The ultimate rumble filter - far more effective than just a high pass filter!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Well I know from Nikki Edge (Gordon's widow) that they asked for her permission to use Gordon's signature, and kept her informed with the design process. And they invited her to the product launch.

And it is nothing to do with the term Edge meaning it has got the Edge, it was named for Gordon Edge (RIP - I miss the guy yet).

Whatever your feeling about the product, most consumer technology products are made in China nowadays. It is not a negative badge. Buy an Japanese camera? Made in China. TV? China. Quad valved power amp - China.

But Technics was the audio brand of Panasonic. Since the Technics brand was discontinued in 2002 (recently revived for high end turntables etc) you amp was probably made in Japan.

We're still way off topic. And is it mostly my fault.

Let's get back to rumble filters.
 
Account Closed
Joined 2018
Well I know from Nikki Edge (Gordon's widow) that they asked for her permission to use Gordon's signature, and kept her informed with the design process. And they invited her to the product launch.

And it is nothing to do with the term Edge meaning it has got the Edge, it was named for Gordon Edge (RIP - I miss the guy yet).

Whatever your feeling about the product, most consumer technology products are made in China nowadays. It is not a negative badge. Buy an Japanese camera? Made in China. TV? China. Quad valved power amp - China.

But Technics was the audio brand of Panasonic. Since the Technics brand was discontinued in 2002 (recently revived for high end turntables etc) you amp was probably made in Japan.

We're still way off topic. And is it mostly my fault.

Let's get back to rumble filters.


But you're missing my point, about the use of the term Edge.
I realize that it's "supposed" to refer to a person, and it's clearly mentioned in the advertising.
But honestly, step back, and view the human aspect of using specific keywords to entice buyers.
Words like... Super, Ultimate, Legendary, Reference, Studio Monitor, Classic,.... etc etc.
All implimented to sway a potential buyer in some way.
Conveniently, the design team naturally used "Edge's" name for those products to the same effect.
However, it's been used before for things, like for instance, "Razor's Edge", specifying a precision sort of design - listen, this is all about marketing, mind controlling, the usual drivvel.


And yes, my Technics was indeed manufactured in Japan, before the massive slide into chinese territory that sucked in almost all companies.
I'm well aware of it all, being a semi retired audio/video sales/service technician for 45 years.
And I've seen the result of the quality first hand....
It's not pretty......
RCA Victor, a once proud corporation with it's slogan "RCA, The most Trusted name in Sound"
And Zenith..... "The Quality Goes In, Before The Name Goes on".


Now, these, and many others, just another chinese company cranking out common "toss away" garbage.
 
Thanks Marcel, I'll start more comprehensive testing tomorrow.

At the moment, the "rumble" signal is at the same level as L & R.

I'm not sure what the power and frequency distribution should be for the rumble. Record warp is usually one excursion up and one back down per revolution so that the frequency would be about 0.5Hz. I do have a couple of LPs that were badly stored where you can see a couple of shimmies per rev.

So the questions are: what kind of vibrations can excite cartridge compliance / arm inertia resonances. I have a Jelco arm to which you can add silicone oil as a damping agent. SME sold a damping attachment that could be fitted to the 3009 Series II (I have the arm but not the damping gizmo).

For mildly warped records I use a locking record clamp that actually flattens the record slightly. The record mat supports the record at the outer edge and the outer edge is slightly higher than the center. The amount of downward dishing is very slight.

I'll actually play some records through my phono stage and into the derumbler to see what happens in real life.

Any further suggestions and discussion would be most welcome.
 
Phono Derumbler Performance Triangle Wave Testing

Hi Marcel,
I pasted a couple of screen shots into the accompanying Open Office document (.odt). I ditched the PDF conversion because it looked horrible and then they had the cheek to ask for money!
The output waveform doesn't look like a triangle (the input obviously IS a triangle).
The interesting thing is that the magnitudes of the fundamental and harmonics are the same (fundamental, 1/9 of the 3rd harmonic, 1/25 of the 5th harmonic, etc).
In other words, the frequency dependent phase shifts have converted the triangle into something that doesn't look like a triangle but has the correct spectral content.
The rumble was 1V sine at 10Hz and the "signal" was 0.9V triangle at 500Hz.
Having gotten the basic idea of how to do it, tomorrow I'll try a bunch of different frequencies.
Incidentally, my arb can create some band limited "noise-like" signals that might be a good challenge.
 
Phono Derumbler Performance

Please see the attachment for the previous post.

The input was a 0.9V triangle. The "hum" was a sine wave fixed at 10Hz with an amplitude of 1V.

Notice that the spectra all follow the same pattern (odd harmonics that decrease as 1/harmonic number squared. You can see the 10Hz sine quite plainly in the Derumbler OFF column up to about 100Hz.

The LR filters were tuned to 150Hz. The triangle shape is quite distorted in the 150Hz and 500Hz tests. Beyond that, the output looks more or less triangular.

The "distortion" isn't harmonic distortion but is phase distortion.

The triangle wave is good test signal for linearity and phase shift testing. I didn't notice anything peculiar with sine wave testing (in other words, sine testing is not much use except for frequency response).

My arbitrary waveform generator can produce a band-limited "rumble-like" waveform so I'll try that out next.

My feeling at the moment is that the derumbler does quite a good job at eliminating vertical rumble.

The filters do, quite obviously, introduce waveform phase shifts. However, these should be on a par with what an active LR crossover does, and probably better than a "normal" passive speaker crossover.
 
Please see the attachment for the previous post.

The input was a 0.9V triangle. The "hum" was a sine wave fixed at 10Hz with an amplitude of 1V.

Notice that the spectra all follow the same pattern (odd harmonics that decrease as 1/harmonic number squared. You can see the 10Hz sine quite plainly in the Derumbler OFF column up to about 100Hz.

The LR filters were tuned to 150Hz. The triangle shape is quite distorted in the 150Hz and 500Hz tests. Beyond that, the output looks more or less triangular.

The "distortion" isn't harmonic distortion but is phase distortion.

The triangle wave is good test signal for linearity and phase shift testing. I didn't notice anything peculiar with sine wave testing (in other words, sine testing is not much use except for frequency response).

My arbitrary waveform generator can produce a band-limited "rumble-like" waveform so I'll try that out next.

My feeling at the moment is that the derumbler does quite a good job at eliminating vertical rumble.

The filters do, quite obviously, introduce waveform phase shifts. However, these should be on a par with what an active LR crossover does, and probably better than a "normal" passive speaker crossover.

Hi there, very interesting but how can you create a "rumble situation" with the signal generator ?
 
Phono Derumbler Performance

This is some additional triangle wave testing but with a multitone signal for the rumble. The bandwidth of the multitone signal is 20 Hz and it is at a level of 1V.

There are four screenshots:
1) 20Hz triangle with Derumbler OFF
2) 20Hz triangle with Derumbler ON
3) 150Hz triangle with Derumbler OFF
4) 150Hz triangle with Derumbler ON

The top trace (RED) is the frequency spectrum. The jagged bit to the left of the triangle wave fundamental at 20 Hz) is the spectrum of the multitone "rumble"

The yellow trace is the triangle input

The pinkish trace is the multitone signal (it repeats about every 500ms)

The blue trace is the output from the derumbler. With derumbler OFF it looks quite scrambled. With the derumbler ON, it looks regular with no baseline shift.

The green is the input to the derumbler

At 20Hz, the multitone signal is very close to the traingle wave.

At 150Hz, the effect of the rumble signal is much obvious in the shifting baseline of the output when the derumbler is OFF. When the derumbler is ON, the baseline shift disappears but the phase shifting of the different harmonics of the triangle wave distorts the waveform.

Any suggestions for further testing would be very welcome.
 

Attachments

  • 20Hz Triangle + Multitone Rumble Derumbler OFF.pdf
    130 KB · Views: 42
  • 20Hz Triangle + Multitone Rumble Derumbler ON.pdf
    129.3 KB · Views: 35
  • 150H Triangle + Multitone Rumble Derumbler OFF.pdf
    141.7 KB · Views: 33
  • 150H Triangle + Multitone Rumble Derumbler ON.pdf
    140.5 KB · Views: 37
Simulating a "Rumble" Signal with an Arbitrary Waveform Generator

Hi there, very interesting but how can you create a "rumble situation" with the signal generator ?

Hi Kissabout,

I have a 2-output channel direct digital synthesis arbitrary waveform generator (it also has an input channel for making measurements but I haven't got around to trying that feature out yet).

While I was playing with the output waveforms, I noticed that one of the signals produced quite a good facsimile of what I imagine a rumble signal would look like. By adjusting the "frequency" you can change the frequency range of the signal. The spectrum looks like it went through a "comb" filter but the time domain output looks quite random. I selected a bandwidth of 20Hz. This seems reasonable sine nobody in their right mind would try playing music on a buzzing turntable. In other words, this represents a worst case test probably by a factor of 100!

The derumbler does a good job of filtering out the rumble since the output baseline is quite steady.

What may be matter of concern is the phase-shifting of the waveform. I use 4th-order LR active filters in my triamped speakers and haven't noticed anything odd in the sound. What I do notice with speaker development is how easy it is to adjust the crossover frequencies knowing that the 4th-order alignment doesn't have any output notches, dips, or bumps.
 
Hi Kissabout,

I have a 2-output channel direct digital synthesis arbitrary waveform generator (it also has an input channel for making measurements but I haven't got around to trying that feature out yet).

While I was playing with the output waveforms, I noticed that one of the signals produced quite a good facsimile of what I imagine a rumble signal would look like. By adjusting the "frequency" you can change the frequency range of the signal. The spectrum looks like it went through a "comb" filter but the time domain output looks quite random. I selected a bandwidth of 20Hz. This seems reasonable sine nobody in their right mind would try playing music on a buzzing turntable. In other words, this represents a worst case test probably by a factor of 100!

The derumbler does a good job of filtering out the rumble since the output baseline is quite steady.

What may be matter of concern is the phase-shifting of the waveform. I use 4th-order LR active filters in my triamped speakers and haven't noticed anything odd in the sound. What I do notice with speaker development is how easy it is to adjust the crossover frequencies knowing that the 4th-order alignment doesn't have any output notches, dips, or bumps.

so you connect together the two output of the generator with different wave form that combined produce a rumble emulation, is that right?
 
Member
Joined 2014
Paid Member
I'm not sure what the power and frequency distribution should be for the rumble. Record warp is usually one excursion up and one back down per revolution so that the frequency would be about 0.5Hz. I do have a couple of LPs that were badly stored where you can see a couple of shimmies per rev.

So the questions are: what kind of vibrations can excite cartridge compliance / arm inertia resonances. I have a Jelco arm to which you can add silicone oil as a damping agent. SME sold a damping attachment that could be fitted to the 3009 Series II (I have the arm but not the damping gizmo).


This is the big question. I've not been following this thread too closely recently as I already have a fairly complex eq to build, but I am intrigued as to the gains to be had. It is worth considering that a stylus in the groove is constantly flirting with a what in motorbike terms would be called a tank slapper. As most recordings start a blend to mono quite high up (150Hz or more) there is the argument that much of the vertical output below that is unwanted and there might be benefits in stereo image stability.



Where my mind is not clear yet is, if we have this blend to mono in the replay chain how much additional filtering might be needed to deal with the 0.5-20Hz perturbations. As the effect of these in modulating the wanted signal cannot be filtered its just removing energy low down, which whilst useful in stopping woofer flap may not improve things audibly. Not sure on that.
 
An interesting observation about vertical rumble is that it is actually added deliberately by most cutting chains!

Specifically the system that drives the groove spacing also dynamically adjusts depth of cut at sub audio rates (Because a deeper groove is a wider groove due to the 45 degree geometry) and a groove wider then it needs to be hurts playing time by forcing a lower groove pitch.

The later generations of lathes were actually getting quite sophisticated about this stuff with digital delays to give the control system information about where the groove was one turn earlier among other tricks. See for example the VMS80 (The 'computer' is a fine example of what is possible with an infinite quantity of 4000 series cmos).

There are after market systems for many other lathes as well, so this ended up not just being a Neumann thing.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.