advertisement


Turntable speed analysis

Status
Not open for further replies.
Phi , logic always applies. As much as one may wish to disregard physics in Favour of audio magic and wishful thinking it will unfortunately never take you anywhere but up blind alleys.

If an idler wheel softened up enough at 78 to remove irregularities then it must by default be too soft to hold its own shape and would then be subject to the same degradation as all other rotating rubber wheels that had reach a level of plastic flow. If you don't understand the materials science that's fair enough, but when the data and the experience pool tells you that you supposition may be incorrect- guess what?
 
sq225917 said:
Phi , logic always applies. As much as one may wish to disregard physics in Favour of audio magic and wishful thinking it will unfortunately never take you anywhere but up blind alleys.

If an idler wheel softened up enough at 78 to remove irregularities then it must by default be too soft to hold its own shape and would then be subject to the same degradation as all other rotating rubber wheels that had reach a level of plastic flow. If you don't understand the materials science that's fair enough, but when the data and the experience pool tells you that you supposition may be incorrect- guess what?
Click to expand...

F1 logic and physics applies to F1, not to TT, so save us the catchy comparisons and don't push the discussion into blind alleys.
For sure there are F1 material science fora where you can practice your F1 knowledge.
Back to the topic, idler has much more to it than just it's rubber surface, there are more factors determining it's pefrormance.
Have you seen how the idler is mounted in EMT's? Have you thought about all the possible sources of idler error? Also what made you think that my idler has been running for years? It's actually pretty new and together with it's washers not really burned in.
 
Rubber is rubber, the physics does not change just the scale of measurements. It might be possible that running a deck at 78 causes the rubber idler to warm up sufficiently to drop a Shore rating point or two and decrease the rumble. I assume that would depend on the mechanism by which the idler is held in contact with the driver surface- ie sprung tension or located in a fixed position by some none-reactive adjustment option. But more than anything it would depend on slippage and drive load, neither of which are high enough to really have this effect in a EMT.

If you doubt that I suggest you get an infra red gun thermometer and measure the idler wheel temp. I doubt it will move more than 5 degrees above ambient even after protracted running.


If you are talking about the bearing smoothing in from new, then yes a very small amount of that will happen, but 'running in' is just a matter of perspective, it would be more correct to refer to it as 'wearing out' as it is a continuous process from the day you first use an item until the day it falls to bits.

There may be a period during which the friction noise generated from the sliding/rotation of new parts subdues due to polishing to such a point that it becomes less audible and at the same time the increase in reciprocally created bearing slop, from angular error, might not have risen sufficiently to allow the idler to wobble on its bearing- so you may well achieve a set of conditions that come together to give reduced noise as these two variables pass each other like ships in the night. This may even last for years.

Sadly we'd need measurement repeated over a period of years/decades to prove this.
 
Anyway, comming back to the initial subject, the experiment has just shown that at least in my context all those processes even if happen do not seem to matter. And this is the most important outcome. Thank you Paul!
 
I'm not into that for sure. Not now. After months of my solo fight the machine has been reworked, including
changing the idler at least twice, by one of
the known EMT experts and is said to meet the original EMT specs, including W&F. So it be.
My goal for now is a 3-phase power synthesizer, which should be ready "soon".
Then I will bombard Paul with data again :)
 
YNWOAN said:
Hmm...well no amount of power supply cleverness will make the idler more concentric.
Click to expand...

Idler no, motor yes. When I look at how the motor is supplied
my eyes are bleeding...

Also it's not clear to me where the idler fails: the surface, the axis alignment, the idler bearing/washers, all the things?
I do not touch it now as I said.

...having said that I do have an old EMT idler which I would happily send to someone brave enough to try to restore it's surface.
Anyone knows a service like this? I rememeber seeing one guy in US specilaized in restoring old idlers for $30 or something ike this. Maybe he can beat the official EMT severely overpriced parts?
 
Green is good, red is bad, it basically shows rumble and tracing components as well as PSU components upto 70hz.

Dan it's be good if these went up to 150hz to show the third on the line frequency.
 
What is shown is the power spectrum between ~5Hz and 70 Hz as a function of time. Reds indicate more power, greens less as sq says.

I'd say red is likely bad if it doesn't appear across all 4 tracks. It's hard to say if the very repeatable signal in the 10 Hz-ish band is on the record or the decks consistent response to record warp or eccentricity. I'm actually surprised how repeatable the response is.

Interesting that the PSU noise at 50 Hz is lower with drag applied, and does not occur until the record tone starts ~1.5 rotations in.

My recollection is that there's not much beyond 70 Hz, but I can take a look.
 


Here it is compared to the A.wav. I attempted to normalize the files (Geddon was about 50% louder) so hopefully the power can be directly compared. Certainly 50 Hz noise is weaker though the ~7 Hz stronger.
 
Normalizing should be fine we used the same copy of the same test track.

Interesting result for several reasons.

I wonder if the lower 50hz is symbolic of my phase caps being matched to within 1n for perfect 90 degree lag via my scope, whereas Mark's uses a generic best fit value from the Armageddon, or if it's down to my Prefix vs his 323 based stage. we should borrow a step-up to put this to the test. Or I could scope out our stages.

Ynwans deck has less energy below 35rpm, his deck is suspended, but also his bearing is magnetically levitated. Notice how mine has tell tales around 33rpm the rotational frequency of the platter and his has none.

I'm pleased with that as a result, I hope Paul can similarly provide the polars to check against earlier versions. Time to investigate the phonostage and noise from laptop doing the recording.

Thanks Dan.
 
You're welcome. Glad it's useful in a small way, since looking at the wavelet spectrum was mostly done out of curiosity. I wonder what looking at the demodulated signal might look like.

Since these are two different decks, I can only assume the two vertical bars per revolution (extending from 15-25 Hz) are on the record. I suppose the rest is coming from the deck/arm/cart.
 
Paul,

Can you expand on this linked from your first post:

I(n)*Q'(n) - Q(n)*I'(n)
----------------------- = omega (F)
I^2(n) + Q^2(n)

Where I(n) is the original signal, Q(n) is the Hilbert transform of I(n), and I'(n) and Q'(n) are their respective bandlimited derivatives.

Is omega(F) the demodulated spectra? How do you go back to the signal? What band pass did you use? If M is a band pass mask in frequency space (simple notch 1s and 0s?), to compute the numerator is it:

I * M(Q) - Q * f'(M(f(I)))

where f is the forward fft and f' is the backward fft.

thanks

Dan
 
In that relation 'omega (F)' is the instantaneous phase, which should be in radians per period. Which can be converted to instantaneous frequency by multiplying by the sample rate and dividing by 2Pi and hey presto, the desired output.

I optionally bandpass the input signal at the measured centre frequency, use an FIR Hilbert Transformer to generate Q, a trivial differentiator and the output seems to be about right. I wouldn't want to stage a serious first principles defence of the choices made in the detail of these processes, I wasn't paying attention in Signal Processing class 25 or more years ago, and the book we had to use was, and remains, spectactularly impenetrable. So I started again from first principles with the book by Richard Lyons also referenced in the first post.

Paul
 
Status
Not open for further replies.


advertisement


Back
Top