Equipment isolation and vibration damping.

[YashN]

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?

 

Don't! The more the merrier, and it is a question I was going to ask myself. Once we know what we want to achieve, we can then see how, and then also think about how to optimise the characteristics.

 

Indeed, if it's the case that we want the slow oscillations to last longer, think we should look for very shallow cups, and it turn that means, very large ones. Probably with the smoothest and hardest surface as well.

[YashN]

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...

 

Certainly because even with less than ideal materials, the process of isolation is occurring to some extent. Perfect for initial experiments compared to no isolation at all.

 

All my cups are acrylic moldings made for retail display purposes. They are available in an array of sizes, shapes and colors. And they are inexpensive ! I can put together a set of 3 for $10 - $15.

 

That's very cool, Daudio. As ringing, is acrylic a good material? I gather we'd need hardest and smoothest, but also something which does not ring too much, or if it does ring, then we need to dampen that by encasing it within some other material.

 

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.

 

I hope to learn that one day, but for now I feel have achieved almost a 'Regen' level SQ improvement, for less then even that small cost.

 

Very encouraging. How large can you make these?

[Daudio]

This CAN be a problem for devices you need to push buttons on, this can cause the gear to move a lot if you use a shallow bowl. Think carefully about using remote control for these functions, or using a rotary control rather than push buttons.

 

John,

 

It can be quite a problem with changing cabling !!, Even with optical disks and USB sticks, if used under a PC.

 

My Mac Mini and DAC sit on the same 10" x 16" acrylic plate, which rests on 3 SS balls siting in acrylic cups. When I changed out a pair of stiff interconnects, with some softer ones (not the intent), my gear kept wanting to slide off to the side. It seems that the shelf it all sits on is not quite level ! A chunk of foam is keeping things in place until I get around to leveling the stack.

 

That quasi-frictionless feel is what tells me that I have the rollers setup properly. Not always easy to do under heavy equipment

 

 

As has been mentioned the lowest resistance is the goal, this is achieved with hard round surfaces rolling on hard smooth surfaces, so things like chrome steel etc are pretty good choices.

 

 

While I'd like to have the 'hardest' material for the bowls, as well as balls, the expense rules it out for me at this time. But I'm pretty happy and hopeful for the acrylic bowls, and acrylic in general (I can work it pretty well).

 

While its hardness compared to metal is quite low, what gives me that hope is the 'smoothness' factor. The cast parts I'm using have very, very smooth surfaces. So smooth it looks like water under surface tension. I can even use a torch to 'flame polish' edges of cut sheet, and achieve the same smoothness and gloss.

 

I have a feeling that the smooth surfaces on the bowl, ball, and suspended plate (if not the bottom of some gear) contribute to the functioning of the isolation device.

 

 

I look forward to more comments from you on vibration control for audio. Thanks for chiming in !

[YashN]

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.

 

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.

 

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.

 

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 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.

 

It is very difficult, however to synchronise the air content of three different tubes though.

 

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?

 

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

 

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.).

 

That's a good reminder from your article, something to be put on the checklist when testing, and generally setting up.

[YashN]

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

 

Checked them out at this Soundstage review. They look cool and not too expensive. Are the tops curved as well?

 

Review Summary

[TABLE=width: 100%]

[TR]

[TD=width: 100%, bgcolor: #FF0000]

[TABLE=width: 100%]

[TR]

[TD=width: 100%, bgcolor: #FFFFFF]

[TABLE=width: 100%]

[TR]

[TD=width: 20%]Sound[/TD]

[TD=width: 80%]"The clarity of the music improved, the soundstage became more specific, and the bass seemed to tighten and become more authoritative."[/TD]

[/TR]

[/TABLE]

[/TD]

[/TR]

[TR]

[TD=width: 100%, bgcolor: #FFFFFF]

[TABLE=width: 100%]

[TR]

[TD=width: 20%]Features[/TD]

[TD=width: 80%]Three pieces -- top, bottom and ball bearing -- that "float" components under which they are used, isolating them laterally.[/TD]

[/TR]

[/TABLE]

[/TD]

[/TR]

[TR]

[TD=width: 100%, bgcolor: #FFFFFF]

[TABLE=width: 100%]

[TR]

[TD=width: 20%]Use[/TD]

[TD=width: 80%]Doug found them to work best under CD players; under solid-state electronics, "the changes range from, 'uh, maybe,' to 'yeah, perhaps.'"[/TD]

[/TR]

[/TABLE]

[/TD]

[/TR]

[TR]

[TD=width: 100%, bgcolor: #FFFFFF]

[TABLE=width: 100%]

[TR]

[TD=width: 20%]Value[/TD]

[TD=width: 80%]"Is there any competition for Daruma 3-IIs in their price bracket?" "Their $99-per-set price is quite reasonable."[/TD]

[/TR]

[/TABLE]

[/TD]

[/TR]

[/TABLE]

[/TD]

[/TR]

[/TABLE]

[YashN]

I agree with Barry that the Tungsten Carbide balls are 'a bridge too far'

 

But they're the most durable.

[YashN]

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

 

From what I've read so far:

 

Hardness: as hard as possible (if it dents or marks, the motion won't be smooth anymore, and may even introduce vibrations)

 

Smoothness: as smooth as possible

 

Curvature: as shallow as possible: you want the oscillations to be as long as possible.

 

Concentricity: eh?

[Daudio]

As ringing, is acrylic a good material? I gather we'd need hardest and smoothest, but also something which does not ring too much...

 

I haven't thought too much about ringing in acrylic. I don't see the bowl blocks as really transmitting vibrations.. well.. just between the base and the contact area with the ball. Don't think there any vibrations of consequence bouncing abound inside the square and disc shapped blocks.

 

And, we are talking mostly about low frequency vibrations here.

 

Also, I have heard many good, synergistic, things about acrylic, in conjunction with vinyl, at least. As a matter of fact, I have a Teres turntable that consists of a 3" thick platter, 2 1/2" thick base, and 2" thick arm board: all made of beautiful clear acrylic. Ringing is not a problem in this Very Well Regarded turntable

 

But, it could be an issue, I suppose... Anybody else have thoughts on this ?

 

 

Very encouraging. How large can you make these?

 

I generally use blocks that are either 2" square, or 2" in diameter. The bowl curvature is the same. I bought some larger ones (3" dia), but they had the same curvature in a deeper bowl, which doesn't work out very well in practice. They are very pretty in a piano black finish, through !

 

I use 4 of the square ones for my 95 lb electrostatics (panel, xformers, base). A three way support is just too unstable for a 6' tall speaker, and 4 spread out the load better. Not easy to set up on carpet

[Daudio]

From what I've read so far:

 

Hardness: as hard as possible (if it dents or marks, the motion won't be smooth anymore, and may even introduce vibrations)

 

Smoothness: as smooth as possible

 

Curvature: as shallow as possible: you want the oscillations to be as long as possible.

 

Concentricity: eh?

 

 

Yash,

 

Be careful here, any engineer will tell you that you can NOT optimize all variables in a system

 

 

Concentricity - the accuracy of the surface as compared to a theoretically perfect sphere. Applicable to the ball and the curvature of the bowl, both.

 

With cheap G25 bearing balls readily available, I'm not worrying about that. The Tungsten Carbide balls are definitely harder, but at their cost, your facing the law of diminishing returns. Yes, people say the sound better. If you want to pay the price, be my guest. I choose to be satisfied with G25 Stainless Steel.

 

 

 

Re: Tungsten Carbide

But they're the most durable.

 

Yeah, but what are you going to do with them ? Launch them from a rail gun ??

 

 

(actually I think they may be more brittle the SS)

[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

[YashN]

First a thought experiment. First take a large flat board that is infinitely rigid and has a top surface that is completely frictionless, then place your audio component on top, for now no wires connected to it.

 

Like Albert, I love gedankenexperiment Thanks for chiming in, John.

 

 

There are some problems with this isolator, if the board is not completely perpendicular to the local gravitational field, the gear just slides right off. Not very useful. THAT is why a shallow bowl is used in the devices mentioned, it provides a DC restoring force to prevent the gear from just sliding right off.

 

Yes, and it appears it also introduces a vertical component at this layer, as the ball rises up and down the cup's surface.

 

So for the best isolation you want the shallowest "bowl" you can get away with that still provides enough restoring force to keep your gear from sliding off.

 

I agree, so if we could get large bowls with slowly curving sides, that would be better. They would ring more though so we'd need to think of a way of damping this?

 

This is also important for cables, the stiffer your cables are they steeper your bowl will have to be so the cables don't fling the gear off the shelf. This may be a good incentive to go with very flexible cables, so you can go with shallow bowls, which work better.

 

Good point. Alternately, provide structural support for your cables.

 

The other aspect of the bowl is that it adds a resonance, in order for the isolator to be effective that resonance has to be significantly lower than the frequencies you are trying to isolate. The seismic noise in question seems to have most of its power in the single digit Hz range, so a resonant frequency below 1Hz will usually be quite effective.

 

Very interesting. I wonder what the resonant frequency of those small commercial pucks are.

 

As has been noted it can take a long time for this resonance to damp out. This is a GOOD thing, it means there is low resistance to the motion. Refer back to the original thought experiment, it is the lack of resistance that makes the isolation work, so anything you do to damp this very low frequency resonance defeats the whole purpose of the system.

 

I was having a hard time to accept this and was under the impression that we would need to damp the oscillations, but the gedankenexperiment cleared it for me.

 

Think carefully about using remote control for these functions, or using a rotary control rather than push buttons.

 

Or use devices with touch-screens (no push required).

 

That is it for the horizontal plane, the next post will cover vertical isolation, which is much harder to achieve.

 

Thanks a lot, John. Don't forget to chime in on tilt if you can!

[YashN]

Converting energy is not draining anything. It is converting it.

 

It's the same thing.

 

Law of Conservation of Energy: 100 units.

 

Case 1: A has 100 units

 

Case 2 (at moment t): A is coupled to B: Every time A has 100 units initially, B drains away 75 units =

A has 25 units, B has 75 units

 

B has drained A of 75 units. Total energy is still 100 units.

 

You can also say 75 units of A have been converted into the form of energy that B stores or expresses.

 

At time t+1, this process continues.

 

It is precisely the same thing, which is why I told you you don't like the term but the way people commonly use it is still OK.

 

I agree to disagree (already did).

[YashN]

I generally use blocks that are either 2" square, or 2" in diameter. The bowl curvature is the same. I bought some larger ones (3" dia), but they had the same curvature in a deeper bowl, which doesn't work out very well in practice. They are very pretty in a piano black finish, through !

 

I use 4 of the square ones for my 95 lb electrostatics (panel, xformers, base). A three way support is just too unstable for a 6' tall speaker, and 4 spread out the load better. Not easy to set up on carpet

 

The larger the dimension, the shallower it allows the curve to be. I think actual dimensions should vary with the components, to be able to make as large an equilateral triangle as Barry usually puts it.

 

I checked one of my amps, and for these, something with a diameter of 4" appears better.

 

I have read a lot of people hearing issues when using 4 contact points: they're hard to set in the exact same plane, and the result is that one contact point always then rattles, affecting the sound negatively.

[YashN]

Be careful here, any engineer will tell you that you can NOT optimize all variables in a system

 

I'm an engineer but I'm not telling you that

 

Concentricity - the accuracy of the surface as compared to a theoretically perfect sphere. Applicable to the ball and the curvature of the bowl, both.

 

Right. I figure high is better here, for the smoothest motion possible.

 

Yeah, but what are you going to do with them ? Launch them from a rail gun ??

 

Nope, I'd put them under my components, but with the less hard ones, if they dent or deform at any time, you'll need to buy new ones.

[YashN]

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.

 

Interesting approach, and more things I need to read upon. S waves do not travel through liquid, so why aren't we using a liquid to deal with the vertical isolation?

[YashN]

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

 

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?

[YashN]

A couple of commercial implementations for the cup-and-ball that I found:

 

 

[YashN]

Another cool technology I found while researching yesterday, using passive negative-stiffness vibration isolation:

 

[video=youtube;m9J8-YAMQKo]

[YashN]

This time, a test on a much bigger arrangement:

 

[YashN]

This turntable integrates Minus K's vibration isolation:

[YashN]

From MinusK's website (click on it to see the animation):

[Daudio]

The larger the dimension, the shallower it allows the curve to be. I think actual dimensions should vary with the components, to be able to make as large an equilateral triangle as Barry usually puts it.

 

In theory, yes, but I told you that the acrylic blocks I bought do not actually do that. Triangle ??

 

 

I checked one of my amps, and for these, something with a diameter of 4" appears better.

 

How did you arrive at that dimension ?

 

 

I have read a lot of people hearing issues when using 4 contact points: they're hard to set in the exact same plane, and the result is that one contact point always then rattles, affecting the sound negatively.

 

Yeah, maybe, but it doesn't mean much after your favorite speaker falls over and self destructs, leaving you without a good audio system !!!

 

 

I'd put them under my components, but with the less hard ones, if they dent or deform at any time, you'll need to buy new ones.

 

I don't think any of the parts we are talking about will undergo any deformation at all, in the modest service of our audio applications. With the possible exception of a severely overloaded acrylic block. Not to worry

[Daudio]

From MinusK's website (click on it to see the animation):

[ATTACH=CONFIG]19231[/ATTACH]

 

Since the Linn LP-12, and a few other audio encounters, I have disliked any kind of springs in things audio. They all have resonances, but some think thay can tame them. I have my doubts...

[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

[Jabs1542]

Another cool technology I found while researching yesterday, using passive negative-stiffness vibration isolation:

 

[video=youtube;m9J8-YAMQKo]

 

Very cool - have you priced these isolation platforms? Roughly $2K and up.