temp/wandf_flex.zip
It's basically the same as before, but will now accept any sample rate up to and including 96k, either 16 or 24 bit. So it should now work for people using, say 48k sample rate.

Only done a few quick tests, but they look like low sample rates and 16 bit still works nicely. I generated some 44k1/16 files from my 96k/24 recordings and they gave pretty similar results. So all being well, people should find that 48k captures work satisfactorily.

I need to tidy up the code in some other ways and write some clearer documentation. I'll then update the Linux version. But so far as I can see the program should now be OK. (Not having *yet* fallen over any bugs!)

In principle the program would work above 96k. But some of the arrays used would need to be doubled in size. And I'm guessing no-one really needs it to cope with > 92k so I've left the array sizes as they are.

Once I'm done with this I'll have a look at making an open source 'fozgometer' program for people to play with cartridge alignment. 8-]

Jim
 
You must log in or register to reply here.


advertisement


Back
Top