Jacques,
I have the feeling that we are coming closer in understanding what's going on.
Your time signal is in anti phase, meaning that your needle is moving vertically on a blank track without modulation.
It is as if your needle is permanently pushed out of the groove and is falling back again in a continuing process with it's resonance frequency, like a wheel of a car when the shock absorber is no longer functioning.
If you could apply some form of damping, it should be possible to reduce or even eliminate it.
Question is of course, how do you apply damping. That depends on the type of arm suspension that yo have. Sometimes a thick viscous oil could do the job.
Another parameter you can influence is the anti skating force, you could twiddle a bit with this setting to start with.
I have already shown a few days ago the horizontal resonance of my arm/element,
however the vertical resonance picture looks quite different. That may be the main cause why I do not have the "vertical" rumble that you have.
Hans
I don't, and I would love to see it.
I am aware of this:
7] Macaulay J P “Differential Rumble Filter”
Circuit Ideas, Wireless World Sept 1979, p75
Hi,
I don't know about the "ideal" rumble filter but have some opinions :
The ideal filter depends on what your fixing and the system, turntable
problems and/or say vented speaker overload below tuning frequency.
(Or excursion limits of sealed drivers being wasted by low bass).
Best you can do with a single stage is 3rd order.
For turntable issues I suspect an elliptical filter in the phono stage
is the best idea, tuned to the arm/cartridge resonant frequency.
For speaker issues your talking the line stage and Chebyshev is
as good a general idea as any I can think of, except of course
for a bespoke matched speaker enhancing filter alignment.
It should be obvious having both is a good idea, AFAICT.
rgds, sreten.
I don't know about the "ideal" rumble filter but have some opinions :
The ideal filter depends on what your fixing and the system, turntable
problems and/or say vented speaker overload below tuning frequency.
(Or excursion limits of sealed drivers being wasted by low bass).
Best you can do with a single stage is 3rd order.
For turntable issues I suspect an elliptical filter in the phono stage
is the best idea, tuned to the arm/cartridge resonant frequency.
For speaker issues your talking the line stage and Chebyshev is
as good a general idea as any I can think of, except of course
for a bespoke matched speaker enhancing filter alignment.
It should be obvious having both is a good idea, AFAICT.
rgds, sreten.
If the motor or belt is off, how do you get the output? Did you spin the platter manually?
After all your idea was not that bad, but it was not the only one.
There are two different noises being produced.
One is the (vertical) 20Hz BW noise because of the stylus moving up and down in the groove with its own resonance frequency.
The other is the high bandwidth (horizontal) friction noise from the stylus permanently rubbing against the wall.
The two noise source are resp 5) and 4) in my previous response #59.
Hans
Reminds me of the Practical Wireless "Texan" amp. A Paul Kemble web page - PW Texan amplifier and the PE Rondo.
Linsley-Hood wrote an article on "Active Filters" published in Electronics World issue of October 1991. One of the Low Pass variety used a combination of his bootstrap configuration and a paralled T section. If of any interest I can post a copy of the circuit.
To Super10018
When belt is removed, the platter doesn't spin and I don't move it while measuring the noise.
So noise in this situation is not made by the displacement of the groove under the stylus.
Stylus is acting like a microphon and collect all mecanical vibrations from the turntable. When motor is on it collects in addition the vibrations at disc level produced by the spining motor.
But, at the output of the PARIAA we also have its own noise. I know it is not perfectily quiet in the range 50 Hz - 300 Hz because of the power supply.
My PARIAA is of my own design : a FET (2SK30ATM) + tube (ECF82), passive compensation network.
Jacques
When belt is removed, the platter doesn't spin and I don't move it while measuring the noise.
So noise in this situation is not made by the displacement of the groove under the stylus.
Stylus is acting like a microphon and collect all mecanical vibrations from the turntable. When motor is on it collects in addition the vibrations at disc level produced by the spining motor.
But, at the output of the PARIAA we also have its own noise. I know it is not perfectily quiet in the range 50 Hz - 300 Hz because of the power supply.
My PARIAA is of my own design : a FET (2SK30ATM) + tube (ECF82), passive compensation network.
Jacques
Last edited:
Hi
I looked at the spectrum of my recordings :
- when belt removed motor off (green graph, post #54) : the rays are 50 Hz and harmonics produced by the power supply of the PARIAA
-when belt removed but motor on (orange graph, post#54) : this is when the 19-20 Hz +harmonics appear
- but when belt is on and disc moving under the stylus, this harmonics are lower (blue graph same post).
Jacques
I looked at the spectrum of my recordings :
- when belt removed motor off (green graph, post #54) : the rays are 50 Hz and harmonics produced by the power supply of the PARIAA
-when belt removed but motor on (orange graph, post#54) : this is when the 19-20 Hz +harmonics appear
- but when belt is on and disc moving under the stylus, this harmonics are lower (blue graph same post).
Jacques
That the motor noise is lower with the belt on, can only mean that the belt driving of the platter dampens the motor vibrations.
Would it still be possible for you to make the same blue frequency recording but with the antiskating in different positions?
Hans
I think it would be a mistake to assume that LP friction has the effect of white noise being filtered by the arm-cartridge resonance. This is because the level of the noise depends in some way on the downforce, which will vary as the arm bobs up and down. We have a form of non-linear oscillator/filter.Hans Polak said:
Is there a mechanical engineer in the house? We need someone who knows about friction and possible stiction too, and nonlinear oscillation.
Yes I can do it, but now it is Easter time 🙂 and I will not be availlable before next Monday to do it.
I will do it but I am not sure we can see a difference because the accuracy of my trials is not very high.
Have a nice week end
Jacques
You are referring to an earlier response that has been surpassed in the meanwhile.
Looking a the frequency plots that Hyperman has published, one can draw a few conclusions.
With the stylus on the disk, but no motor running, represented by the green line in the graph, you can see the base noise produced by the combination element/RIAA preamp.
The part from 100Hz to 1Khz nicely follows the RIAA curve, and frequencies below 100Hz are a bit higher probably because aucoustic/mechanic coupling of noises in the surrounding.
Since the green curve is folliwing the RIAA curve, this noise can be seen as white noise at the input of the preamp.
When playing a unmodulated record groove the whole noise level goes up some 15 dB, parallel to the green curve. So the assupmtion that this is also white noise coming from the stylus is not that far off.
But the area from 5Hz to 50Hz is way to high, and must be caused by the vertical oscillation of the element/arm.
That is why 2 noise sources were needed to simulate the blue graph.
So in effect, not because of any assumption, but coming from a real measurement, it seems that friction is causing white noise.
Hans
My point is that the effect of friction may be more complicated than merely producing white noise which is subsequently filtered. It may be that the system reduces to filtered white noise for small vertical displacements (i.e. near ideal conditions) but has nonlinear issues for larger movement. The friction is acting almost at right angles to the movement it is producing so there seems to be scope for mechanical complications.
Am I right in thinking that you say that your system doesn't need the extra low frequency noise source in order to model the output?
Am I right in thinking that you say that your system doesn't need the extra low frequency noise source in order to model the output?
Yes you are right, but this is based on a time recording of my system that looks like the 3rd figure in posting #59 and not at like the 2nd one.
This 3rd figure was the result when simulating without the low frequency noise source.
I hope to come next week with real life recordings presented in the frequency domain showing how my system is performing.
I'm also anxious too to see what the exact outcome will be.
Hans
Yes, I think there is no doubt at all about that.
I would not expect a rumble filter to do anything for motor noise, for it is typically in the audio band, as seen in this thread.
BTW, apologies for abandoning this thread for a few days- have been on holiday.
Back now.
So if the rumble filter is matched in some way, how does it vary?
No, you can certainly go to 4th-order because I've done it. (see my crossover book, Chapter 8) Whether it's a good idea in terms of component sensitivity is another matter.
Can you show us an example? I would have thought that combining a phono stage with an elliptical filter would be very difficult indeed. What order elliptical filter are you thinking of?
[/QUOTE]
I think it is generally accepted that the major problem is the cart-arm resonance being excited by disc warps. Groove roughness (while real) is orders of magnitude lower.
I think if phase was an issue at these frequencies we would not be seeing mono subwoofers. I want to aim for the ideal amplitude response, and worry about the phase later, if at all.
Why ever not? Surely the whole point about about vinyl is that it HASN'T changed over the last 40 years. (and it certainly hasn't improved)
No, it is pretty much standard practice for recording.
About 20 dB below the audio with bad cart-arm combinations. I am concerned only with filtering out subsonic rubbish. Motor noise is all too often in the audio band (eg 50 Hz) and cannot be filtered. That's a matter for the TT designer.
- Status
- Not open for further replies.