Need to know how to use a spectrum analzser to measure turntable wow & flutter

[Hgspine]

Hello,

I would like to learn how to measure phonographic turntable speed stability (wow, flutter, drift) using a spectrum analyzer.  The signal source would be from a test record with a 3150 Hz tone.

Does anyone have experience with this?

Thank you so much.

Howard

[David Hess]

I am not convinced that a spectrum analyzer is the best choice to measure that but any low frequency spectrum analyzer with a relatively long record length should be able to show it directly.

[DaJMasta]

Yeah, dynamic signal analyzer is the audio to ultrasonic band equivalent, and would do the job.  As would a proper audio analyzer, or any kind of sound card.  Almost any crystal based time source (sample clock for a sound card ADC) is going to be a few orders of magnitude more stable than a turntable, so a regular internal/external sound card and a program to show you a spectrogram will do the trick (I'm a fan of the free Spectrum Lab).

[Renate]

The question is, what are you looking for?
Absolute numbers that you can quote to somebody?
Some sort of relative comparison between two pieces of equipment?
Something to troubleshoot/improve a piece of equipment?

Me, I'd start with the basics.
I've got an AD9850 DDS oscillator.
I'd set it at 3150 Hz, put it in one channel of a scope, sync on it, maybe 10 cycles in a sweep.
I'd feed the turntable in the other channel.
I'd look at it and see the magnitude of the problem.

[Mechatrommer]

I would like to learn how to measure phonographic turntable speed stability (wow, flutter, drift) using a spectrum analyzer.  The signal source would be from a test record with a 3150 Hz tone.

no experience. but isnt it any drift you see in SA will reflect drift of the turntable? say you are reading 3KHz on SA, that will be 95.2% from supposedly 33.3 rpm speed (31.7 rpm) or 4.76% drift? vice versa. but you must have good and calibrated SA to begin with and the audio type low frequency SA and low RBW? i think PC sound card + SA SW can do the trick. no?

https://hi-fiworld.co.uk/index.php/vinyl-lp/70-tests/105-turntable-tests.html

ps/edit: i think using Rigol DS1000Z scope, cranked to appropriate low sampling rate and record length sent to PC, FFT processed with VisaDSO Capture aimed for 1Hz RBW or lower (<0.03% drift resolution @ 3.15KHz) should also do the trick? but i never tried. who knows?

[Hgspine]

I service a particular type of direct turntable that has a trim pot to minimize wow and flutter.  The trim pot adjusts gain to one of two motor coils.  I am looking for the most reliable way to measure wow and flutter so I can optimize this adjustment when I do the service.

[Alex Nikitin]

Hello,

I would like to learn how to measure phonographic turntable speed stability (wow, flutter, drift) using a spectrum analyzer.  The signal source would be from a test record with a 3150 Hz tone.

Does anyone have experience with this?

Thank you so much.

Howard

You need a proper W&F meter to do that, or use the free WFGUI program by Alex Freed (available to download on my website), it is as accurate as a hardware W&F meter and has some nice options, including the data and log file outputs. You need a computer with a sound card capable of 44.1kHz 16bit input sampling. The program works on Windows, however several Mac users reported that it works just fine under Wine.

Cheers

Alex

[drknob]

Find someone with an Audio Precision P1P. I have one and it does W&F. However, you need a low-flutter test disc with a steady tone to measure.

[David Hess]

If I had to hack something together, I would FM demodulate the 3150 Hz tone and measure the result on an oscilloscope.

[Alex Nikitin]

If I had to hack something together, I would FM demodulate the 3150 Hz tone and measure the result on an oscilloscope.

There is a free and accurate tool to do this measurement, why re-invent it, especially in hardware? It also can output a file with the demodulated signal samples, which can be analysed by a software FFT program if you wish to look at the details.

http://www.tapeheads.net/showthread.php?t=55581

Cheers

Alex

[Renate]

There is a free and accurate tool to do this measurement...

That's all fine, but you notice that the posted graphs are for 30 minutes.
There's nothing wrong with that, but in the time we've all been discussing this, somebody could have knocked off the simple scope check that I mentioned.
(i.e. reference 3150 Hz into one channel, turntable into the other channel and look at how the signals slew by each other.)

1 rotation / 33-1/3 RPM = 1.8 seconds
1.8 seconds * 3150 Hz = 5670 cycles
5670 cycles * .1% wow = 5.67 cycles

So with a low figure of wow you should be able to see a gentle slewing.
If you find that things are insanely clean, you could run the test as an XY Lissajous.

I personally would like to see such a scope video.

[Alex Nikitin]

30min is just the length of the test cassette.  The WFGUI has a multi-way display, showing the current waveform of demodulated signal, the current measured frequency, and both digital and an imitated needle meter value for W&F, plus maximum values in last 10sec, both q-peak and RMS.

Here is the manual available on the web for the previous revision of the program (the current version 8 adds the logging capabilities, otherwise it is the same).

Cheers

Alex