Vibration Management

[Mike Lavigne]

This I find a very cool idea! While I'm more with Klaus re the mental picture why this type of a "skeleton" rack works,
clamping is a very clear and well practiced engineering idea and its sort of strange nobody used it before John in audio!
Clamping both damps some chassis resonances (the driving force has to work against the constraints) and moves them higher where
they can have a chance to be better damped, i.e. by the chassis material itself. Marketing apart, I would not at all be dogmatic with either clamping
or isolation. I don't see those techniques as mutually exclusive. To the contrary, I'd rather see them combined:
Equipment clamped to an isolation platform. Roughly speaking, clamping would take care of (some of) the airborne resonances in
the chassis, while the platform of the structure-borne. Just by plain physics, there is simply no chance for a single technique to be good for all.
We realized that at Stacore and combine different techniques to work for the target effects (repeatability of the sonical results is one of them).

Cheers,

it seems to me that the idea of the benefit of clamping a chassis is directly related to build quality. most gear has quite a bit of it's own resonance, which would then make it an ideal candidate for clamping. OTOH some gear is much more robustly built and clamping would have much less effect. clamping would be akin to improving the build quality.

for instance my darTZeel preamp chassis is a solid block of milled plates, that fit together like a Swiss watch. the inside is built and finished as well as the outside. over the years I've seen it be much less affected by tweaks than other gear I've had. as I recall the Linn CD-12 was similar......which was milled from a solid billet. clamping these products is not going to make them 'more solid'.

I'm not sure one could tell by looking whether clamping would have a significant effect, but maybe you could get an idea.

this is just my own view based on observation of cause and effect, and intuition. and I could be completely wrong on a broad spectrum of gear.

 

[Stacore]

Yes Mike, you are right of course. If the designer has done the right job with the chassis design,
any additional damping techniques can be just the icing on the cake.
But from my experience wise anti-vibration chassis design is a rare event,
so as I said I do see a very good potential in additional efforts like mass loading or clamping.

Cheers,

 

[KlausR.]

While I'm more with Klaus re the mental picture why this type of a "skeleton" rack works,
clamping is a very clear and well practiced engineering idea and its sort of strange nobody used it before John in audio!

In my book the very first that should be done is to investigate if there is a problem at all that needs to be adressed. Vibrations come as airborn (sound from the loudspeakers) and structural (someone walking through the room, train/bus/tram passing by outside just under the window of your listening room).

Airborn sound: set up the gear as usual, measure and listen (with control of the listening parameters), move the gear to the neighbouring room while leaving the speakers in place, measure and listen again.

On your site you decribe how you investigate your devices:

We chose a slightly shabby laboratory table that could mimic a "bad piece of furniture", and forced the vibration through a special hammer. The hammer strike at the table aroused in him the vibration from which the platform tried to isolate the sensors and load placed on it. Such a laboratory procedure imitated the vibration of the furniture under the audio equipment through a very strong sound pulse, such as the dynamic input of a full-size orchestra.

A hammer is fine, but did you ever use the real thing, i.e. a stereo system providing that input of a full-size orchestra from a CD? If yes, at what speaker distance and sound pressure level, and what did the measurements look like? If no, why not?

The clear positive effect heard during the hearing here was confirmed by the results of the study.

What was the setup used in those listening test? The scientific literature points to the factors relating to bias in sensory evaluation tests in general and audio listening tests in particular. Did you consider these factors and implemented means to control them? If no, why not?

As for the clamping method: are there controlled listening tests that prove that clamping makes an audible difference?

Klaus

 

[PeterA]

Yes Mike, you are right of course. If the designer has done the right job with the chassis design,
any additional damping techniques can be just the icing on the cake.
But from my experience wise anti-vibration chassis design is a rare event,
so as I said I do see a very good potential in additional efforts like mass loading or clamping.

Cheers,

I may be misunderstanding Stehno, but I do not think his clamping system is primarily for damping a ringing component chassis. I think it is more about providing an optimal interface or conduit through which internally generated vibrational energy can most efficiently leave a chassis and travel away into some mass sink like a concrete floor system. He is essentially soft welding the metal chassis to a metal clamp to improve the contact for energy transmission. If you look at his rack, it appears to have clamps near the perimeter of the top plate where most chassis are already pretty rigid. The bottom plate makes contact with the rack near the center of the component where transformers are located. If damping were the only concern, that could be accomplished by simply placing weight on the top of a component. The clamps are doing that but a lot more too.

Stehno clamped the Esoteric CDP which has a very robust chassis. He reports that the Oppo sounds better with his rack system, so it seems to me that it is more about energy transmission than it is about chassis damping. But as I say, I might not fully understand his goals.

 

[Mike Lavigne]

I may be misunderstanding Stehno, but I do not think his clamping system is primarily for damping a ringing component chassis. I think it is more about providing an optimal interface or conduit through which internally generated vibrational energy can most efficiently leave a chassis and travel away into some mass sink like a concrete floor system. He is essentially soft welding the metal chassis to a metal clamp to improve the contact for energy transmission. If you look at his rack, it appears to have clamps near the perimeter of the top plate where most chassis are already pretty rigid. The bottom plate makes contact with the rack near the center of the component where transformers are located. If damping were the only concern, that could be accomplished by simply placing weight on the top of a component. The clamps are doing that but a lot more too.

Stehno clamped the Esoteric CDP which has a very robust chassis. He reports that the Oppo sounds better with his rack system, so it seems to me that it is more about energy transmission than it is about chassis damping. But as I say, I might not fully understand his goals.

reading Stehno's comments, he speaks mostly to 'pressure points in the chassis where power supplies, etc. might be' where the greatest gains to clamping are found. this speaks to build robustness where clamping is helping, where it's damping resonance at these points. whether any resonance is then carried away seems intuitively besides the point of stopping that resonance from ever starting.

as far as judging what is actually a robust chassis, and what is just 'heavy'........we would need to look closely in and out of the chassis. so this issue can be hard to pin down.

 

[Stacore]

First of all I apologize that the Measurement section of the page is still in Polish - I'll translate it as soon as I get some bit of free time.

As for your questions:

1) We used impulse hammer as the lab standard for the *controlled*, impulse excitation. The problem with the direct excitation by music is how to control the amount of excitation for the measurements to be any close to reliable. We actually tried a technique, closer to what you say: we put an accelerometer
on the table, exciting the table and comparing the readouts (acceleration - acceleration) with those on the platform. Unfortunately according to the scientist conducting this the tests, the data were not healthy enough (low correlation). We plan on more tests, this time with a linear exciter.

2) We have tested our gear in about a dozen or more systems so far, each time with different listeners. Keep in mind that those are "audiophile tests" not scientific researches but we try to keep them as clean as possible. We come, install, and shut up. We don't speak of sound prior to evaluation. We don't choose the music nor make any suggestions. I'd love to put our platforms under a rigorous scientific test one day, with the trained listening group, blind testing etc.

Cheers,

 

[KlausR.]

Jarek,

First of all I apologize that the Measurement section of the page is still in Polish - I'll translate it as soon as I get some bit of free time.

Throwing the text into Google translate works fine.

We used impulse hammer as the lab standard for the *controlled*, impulse excitation. The problem with the direct excitation by music is how to control the amount of excitation for the measurements to be any close to reliable.

Having a standardized input signal is certainly an valid aproach. However, noboby strikes his audio rack with a hammer or similar. Such standard could certainly be provided by playing a music impulse, say from a drum kit, from a CD through speakers. The manufacturer of my speakers uses a tone burst for measuring their loudspeakers in the lab. If your hammer signal is far from anything a stereo system would ever provide, then why use such a signal in the first place?

Physically striking the table with an object is not quite the same as hitting it with sound waves. The big question is: do sound waves have a physical impact on audio components, is the energy transmitted into the component, and if yes, to what degree? I can see the problem with using a loudspeaker: the sound wave itself hitting the accelerometer would probably generate an output, not the vibration of the table the component is placed on. Surely you could find a solution for that issue.

I'd love to put our platforms under a rigorous scientific test one day, with the trained listening group, blind testing etc.

There are 3 AES papers on bias in listening tests, if interested, drop me a note and I can send the pdfs.

http://www.aes.org/e-lib/browse.cfm?elib=14393
http://www.aes.org/e-lib/browse.cfm?elib=18105
http://www.aes.org/e-lib/browse.cfm?elib=14237

Klaus

 

[Stacore]

Having a standardized input signal is certainly an valid aproach. However, noboby strikes his audio rack with a hammer or similar. Such standard could certainly be provided by playing a music impulse, say from a drum kit, from a CD through speakers. The manufacturer of my speakers uses a tone burst for measuring their loudspeakers in the lab. If your hammer signal is far from anything a stereo system would ever provide, then why use such a signal in the first place? Physically striking the table with an object is not quite the same as hitting it with sound waves.

Not sure if you are German-speaking as your name suggests, but there is a brilliant German world: jein Yes and no.
Measuring loudspeakers and measuring structural vibrations are two different things but the basic ideas are the same: you have
impulse tests and frequency scans. If I'm right, burst tone = impulse test just like the signal hammer or simple knuckle test.
One of the purposes is to excite with the impulse as many resonances as possible/needed/wanted in a sense
circumventing the scan. Math speaking, impulse will have quite a lot of harmonics in its spectrum (depending on the shape of course)
and you want to shake with those frequencies your object to see how it responds. The resonant structure of the object is what it is and
(in the small excitations, elastic regime) is independent of the excitation. So if I hit the table with a knuckle, hammer of sound burst
is in a sense secondary for the measurement - these all methods excite the structure at its natural resonant modes. One cannot hit the loudspeaker
hit with knuckle or calibrated hammer so ppl use bursts.

I can see the problem with using a loudspeaker: the sound wave itself hitting the accelerometer would probably generate an output, not the vibration of the table the component is placed on. Surely you could find a solution for that issue.

You've hit the bulls eye! I've been struggling with the same problem myself and lab guys can't really help much (those I spoke to) as for them acoustical feedback is of no concern.

There are 3 AES papers on bias in listening tests, if interested, drop me a note and I can send the pdfs.

http://www.aes.org/e-lib/browse.cfm?elib=14393
http://www.aes.org/e-lib/browse.cfm?elib=18105
http://www.aes.org/e-lib/browse.cfm?elib=14237

Check PM

Thanks,

 

[KlausR.]

One cannot hit the loudspeaker hit with knuckle or calibrated hammer so ppl use bursts.

They actually use tone bursts for measuring maximum sound pressure level. I thought of such a burst as replacement for the hammer.

KlausR.: "I can see the problem with using a loudspeaker: the sound wave itself hitting the accelerometer would probably generate an output, not the vibration of the table the component is placed on. Surely you could find a solution for that issue."

You've hit the bulls eye! I've been struggling with the same problem myself and lab guys can't really help much (those I spoke to) as for them acoustical feedback is of no concern.

What about Laser Doppler vibrometer?

Klaus

 

[Stacore]

What about Laser Doppler vibrometer?

I take all that you can give me

Great idea! Have to talk to the lab we are preparing to measure in
Thanks,

 

[Folsom]

stehno's clamping does a few things... First it makes all the components seem like one. Instead of whatever vibrations in them bouncing between 4" they are going through 20" or whatever. The resonate frequencies drop as they have to pass through more mass. Then when it couples to the floor it further drops, and can even leave out the bottom (it can come up in, but there's more horizontal energy in the floor than vertical unless you have structural issues). Mass helps this. The only bad thing about the mass is they want to make a pendulum function when up high. But keep in mind the heavier a pendulum the more force it takes to move it. But in order to move the pendulum in this case, it has to fight the two other anchor points that are farther spread out. Their distance gives them an advantage or torque application so long as the arms are stiff. As the equipment tries to move, it has to fight a higher form of dampening as the amount of force required to move is increased the farther out the other points are. Think of it like how easy is it to topple a TV stand, now how easy is it to topple a 12 foot long table? Even if they weight the same it's harder to apply the local force required on the long version to make it happen. This also reduces the stands attempts to "walk". So it's pretty hard to move his weights enough to get more than an omni-vibration in them, and they will work to benefit for vertical force. As stated earlier, most of energy from the speakers is reflected, so the horizontal movement isn't as big of a factor; and the horizontal energy from speakers that's in the floor isn't that inclined to go UP into the rack system.

While the enclosures themselves can aid in all of those functions, their size simply can't accomplish as much as you'll get from multiplying them like stehno has done. It would be best to have great to have well done chassis of electronics put into his design.

To me this is basically the opposite of say CMS racks. Both are totally valid ways of achieving a goal. One burns the energy off (CMS) by dampening it, one transforms it into something pretty benign (stehno). Time for a shootout?

 

[Barry2013]

That's exactly what physics tells us and that's exactly what I said:

If you want to drain vibratory energy away you need materials with similar impedance. The first one you usually can't choose since it's what the audio component is made of. If the component has rubber feet a steel interface is not as good as a wooden interface. Placing a steel rack on a concrete floor isn't a good idea either.

Klaus

I have a lot of doubts about heavy steel racks and find their aesthetics not particularly appealing.
I have the Entreq Athena racks,the last bought very recently and posted on the Entreq Tellus thread earlier today.
I may have missed it but I have not noticed any discussion about the effects of steel racks on the electro mechanical properties of the components and the system as a whole and which I would have thought would be detrimental.
Any thoughts?

 

[Stacore]

Barry, depends on what you need. E.g. if you need a stiff rack or rather soft.
For a given geometry, steel is the stiffest (3x stiffer than aluminium) construction material.
It is also quite well damped (quite counterintuitively), not as well as brass or copper but it's
much cheaper plus can be easily additionally damped with various fillings (sand, cat litter, gravel, etc).
Not sure what you mean by the influence on the elctro-mech properties.

Cheers,

 

[Folsom]

Steel is stiffer at the same dimension. The problem is if you have 2x1" aluminium rails and switch to steel... it goes from a two man job to an electric palette jack for a big rack.

 

[miniguy]

Pretty cool little online calculator for Acoustic Impedance measurements showing the amount of reflection involved

https://www.nde-ed.org/EducationResources/CommunityCollege/Ultrasonics/js_apps/acousticImpedance/

Brief explanation of calculator function:

"The following applet can be used to calculate the acoustic impedance for any material, so long as its density (p) and acoustic velocity (V) are known. The applet also shows how a change in the impedance affects the amount of acoustic energy that is reflected and transmitted. The values of the reflected and transmitted energy are the fractional amounts of the total energy incident on the interface. Note that the fractional amount of transmitted sound energy plus the fractional amount of reflected sound energy equals one."

Be careful not to put too much faith in this very simple calculator. It holds only if the incident wave is a compression wave and normal or perpendicular to the interface between the two materials and the boundary is slip-free. It is unclear what form the vibrations and their directions in our stereo components take. The general situation is complicated because sound or vibrations can travel in solid media as both compression and shear waves, unlike gas or liquid, which cannot support shear stresses. An incident compression wave striking a boundary at an oblique angle between two solids gives rise to four waves: reflected compression and shear waves and transmitted compression and shear waves. This is necessary because the physical boundary conditions have to be satisfied for normal and tangential displacements and stresses. This was treated as early as 1899 by C.G. Knott.

 

[stehno]

stehno's clamping does a few things... First it makes all the components seem like one. Instead of whatever vibrations in them bouncing between 4" they are going through 20" or whatever. The resonate frequencies drop as they have to pass through more mass. Then when it couples to the floor it further drops, and can even leave out the bottom (it can come up in, but there's more horizontal energy in the floor than vertical unless you have structural issues). Mass helps this. The only bad thing about the mass is they want to make a pendulum function when up high. But keep in mind the heavier a pendulum the more force it takes to move it. But in order to move the pendulum in this case, it has to fight the two other anchor points that are farther spread out. Their distance gives them an advantage or torque application so long as the arms are stiff. As the equipment tries to move, it has to fight a higher form of dampening as the amount of force required to move is increased the farther out the other points are. Think of it like how easy is it to topple a TV stand, now how easy is it to topple a 12 foot long table? Even if they weight the same it's harder to apply the local force required on the long version to make it happen. This also reduces the stands attempts to "walk". So it's pretty hard to move his weights enough to get more than an omni-vibration in them, and they will work to benefit for vertical force. As stated earlier, most of energy from the speakers is reflected, so the horizontal movement isn't as big of a factor; and the horizontal energy from speakers that's in the floor isn't that inclined to go UP into the rack system.

While the enclosures themselves can aid in all of those functions, their size simply can't accomplish as much as you'll get from multiplying them like stehno has done. It would be best to have great to have well done chassis of electronics put into his design. :

Folsom, I perceive tho art highly insightful. For the most part anyway as there are a few things I couldn't agree with, not to mention a few key areas you overlooked. But excellent analysis overall.

To me this is basically the opposite of say CMS racks. Both are totally valid ways of achieving a goal. One burns the energy off (CMS) by dampening it, one transforms it into something pretty benign (stehno). Time for a shootout?

Ok, well, you were highly insightful until these last few statements anyway. Both cannot be totally valid especially since one's history has been buried in isolation methods and is just now trying to migrate away from that, not to mention the CMS rack only functions at a tweak / accessory performance level as confirmed by several here and elsewhere.

A shootout? It would be fun to say the least. But remember that it takes my rack 2 months settling in to reach about 80% of its full potential and the often times multiple daily incremental improvements don't even begin to kick in until day 5 or 6. IOW, if you wanna make it a fair shootout, you might wanna consider holding the actual event no later than day 8.

 

[Folsom]

How substantial do you rank your rack system?

 

[cjf]

Be careful not to put too much faith in this very simple calculator. It holds only if the incident wave is a compression wave and normal or perpendicular to the interface between the two materials and the boundary is slip-free. It is unclear what form the vibrations and their directions in our stereo components take. The general situation is complicated because sound or vibrations can travel in solid media as both compression and shear waves, unlike gas or liquid, which cannot support shear stresses. An incident compression wave striking a boundary at an oblique angle between two solids gives rise to four waves: reflected compression and shear waves and transmitted compression and shear waves. This is necessary because the physical boundary conditions have to be satisfied for normal and tangential displacements and stresses. This was treated as early as 1899 by C.G. Knott.

I think the calculator must be taken for what it is. A free program that can provide a simple visual and general idea of what may be taking place under the covers when interfacing/combining various materials.

I believe I've read that the interface angle between two objects should be 90deg if possible. Something like a Spike (Material #1) resting directly on top of a floor, platform or other object that could be considered Material #2 so that the directionality of the energy is more predictable.

I haven't been able to find anything online that talks about how differing shapes alter the direction of the energy be transmitted/transfered. It seems the most common shape found for audio applications is the circle/sphere or cone. I wonder if this is because these shapes radiate energy more uniformly in a 360deg pattern? Then again, if using the simple tool posted earlier as an example, when both materials are the same the reflection is almost non-existent which would lead you to believe that the shape of the materials doesn't matter. Just speculating at this point unless solid info is found to say otherwise.

 

[stehno]

I may be misunderstanding Stehno, but I do not think his clamping system is primarily for damping a ringing component chassis. I think it is more about providing an optimal interface or conduit through which internally generated vibrational energy can most efficiently leave a chassis and travel away into some mass sink like a concrete floor system. He is essentially soft welding the metal chassis to a metal clamp to improve the contact for energy transmission. If you look at his rack, it appears to have clamps near the perimeter of the top plate where most chassis are already pretty rigid. The bottom plate makes contact with the rack near the center of the component where transformers are located. If damping were the only concern, that could be accomplished by simply placing weight on the top of a component. The clamps are doing that but a lot more too.

Stehno clamped the Esoteric CDP which has a very robust chassis. He reports that the Oppo sounds better with his rack system, so it seems to me that it is more about energy transmission than it is about chassis damping. But as I say, I might not fully understand his goals.

Peter, I think I may have given you the shorthand version. The Esoteric is built somewhat like a tank with what seemed to be the 60 lbs. chassis built around and anchored to its infamous VRDS metal transport which responded relatively well to my rack and with superior construction, heft, and materials I expected as much. Whereas the OPPO 105D at maybe 16 lbs. and plastic transport did not respond near as well as the Esoteric. As I expected. However, at this stage and just using the transports of either CDP, the OPPO was a tad better so I kept the OPPO and sold the Esoteric.

Where I was suprised was that the Esoteric's performance, with its easily superior construction, materials, transport, and heft, was still a little inferior to the OPPO's less robust construction qualities includng the plastic tray transport.

Seperately and months later, it wasn't until I utilized the OPPO's passive volume attenuator and USB ports (for my SSD) that the OPPO was clearly a superior performer over the Esoteric.

Thus far, the OPPO has not responded as well as I hoped in my rack. In fact, last week I did a partial tear-down and re-install to try an improved mounting method for both the line conditioners and OPPO to see if that can be improved. This being day 6, it's too early to tell but it's sounding like things are just starting to kick in. However, since I re-installed both line conditioners and CDP, if the improvements start soaring I won't be able to discern which objects benefited most from the new method. But if it works, I won't care.

 

[DaveC]

I think the calculator must be taken for what it is. A free program that can provide a simple visual and general idea of what may be taking place under the covers when interfacing/combining various materials.

I believe I've read that the interface angle between two objects should be 90deg if possible. Something like a Spike (Material #1) resting directly on top of a floor, platform or other object that could be considered Material #2 so that the directionality of the energy is more predictable.

I haven't been able to find anything online that talks about how differing shapes alter the direction of the energy be transmitted/transfered. It seems the most common shape found for audio applications is the circle/sphere or cone. I wonder if this is because these shapes radiate energy more uniformly in a 360deg pattern? Then again, if using the simple tool posted earlier as an example, when both materials are the same the reflection is almost non-existent which would lead you to believe that the shape of the materials doesn't matter. Just speculating at this point unless solid info is found to say otherwise.

I think it's a difficult subject because the wide variety of materials and shapes that all interact differently with one another. The shapes must matter as they effect surface area of the interface. There's a good thread here on how differently cones/spikes can act depending on the materials involved. Just a spike and a flat surface can have widely varying outcomes.


Stenho, have you tried bolting the stand to the floor or even just giving it a few taps with a hammer to speed the settling process? What would you ideally prefer for a chassis material?