Equipment isolation and vibration damping.

[bdiament]

I am in the process of building my isolation platform and devices. I am rather sure the results will be in the same vein as room diagnosis and physical acoustic room treatment, i.e. fundamental and quite a large effect, and necessary in order to properly even start to hear your gear.

 

Considering the air cushion and the cup and ball arrangements, the effects are the following:

 

1. Air cushion: predominantly deals with vibrations and isolation along the vertical axis.

 

2. Cup and Ball: only deals with vibrations and rotations in the horizontal plane. Does nothing for the vertical plane.

 

Hence, you can see people combining these to great effect. Some use Springs to reach the same total effect. The balls move in the cup surface and then, have to oscillate back to equilibrium. Thus, the vibrational energy which would otherwise affect the components or radiate outwards affecting other components and sound, is instead spent as kinetic energy and heat through friction mostly and a little through air conduction/convection.

 

The air cushion is usually done either with tire tubes or similar air-filled bags or else squash balls, raquetballs or tennis balls. These support a first platform on which the cup and ball arrangement rests, and then either the component bottom rests directly on the balls if it is smooth enough, or else another platform is used.

 

Going from the principles of both vertical and horizontal isolation, I am thinking of mounting each cup and ball arrangement on its own rubber feet. The rubber feet will then do the vertical axis damping.

 

Another thing I thought about after doing a simple fluid damping for the tonearm of my turntable is to add some oil in the cup and ball arrangement. Perhaps having a fluid filled cushion would be beneficial too.

 

Good ideas?

 

Hi YashN,

 

In my experience, proper isolation provides benefits in every area of sound I know how to describe. Well, it doesn't really add anything; what it does it prevent degradation. Having heard what it does, I'd never go back. Everything in my studio/listening room is isolated, including the speakers and subs. I use isolation for the gear during remote recording sessions too.

 

As with anything else in audio, there is a wide range of opinions on exactly what types of vibrations are of concern and just where they originate. Some folks are concerned with vibrations from the speakers reaching the rest of the gear. Some are concerned with "draining" vibrations out of individual components. My experience has been the it is the low frequencies *entering* the gear that are the ones to be concerned with... simply because blocking these is what showed me the benefits that made me isolate everything.

 

As to "draining" my take is that if something is being drained, there should be less of it in the place from which it is allegedly being drained. Personally, I have not heard any audible evidence to support the idea that anything is being drained or diminished. On top of this, any path out is also a way in. The ad copy speaks of "mechanical diodes" but this is pretty easily dismissed by moving the shelf on which a component on such devices sits. There is no delay in motion and in fact, the component responds directly and instantly to the motion. So such devices are acting as couplers -- the diametric opposite of isolators.

 

Now, what I've learned in my own experiments is that just about anything you put under (or atop) a component will change its sound. The operative word is "change" which should not be confused with "improve." Couplers like spikes and cones *will* change the sound but what I've found is the change is somewhat random and inconsistent from component to component. These are ideal for folks who want to play with the "color" of their components. But they aren't the same as isolating a component and freeing it to do what(ever) it can do. (By the way, in my view, the idea of an "isolation cone" is an oxymoron, like "jumbo shrimp" or "civil war.")

 

To your question, I would advise against a rubber base for the roller bearings. An isolator needs to have a resonance in the low single digits because it doesn't begin to become effective until about 1.4x its resonance frequency. In order to block low frequencies, it needs to take effect at as low a frequency as possible. Rubber, sorbothane, squash balls and similar materials have resonances that are way too high. They will allow low bass to pass as if they weren't there and they will not become effective until the mid-bass. This means the response will have a peak in the mid-bass, adding "bloat" to the sound and a loss of definition in the bass. (Some may like this sound but a proper isolator will not alter tonality, it will free it.)

 

For more info see Vibration control for better performance, which I hope will be of interest.

 

By the way, after isolating the CD player and other electronic components, I experimented with isolating the speakers (just on rollers as I don't want to risk a tilt and fall on an air bearing, which is much easier to control with a regular component). Once I hear the changes, I described the non-isolated speakers as "bound and gagged" by comparison. Max Townshend has spoken about this stuff for years and years. Clark Johnsen too. I didn't start experimenting until about 2002. It has all been "afloat" on isolators ever since.

 

Have fun!

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Don't forget: A good roller bearing will also work in the rotational plane, not merely the horizontal.

 

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

...Makes sense if you think that the lower frequencies have a tendency to travel and make standing waves. Typically, there is a lot of energy around the lower frequencies...

 

Hi YashN,

 

The second sentence says it: there is more energy in the lows... enough to disturb gear. (Standing waves can occur at any frequency and are, in my estimation, not part of what is of concern with isolation.)

 

 

 

...For me, the analogy and term is correct. Take the example of a bookshelf speaker cabinet resonance...

 

In other words, we converted that detrimental sound from the cabinet into heat. Essentially, we drained the cabinet's energy down to the stand and sand where the loss as heat makes it sonically harmless...

 

While you may like the results of doing this, in my view, nothing was drained. What occurred by coupling the cabinet to the stand is an alteration in the resonant characteristics of the cabinet. (The cabinet is still resonating, albeit with slight changes in the frequencies because it is not part of a larger assembly.)

 

 

...Draining is also what's happening with the cup and ball arrangement (draining vibration into other motion and heat loss)...

 

Actually, what is occurring is what I described with the speaker and stand: the roller bearing acts as a coupler in the vertical plane. The motion of the ball is due to horiztonal (or rotational) excitement and what it is doing is *converting* the motion to heat. Again, in my view, nothing at all is being drained.

 

 

...I am a bit more concerned about the oscillation of the arrangement at low frequencies. I have manually and visually tested the air cushion and cup+ball this morning. I am wondering whether it would be better to dampen the ball motion further. Adding oil or some similar, perhaps more viscous material like silicone could help here...

 

It is precisely the oscillation that is creating the isolation. Adding damping will work against this as it will diminish the steepness of the rolloff above resonance (that last *is* the isolation). In other words the *less* the damping on ball motion, the steeper the roll-off, i.e., the greater the degree of isolation. This is one reason why I do not recommend using a "bowl" on the top and the bottom of the ball--it adds damping, thus diminishing isolation.

 

 

...When there is vertical motion, the air cushion dampens it to some degree...

 

The purpose of the air bearing, as I see it, is to prevent such motion from being transferred. It should *block* it, not merely dampen it.

 

 

...So, ideally, what I would like here, is that the kinetic energy in both the lower platform (air-cushion) and the upper one (cup+ball) be killed as fast as possible, i.e. I'd prefer 3 oscillations until rest rather than a 12-oscillation motion until rest of the balls in the cups.

 

Maybe a steeper curvature of the cup can help here or adding the oil or silicone.

 

As for the air-cushion, I would take the enclosure, ensure it is hermetic and try to fill it with fluid as well (perhaps just oil)....

 

Why 3 oscillations? I would suggest creating different designs and comparing them yourself. What I found when I did this is that the greater the number of oscillations (i.e., the less the damping) the greater the isolation. I would not want to deliberately add damping and thus waste the effort spent trying to isolate by effectively minimizing the isolation.

 

A steeper curve on the "bowl" will add damping to ball motion. I recommend a shallow curve. I want the ball to move as slowly as possible (lowest resonance frequency) and for as long as possible (least damping on ball motion). In this way, isolation takes effect at the lowest possible frequency and the rolloff above resonance is as steep as possible (i.e., maximum isolation).

 

 

...Taking the example of sand above, where ultimately, energy is dispersed at heat, this morning, I was also wondering whether I could substitute either or both of our platforms with sand or bags or sand.

 

I've seen people mention a bed of sand as support, so this is of the realm of already tested and proven tweaks.

 

For the upper platform, supposing you keep the air-cushion, 3 little bags of sand in place of the cups+balls perhaps?/

 

In my experience, sand will dampen energy but it will not provide effective isolation.

If you have the inclination, by all means, try all these things out. That will tell you much more than anything I could say.

 

 

...I think I would need a platform to isolate all these components, although I initially thought of doing the isolation for the amplifier first and then the speakers.

 

Which do you think I should start with?

 

I'd definitely use a separate platform for each component being isolated.

While I've heard benefits with everything I've isolated, I would say the greatest benefits have been with digital gear (CD players, transports, DACs, ADCs) and with loudspeakers. Speakers were a revelation because I'd always heard that speakers should be mounted rigidly. Now I would say the *drivers* have to be mounted rigidly to the baffle, but want my speakers "afloat."

 

So between amp and speakers, I'd try the latter. If they are stand-mounted cabinets, I'd first try isolation between the speaker and stand, rather than isolating the whole stand. The reason is that I've found the closer the isolation is to that which is being isolated, the more effective it is.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi YashN,

 

I've seen you mention this a few times, but I cannot visualise it or translate it into normal mathematical/physical terms.

 

Let's take the horizontal plane of the upper platform, i.e. the one above the balls, and let's look at it from above. This constitutes a horizontal plane.

 

This horizontal plane can have a few types of motion which we want to dampen:

 

1. The horizontal plane can suffer translations along that plane, i.e. horizontal displacements, i.e. North-Sound, East-West, any direction and its opposite, really or outward-inward if you prefer.

 

2. The horizontal plane can suffer motion by rotating around a vertical axis, i.e. clockwise-anti-clockwise rotation. Even when rotated, it is still a horizontal plane.

 

 

Is 2. what you meant by 'rotational plane'?

 

First, per my previous post, I don't want to dampen anything. I want completely free, undamped resonance in the isolator in order to achieve the steepest rolloff above resonance, i.e., the greatest degree of isolation.

 

Actually, there is a difference between horizontal energy (front-to-back and side-to-side) and rotational energy. You've described the difference in your post.

I suppose you could see them both as horizontal in some respects. I see them as different.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

You may not like the term 'drain' as people commonly use it, but that's correct and it's exactly what is happening....

Hence the detrimental energy at the cabinet is drained into heat, if you want a shortcut....

 

Hi YashN,

 

If something is being drained, I expect there to be less of whatever was drained in the place from which it supposedly was drained.

If you check that speaker cabinet, you'll find it is still vibrating. This is a characteristic of the cabinet, i.e. the speaker design, and it won't go away simply because the cabinet has been coupled to a larger structure like a stand. The energy conducted into the stand is *converted* to heat. No energy is lost.

In contrast, when I drain my kitchen sink of water, it ends up with no more water in it. ;-}

 

 

 

The conversion to ball motion and heat is the draining. With the ball motion, since it has to rise along the curve of the cup, there is also a vertical component to the motion...

 

You can see it that way if you want to. I would not agree as I differentiate between conversion (where no energy is lost) and the concept of "draining" as it is applied in the marketing literature (where energy is supposedly being removed -- which by definition, is what draining is). To my knowledge, no one making such claims has yet demonstrated that there is less of what is supposedly being drained in the place from which this is said to occur (i.e., the component).

 

 

Now this is interesting, because in my visual and manual tests today, the freer the motion, the longer my upper platform oscillates. Much like a pendulum in air vs a pendulum in water: the oscillations in air will last longer than in water. In water, the pendulum loses more energy working against water and hence the oscillations die down quicker. In other words, the freer the motion of the balls, the longer the platform and component are oscillating. Do we want our component to oscillate? Or do we want our component to oscillate the least?

 

Yes, exactly: the freer the motion, the longer the item oscillates. Less damping, manifesting visually as longer oscillations, equals a steeper rolloff above resonance, manifesting as a greater degree of isolation.

 

 

The observation about roll-off is I believe important but are we here looking to :

(a) tame the low-frequency oscillations which are particularly detrimental at a specific resonant frequency?

(b) to isolate from vibrations effectively across a large bandwith of frequencies?

 

These two seem different to me....

 

We want to block performance-degrading vibrations from entering the gear. In my opinion, these tend to be primarily very low frequencies coming up from the ground. (The higher frequencies of vibrating air in the listening room have *much* less energy than the low frequencies I'm speaking of.)

In order to block those low frequencies from interfering, the isolation must begin to take effect at a lower frequency. (The isolation will be effective from ~1/4x the resonance of the isolator and increase in effectiveness with increasing frequency. Think of a low-frequency resonant peak with a roll-off at frequencies above the resonance.

So, we *are* blocking transmission of vibrations across a wide bandwidth. The key is: At what frequency will the blocking begin? To be effective, this has to be as low as possible.

 

 

It won't prevent it, only diminish it as it is transferred with loss to the upper arrangement...

Technically, yes. Unless we have a perfect isolator with an infinitely steep roll off. But *effectively*, with a properly designed and properly implemented isolator, we're preventing those frequencies from harming performance. At least that is how it sounds to me in comparison to no isolation.

 

 

 

 

The number is unimportant, it was only an example, together with 12, to show what I meant when I said I wanted less oscillations: if the ball + cup arrangement has to work against a viscous material there, then the kinetic motion is lost to move the material and as hear, the oscillations are damped and hence the upper platform and my component comes back to rest more rapidly instead of continuing to oscillate with the balls.

 

Yes. And there is little to no isolation. Try it and see/hear for yourself.

Taken to its logical extreme, it you damp the motion completely, there is no oscillation at all. And no isolation. Again, it is precisely those oscillations that provide the isolation. The goal is more of them, not less.

 

 

By isolation here, do you mean as perceived sonically, or as vibrations as measured through equipment?

Both sonic and measurable visually. No need to lab gear. Try it different ways and compare for yourself.

 

 

Yes, except, people mean different things when they say 'isolation'. Here, if my component is oscillating in sympathy with the balls and for too long, to me, it's not isolated. Ideally, I want the ball motion to occur but I don't want the upper platform and the component to do the same in sympathy. In the ideal case, I'd rather the energy loss occur at the ball and cup the in the smallest amount of time, so that the upper platform and component move the least.

 

All I can tell you is what *I* mean by "isolation" and that is that ground borne vibrations do not impede the performance of the gear because they are effectively blocked from entering the gear. The only point of the ball motion is to impart that motion to the component being supported.

Put another way (and this is something only you can determine whether or not you want), isolation = oscillating gear. More isolation = more oscillations.

Fewer oscillations = less isolation. The relationship between the two is direct and will not change.

 

 

I suppoer in this viewpoint, we can consider a low frequency of oscillation below what's considered acceptable for turntable's tonearm + cartridge, e.g. something like a frequency of <10Hz to be acceptable oscillation?

Complete different things. A cartridge/tonearm resonance needs to be ~10 Hz so that it is above the frequency of typical record warps, yet below the audible range. We're not trying to block anything there. The cartridge/tonearm (or more specifically, the cantilever/tonearm), being a compliant structure is going to have a resonance. Isolation, while also involving resonance (of the isolator, if we're using a mechanical low-pass filter like a roller bearing or an air bearing) is something else.

 

A good isolator will have a resonance in the low single digits. Something like 2 Hz or 3 Hz would be great. (That's why that roller ball needs to oscillate *slowly*.)

 

 

It amounts to the same effect: I don't think we can have perfect isolation. Damping is the working principle behind normal, imperfect isolation.

 

Rather than repeat myself, I will suggest you try experimenting with sand vs. well-designed and well-implemented mechanical low-pass filters (with different amounts of damping). Sand if not a very effective isolator. Again, high damping = slow rolloff = little if anything in the way of isolation.

 

 

I think I will try some arrangements if I can get my hand on some sand (doesn't seem to be easy here). If not, I may try with cat litter.

 

Have fun!

 

 

OK, thanks for your feedback....

 

I hope some of it proves useful to you.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Yes, but less, which is the whole point, hence we are draining the cabinet of its harmful resonance (nobody said we were draining perfectly).

 

More responses at length during the week

 

Thanks for the cool exchange of perspectives.

 

Hi YashN,

 

My point is that there is no less. Different, yes. But not less. In fact, I say exactly as much.

I note that no one using the term "draining" for audio has yet released any measurements to support the claim that anything is being drained (other than the customer's wallet ;-}).

 

Thank you too for the dialogue!

I hope you find the same pleasures I've found in isolating gear.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi YashN,

 

It would be interesting to see why you think it is so.

 

Based on Physics principles, when you couple the speaker cabinet with additional mass, it is the whole arrangement which vibrates, with a lesser resonance. The energy that the cabinet itself used to vibrate previously is now dispersed among both the cabinet and the stand.

 

But there is not a static quantity. It is constantly being replaced.

 

It follows that the cabinet has less energy in its vibration than before. Hence, it has less deleterious effect in combination with the driver's performance....

 

I would say it follows that the resonant characteristics of the cabinet have changed but in now way is there less energy. It is not a cup with a static amount of water in it. It is a dynamic system, always in "motion".

 

 

Another way of proceeding is to do without the cabinet altogether: open-baffle...

 

My speakers (Magnepans) are boxless. They still benefit a great deal from blocking vibrations from entering the speaker (and crossover).

I make no attempt to drain anything; only to block seismic vibrations from entering the gear.

 

 

Now, if your speakers oscillate back and forth, what is it that happens to the soundwaves the drivers are supposed to send to you?

 

I would guess the question (a quite reasonable one) comes from the assumption that the motion of the drivers will cause the balls in the roller bearings to roll--a Newtonian reaction to the action of the drivers. A properly designed roller bearing will have a resonance in the very low single digits, well below any frequency edited by any speaker (or subwoofer) I know of. In effect, the rollers to not "see" the speakers. The drivers in the speaker move against the relative stillness of the baffle on which they are mounted.

 

So what I hear happening with speakers that *can* oscillate due to being isolated on roller bearings, is the speaker now sounds like it has been unbound and ungagged. Performance in every area of sound I know how to describe is improved: extension (at both ends), dynamics (at both ends), low level detail, bass pitch definition, soundstaging expands (in all three dimensions, provided the recording contains the information), images are more "solid", and overall "focus" improves appreciably.

Townshends Seismic Speaker Stands accomplish the same thing. I've heard different speakers demoed on the stands vs. off the stands and found the improvements on the stands were not subtle.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

I feel terrible about injecting this very simple question into a conversation in which I really have no business being involved in, but Barry, what have you found to be the optimal "cup" for your roller bearing method? I am using furniture cups now but they are almost infuriating to work with, in that they don't hold the bearing securely enough to be able to move equipment even a 1/4 inch without the bearing moving to an area where it is no longer supported properly. The pictures in your first post look like a metal receiver for the bearing, any insight on where to purchase something like that or was it custom made?

 

Hi Todd,

 

Don't feel bad -- I think your post is entirely apropos in this thread.

To answer your question, after my initial experiments with "econo" rollers (using marbles and Easter egg holders -- which very easily outperformed a set of $400 "racing cones" -- for a total cost of under $2) I tried a number of materials. When I came up with the design I wanted I drew up a spec and found a local machinist to make some prototypes for me. I had some made in 6061 aluminum and in 7075, ultimately deciding on the latter. Then I went back and had enough 7075 sets made to support all the gear in the system, including the speakers and subs.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi Daudio,

 

Going for 'optimal' whether with commercial products, or having your own cups custom machined is going to involve some serious money, especialy for enough devices to outfit an audiophiles complete system...

 

When I had mine made I spoke with several machine shops before finding the one I ended up using. Some gave me silly quotes like $150 per piece (!) and added a $75 "setup" charge. Preferring to *not* be "set up" ;-} I thanked them and went elsewhere. The machinist I chose made me sets of three at well under $50 per set. Of course, this was over a decade ago, so one seeking a shop today would have to consider the degree to which the dollar has been devalued in the intervening years.

 

 

I have never quite understood how the rough wood cups and imprecise marbles are able to deliver any audible performance at all, but Barry says he has heard it...

 

Indeed, it was the initial Easter egg holder and marbles prototypes that convinced me the idea of roller bearings is so good that even a trio of 29 cent holders from a "crafts" store, with an ordinary marble in each very handily outperformed a $400 set of "racing cones." The former provided consistent and repeatable isolation--albeit not from an optimal frequency, with less than optimal surfaces, and "bowls" of less than optimal shape--while the latter allowed somewhat random color changes.

 

 

These retail 'dimple blocks' do not have as shallow a cup as would be preferred, but a smaller size of the SS 'ball' (while still usable) can optimize the relationships of the curves to make the best of the situation...

 

While they will certainly work--as you already know, I would not agree that changing the size of the ball will do anything significant with regard to the curve. (Well, maybe if the ball was so small, it no longer protruded from the top of the "bowl" -- but even then, I say still no significant effect in terms of resonance frequency or damping. That's basically all in the bowl.

 

 

I have my entire system, electronics and speakers, outfitted with these DIY Acrylic/Stainless Steel devices, and I clearly heard veils lift when, first the digital stuff and amp, and then the speakers and subs, were floated. No going back ...

 

Sounds like your experience has been very similar to my own. Definitely no going back.

 

 

(P.S. the inspiration for me to do this experimentation came directly from Barry's BLOG and forum postings. Thanks BD)

 

You are very welcome! I'm so glad the writings were of use. It really is an easy (and for do-it-yourselfer's an economical) way to significantly up a system's performance.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Roller bearing devices are not always a silver bullet. It can be device dependent. A friend of mine put his SACD player on some (don't recall the brand) and they sucked all the life out of the sound. Replacing them with brass cones revived it.

 

Hi Allan,

 

Actually, properly designed and implemented, I think they are a silver bullet and utterly indifferent to the device. In the case you mention, it might be the particular devices, it might be how they were placed, and/or it might be the bottom of the SACD player chassis. If the life was sucked out of the sound, I would say something was wrong but would not blame the idea of roller bearings.

 

I speak for myself of course but I've never heard a component that was not appreciably freed to better performance when used with a properly design and properly used set of rollers. Basically, they prevent vibrations from entering the component via its support. They should not alter tonality and certainly not remove life. Again, I say something in the design or the implementation was at fault in that case.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi Barry:

 

Then it sort of becomes Catch-22, doesn't it? If the device doesn't improve the sound, then it is "not properly designed and implemented". One could argue that, if the component itself was properly designed and implemented, these devices wouldn't be necessary.

 

Hi Allan,

 

I don't see a catch-22 here. It is simply a matter of approach.

As to these devices not being necessary with optimally designed components, agreed 100%.

It is nice to see some designers (notably Carl Marchisotto with his Nola speakers) incorporating seismic isolation into their products.

 

Back to outboard devices, if the sound isn't improved by isolating devices, any one or more of the following might be true:

 

1. The device isn't a particularly good isolator.

Many devices sold as isolators are actually couplers - the diametric opposite. Other devices sold as isolators have resonances that are too high, often in the mid or upper bass. Their isolation does not take effect at a low enough freqneucy to be effective. And these devices result in bass bloat to boot.

 

Among rollers, some are sold with "bowls" to be used above the ball as well as those used below it. My experience has been that the second bowl adds damping, which diminishes the rolloff above resonance, i.e., diminishes the degree of isolation.

 

2. Assuming a well designed isolator, placement becomes important. With good rollers, I've found it is important to place them in an equilateral arrangement to allow for the greatest freedom of motion. With air bearings, I've found it is important to inflate them *only* enough to achieve air support -- to lift the load so it isn't resting on the inflation valve. One pump too many and the resonance goes up, bass bloats, and isolation diminishes.

 

Rollers need a *very* smooth surface resting on them. If the bottom of the component to be isolated has a textured surface or has ventilation holes, motion (and hence, isolation) is impeded. With such components, I use a dead marble tile, smooth side down against the roller bearing balls. The component sits atop the tile on its own feet.

 

Back to air bearings: Some folks use multiple small air bearings to achieve balance (so the gear doesn't tilt). This alters the function of the air bearing from isolator to level. Once that is done, it isn't such a good isolator. Better, in my view, to use an air bearing (aka inner tube) that describes a large enough circle to accommodate the component to be isolated. In this way, components where the bulk of the mass is off center, can be placed off center atop the platform atop the air bearing. In other words, if an amp's left side is heavier than its right side, I'd place the amp to the right of center atop the platform, thus achieving balance without having to alter the minimal inflation of the air bearing. (This means the shelf on which the air bearing sits, as well as the platform between the air bearing and the component being supported, needs to have adequate size.)

 

3. As I noted in my article on the subject, it has been my experience that certain other things must be attended to before the benefits of isolation will be plain. Among those other things are clean AC power and proper routing of system cables (signal cables separated from power cables, etc.).

 

As with any system setup, there are priorities. The smaller things won't be audible until the larger things (starting with speaker and listening position placement) are properly addressed. Once all is in order, the benefits of isolation become clear.

 

A last thought: In my experience, not everyone will like the results of increasing the neutrality of a component or system. Some folks prefer certain colors in their gear (such as can be achieved, albeit somewhat randomly, with spikes and cones) and I'd never argue with whatever brings anyone their listening pleasure. That said, for those who have taken great care in their system setup and want to hear their gear do even better at what I call "getting out of the way" I recommend trying some good seismic isolation techniques.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

FWIW, the devices my friend used were of this category. The difference heard when they were replaced was both immediate and significant. My friend's wife, who was in the kitchen, noticed and commented on it.

 

Hi Allan,

 

I'm not surprised.

The larger contact surface presented by a "top" (vs. the ball) also tends to alter tonality, adding brightness and sometimes, outright "chatter." I would not think the difference would be subtle.

 

Like I said, "properly designed and properly implemented." The fault was not at all in the concept of roller bearings for seismic isolation. It was in the approach.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi Daudio,

 

Barry,

 

A decade ago in an craft that is going more to the big CNC shop and less to the 'old lathe guy' kind of thing, I used to see a lot of around the 'Motor city'. Not so much anymore

 

Anyway, how many units did you buy at that > $50 per set price ?

 

That was "well under $50 per set."

I made enough for everything, including the speakers and subs. If I recall correctly, probably 12 sets.

 

 

OK, how's about we put some numbers to them ? What diameter are your bowls (the full negative space, not the block intersection) and bearing balls ?

 

I'll figure out my variations...

 

Iteration #1 uses a .96" diameter bowl that is a section of a 2" sphere. Bowl 1/8" deep at the center.

Balls are 1/2" chrome steel. I auditioned Tungsten Carbide and they clearly sound even better -- but $9 per ball was more than I chose to spend. A bag of 100 chrome steel balls was a couple of bucks at the time.

 

Iteration #2 to come. I have some ideas.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

So, this is all great reading, and I just happen to be playing with my old sets of Daruma 3 mkII roller bearing isolators on a pair of Omega Super Alnico monitors.

 

Sounds like I should attempt to remove one of the brass pieces and just use the ball with the brass cupped piece. This will be on my "must try soon" list!

 

And finally the question has to be asked.... What is the "best" ball material?? Brass? Titanium? Maple - other hardwood?? I do have a couple of sizes of maple balls coming this week to play with; either will fit into the cupped brass end of the Daruma... Cheap experiment!

 

Keep the great commentary coming!!

 

Hi 1markr,

 

In my view, the best ball will be the hardest and the smoothest.

The best I've heard are Tungsten Carbide but the price resulted in my deciding I was quite happy with Chrome Steel.

 

I would not deem wood an optimal material for such. Not hard enough and not smooth enough.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Thanks Barry.... good to know about the Tungsten Carbide and Chrome Steel. I will have to check Grainger or similar if they have them in 3/8", 7/16", or 1/2" diameters. Any other known sources?

 

As for the maple, it cost me $6, after shipping, for 12 balls of two sizes, so mehhhh, we'll see what happens!

 

Cheers,

Mark

 

Hi Mark,

 

At the time, I got them from smallparts.com, which I believe has since been absorbed into Amazon.

Grainger may be a good bet.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi Dave,

 

Hi Barry,

 

Ok, so we're talking about less then $500 for a 'system set' (as opposed to a 'device' set of 3). Ten years ago. And access to custom machining services is variable. Unless some enterprising job shop got setup to make, a nice, simple design... (Alex, are you listening ?)

 

I spent about $200 on blocks, balls, and inner tube stuff. Some was experimental, and wasted. But I ended up with about the same quantity.

 

There are still local machinists to be found. There are also on-line services some folks I know have used.

We spent different amounts but didn't you say you are using off-the-shelf acrylic "bowls"?

I think that might account for it as I brought a specification and drawings to a machinist and had them made per my design, out of the material I chose.

 

 

All right ! So you have 2.00" dia bowl, and a .500" ball. 4 to 1 curvature ratio.

 

My 'medium' size Ball Block has a 1.300" section of a sphere, and the ball is 3/8" - just enough for a nice low profile device.

 

So the DTA DIY do-hickey, is 1.30" to .375", for a 3.47 to 1 ratio.

 

That doesn't sound like a huge difference, to me And I'll bet that your upcoming 'Iteration 2' will have a larger sphere diameter bowl, right

 

Cheers,

Dave

 

Sorry if I wasn't clear but that is not correct. As I said, "Iteration #1 uses a .96" diameter bowl that is a section of a 2" sphere."

That's a .96" diameter bowl, not 2". Further, to look at the bowl diameter to ball diameter ratio without looking at the shape of the bowl is, in my opinion, to miss the mark entirely. As I said earlier, from my perspective, it is all in the bowl.

(Iteration #2 will have to remain close to the vest for now.)

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi Dave,

 

...I'm not sure we're in sync with our words here.

 

Imagine you have a big steel ball that fits perfectly into the bowl, touching on all sides and on the bottom of the bowl. The size of that 'ball' would be what I call the spherical diameter (or radii). and it would be larger then the circle made by the intersection of the bowls concavity and the top surface of whatever it is carved out of (the block). The diameter of that intersection circle and the depth of the bowl in the block form a chord, and can be used to calculate the spherical diameter, but I forget the formula .

 

It is not so easy to measure the bowls spherical diameter, since there is nothing there to lay a rule on ! My bowls diameter was larger then any of my radius gauges, so I went looking for thin disks (could be washers) till I found one that fit perfectly. It was 1.3" in diameter. One could draw, and cut out, paper circles to do the same thing. I don't think tight tolerances are necessary here

 

So where you are saying 'the shape of the bowl', I think I am saying 'the spherical diameter', and I believe both boil down to the curvature of the bowl, as expressed by a diameter or radius.

 

Does that work for you ?

 

I'm not sure either.

If you place a 2" diameter spherical ball into the bowl of one of my Hip Joint roller bearings, it would fit into the bowl touching it on all surfaces.

But I wasn't thinking of spherical diameter when I described the Hip Joints. Well, perhaps this is what came to your mind when I mentioned the shape of the bowl is a section of a 2" sphere. ? (I didn't have to measure this because the number came from my head during the design phase, as I made a drawing of what I wanted.

 

When I refer to the diameter of the bowl, I'm talking only about the top edge of the bowl. The whole Hip Joint is a 2" circle, 1/2" thick.

 

Not sure if I'm any more clear than before. I hope so. ;-}

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Barry,

 

How about we call the bowls spherical diameter (size of the imaginary ball that fits into the bowl), the 'bowl curvature diameter', or just 'bowl curvature' ?

 

We can if you like. I prefer to describe the curvature I used as "a section of a 2 inch sphere" since that is the geometric term I've always heard.

 

 

OK, so we refer to the opening of the bowl as "bowl diameter"

 

I'll only use those terms so we should be in sync Bowl curvature is the measurement I'm interested in here, to compare with the ball diameter, as I did eariler:

 

Barry's Hip-Joints = 4.0 to 1 curvature/ball ratio.

DTA DIY Ball Block = 3.5 to 1 curvature/ball ratio.

 

Do you think that 12% difference might have have a significant effect on my devices's resonance or rolloff ?

 

I wish I had a better idea of which characteristics are more important for the bowls and balls: hardness, smoothness, curvature, concentricity, or ? ??

 

(good night)

 

Again, I do not think the ratio of bowl to ball diameter is an adequate criterion for comparison.

To answer your question, in my opinion, the difference is significant.

 

In my experience, hardness and smoothness are key for both the bowl and the ball. Add shape (or curvature) and size for the bowl, which I consider #1.

Concentricity in the ball is icing on the cake. But I'd trade some perfection there before I traded bowl geometry, smoothness, and hardness.

 

As always, my best suggestion is to try out different geometries and materials. There's nothing like direct experience to find out how *you* would rate the importance of each of these.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

I wouldn't get too stuck on the term if I were you: there is a Law of Conservation of Energy, but there is also something called 'Heat Loss'. When someone says 'drain', they do mean the conversion into heat or some other form of energy, thus less energy at the point where it is 'drained' from or converted from.

 

Thus, the cabinet vibration is diminished or drained in sand motion and heat. Yes, it is conversion and overall conservation, but the energy which would otherwise have been used for vibrating the cabinet only is now partially going elsewhere, so both are correct. It is a fact of Physics.

 

Hi YashN,

 

I hope we can agree to disagree.

Again, to my mind, nothing at all is being drained unless one can demonstrate there is less of it in the place from which it has supposedly been drained. This is by definition of the term.

 

Converting energy is not draining anything. It is converting it. By the way, in your example, it is the sand that is heating up, but the cabinet is still vibrating.

 

Anyway, as I said, rather than go round and round, let us agree to disagree on this.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi John,

 

Great post!

 

I'd differ on two things though:

 

...Note that the bowl shape is not part of the function of the isolation, it actually hinders the isolation somewhat but is a necessary evil for a practical implementation...

 

The other aspect of the bowl is that it adds a resonance...

 

To my mind, these two statements are at odds. I understand that if we had a frictionless flat surface, we wouldn't need the bowl, but in practical terms, the bowl--more specifically the rolloff ensuing from the resonance created by motion of the ball in the bowl--is exactly what provides the isolation.

 

...The "steepness" of the bowl also affects the resonance frequency, so again a shallow bowl is a good thing...

 

In my experience, the steepness of the bowl primarily affects damping on ball motion. I see the diameter of the bowl as affecting the resonance frequency. Hence a wide, shallow bowl is the idea.

 

Or perhaps I'm misunderstanding and we're saying the same thing with regard to both of these.

 

By the way, with regard to front panel buttons of devices on rollers, I've often said one must develop a gentle touch. A few fingers on top of the device while pressing buttons with the thumb -- or one hand to steady the device and the other to push the button, will work when one is at the equipment rack. Of course, the remote works just fine too.

 

I've also mentioned folks should be wary of pets and people near the gear, especially loudspeakers "afloat" on roller bearings. (Though for some reason my dog, who loves to join me in the studio/listening room, has not presented any issues at all.)

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi YashN,

 

I gather by this we are OK with oscillations that are sub-10 Hz or so, assuming these are not modulating the audible spectrum? If we are looking to maximise the oscillations for longer, then why aren't we using very large and very shallow cups?

 

Sub-10 Hz is not good enough. The resonance needs to be in the low single digits.

Wide, shallow bowls are what I recommend.

 

 

I've seen some commercial products managing just that for Lab equipment. Concerning our DIY arrangement, do you find that the amplitudes we are isolating for are rather small? That is, in normal use (without touching or pushing the component), do you find them oscillate visually at all?

 

I can see the upper platform move around at low frequency when testing manually (providing a large impulse by hand), so in that case, the platform moves, our component is moving as well as around, say 10Hz for several seconds. Can we live with that?

 

The only time I see motion is when I physically push a component atop a set of rollers.

In normal use, the amplitudes and frequencies are so small, I have not seen anything. Again, in my opinion, 10 Hz is way too high for the purpose of an isolator.

 

 

Depends on the nature of the damping (c.f. the Wikipedia extract above). I have seen a couple of contradictory messages about isolation/damping around the web. Perhaps someone whose Physics is less rusty than mine can chime in.

 

I don't know about Wikipedia. There is no dependence. Degree of damping is inversely proportional to the steepness of rolloff above resonance. More damping = less steep rolloff. Less damping = steeper rolloff. This does not change.

 

 

OK, that answers my questions above. So here again, assuming we have a ground wave of 10Hz which is making our speaker oscillate front and back through the cup and ball. Doesn't this modulate the drivers' sound? Can we live with that?

 

In my opinion, no. In order for the isolator to be effective, it has to have a resonance in the low single digits. I'm talking more than two *octaves* below 10 Hz. Lower is better.

With a proper resonance, there will be no impact on the drivers, which even with a subwoofer, are well above what the isolator will "see".

 

 

Both these cases have resonance and frequencies that we want to have at less than audible. We are indeed trying to block something in the turntable example: we do not want the tonearm/cartridge assembly's resonance to impact what we hear. So these two cases are similar, not different.

 

Here again, I hope we can agree to disagree. We're not trying to block anything with a phono cartridge. We want it to be immune to record warps. This is not the same as blocking the warps because we're setting up a resonance that will not "see" the warps but will still be low enough to get out of the way of the audible frequencies.

 

 

If this is the case, that's not what we're doing here: the cup's curvature is too steep and the cup is too small. You'd need a much shallower and larger cup and allow the ball to move as far as possible before returning.

 

Perhaps this is true of whatever you've tried so far. It would seem so if you see oscillations on the order of 10 Hz.

It's what *I'm* doing with my Hip Joints design though. Some of the commercial rollers are doing this too.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

Hi YashN,

 

I was going to write about just that: tilt. There are six degrees of freedom that we want to isolate from. If the air cushion deals predominantly with vertical motion, and the cup-and-ball with the horizontal plane, then we still ideally need to deal with tilt...

 

I'll speak for myself of course but I'm not seeking to isolate from anything but seismic vibrations entering the gear via its support. These tend to take the form of what are called P waves (which travel horizontally) and the lesser S waves (which travel vertically). So tilt has nothing to do with the paths by which vibration might be transmitted into the gear.

 

The fact that the inner tube isn't completely inflated: does that help with tilt? Would a completely inflated tube deal with tilt better? Less?

 

In other words can we consider that the inner-tube is also isolating from tilt to some extent?...

 

If the tube is inflated too much, its resonance frequency goes up, making it ineffective as an isolator. Tilt can easily be accommodated by using an air bearing (aka inner tube) that describes a large circle. (I believe I wrote about this in an earlier post in this thread.) A larger diameter air bearing and a larger platform for the component on a larger shelf will do the job. A component with off-center weight concentration can be placed off-center in the opposite direction atop the platform to compensate.

 

If the air cushion's effect on tilt is not satisfactory, the separate smaller air cushions appear to be interesting, and namely why was thinking of making the air cushions part of the ball and cup assembly...

 

As I mentioned earlier, once you start using the amount of inflation to compensate for tilt, isolation is either compromised or eradicated.

 

 

 

Another question occurs if we consider tilt: if we find that the air cushion, which is in fact rubber together with air, doesn't deal with tilt efficiently, what could we do to implement tilt control?...

 

Answered above (and in a previous post). What I described has worked for me. For example, my amps have their transformers on the left side of the chassis. As a result, they are heavier on the left than on the right. With extra room built into my rack and platforms for exactly this purpose, I place the amps right of center on the platform (actually on the roller bearings which sit on the platform). The result is a nice, level (non-tilted) platform.

 

 

For tilt control, a simple pad between two hard surfaces should suffice, something like Sorbothane...

 

This will eradicate any isolation. The resonance is too high. It will also cause bass bloat.

 

As always, my best suggestion is to *not* simply take my word for what I'm reporting. Try all these things out for yourself so you can draw your own informed conclusions. And remember to have fun!

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

I don't see why having a tilt management layer is going to affect the other layers.

 

When you noticed bass bloat, what dimension of Sorbothane (or other) were you using?

 

Hi YashN,

 

All I'll say is try it yourself and see.

 

It isn't the dimension of the Sorbothane, it is its inherent resonance, which in my opinion, is way too high to be effective for my purposes.

Basically, anything that has a resonance in the audible bass range is going to cause bloat -- the nature of the bloat being tied to exactly where the resonance peaks and to its Q (i.e., bandwidth).

 

If you try it, you'll know what I'm referring to.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

What is the resonance of the inner tube air-cushion?

 

Hi YashN,

 

Unlike a solid material like Sorbothane, the resonance of an inner tube used as air bearing will depend on the amount of inflation.

That is why I say to use *only* enough inflation to lift the load off the inflation valve. Properly done, we can get the resonance in the target zone.

 

Townshend has done this too with some of his products.

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio

[bdiament]

...If we agree on this, then the cups should be large, allowing for greater shallowness. I don't see this in either the commercial versions or your version...

 

So, the current implementations are a compromise and sub-optimal...

 

If you read my earliest posts, you'll see that I was saying this all along. (Remember when you wanted damping and a steeper-walled bowl?)

 

Practical considerations come into play in order to fit a trio under gear. Yes, a larger diameter bowl would lower the resonance. The idea however is to get the resonance down into the low single digits. Taking it from 2 Hz or 1.5 Hz to 0.75 Hz would offer a *theoretical* benefit and *perhaps* a practical one but the former will already be effective. (We're talking about isolation taking effect at 2.8 Hz or 2.1 Hz vs. 1.05 Hz.) Still, if you want to, consider the largest bowl you find practical.

 

This is an example of how understanding the physical phenomenon and thinking things through can help the design and, am hoping, the efficiency and the results.

 

I've long been an advocate of understanding the phenomenon and thinking things through. Add some direct experience to the theoretical analysis and you'll have it nailed.

Remember to have fun!

 

Best regards,

Barry

Soundkeeper Recordings

http://www.soundkeeperrecordings.wordpress.com

Barry Diament Audio