Vibration Air & Roller Bearings - Thanks to Barry & Warren

[Ralf11]

Thx for that info @jabbr

[Solstice380]

I was able to complete some resonance frequency spectra on an aluminum spike footer disc and a cone from 500 Hz - 1 MHz.  They are both ~1" diameter and the cone is ~1.25" tall.  Because they share a common major dimension the resonance patterns are similar.

 

The system excites the object under test with a swept sine wave drive through a piezoelectric transducer and receives the vibrations through 2 piezoelectric transducers.  The drive voltage on the piezoelectric transducer is constant across the frequency spectrum, but the transducers excitation output and input sensitivity vary over that range up to about 1.2X the lowest level.  That is well below the range of relative amplitudes in the disc and cone, so this is semi-quantitative data with respect to the amplitude.  Only the frequency is calibrated for this equipment setup.  I've noted the Max Peak Amplitude for the disc and cone in each of the frequency ranges.  The main point here is that all of these things ring like crazy, so I prefer to use isolation.

 

  

      

 

500 Hz - 300 kHz  Cone top, disc bottom

 

300 kHz - 600 kHz

 

600 kHz - 900 kHz

 

900 kHz - 1 MHz

[Ralf11]

Thx for doing this - it is excited on the edge of the disc, is that right?

[Solstice380]

7 hours ago, Ralf11 said:

Thx for doing this - it is excited on the edge of the disc, is that right?

 

The flat faces were in contact with the transducers.  If you change the excitation location you get about the same resonance frequencies but the relative amplitudes will change some.  

[Johnseye]

14 hours ago, Solstice380 said:

I was able to complete some resonance frequency spectra on an aluminum spike footer disc and a cone from 500 Hz - 1 MHz.  They are both ~1" diameter and the cone is ~1.25" tall.  Because they share a common major dimension the resonance patterns are similar.

 

The system excites the object under test with a swept sine wave drive through a piezoelectric transducer and receives the vibrations through 2 piezoelectric transducers.  The drive voltage on the piezoelectric transducer is constant across the frequency spectrum, but the transducers excitation output and input sensitivity vary over that range up to about 1.2X the lowest level.  That is well below the range of relative amplitudes in the disc and cone, so this is semi-quantitative data with respect to the amplitude.  Only the frequency is calibrated for this equipment setup.  I've noted the Max Peak Amplitude for the disc and cone in each of the frequency ranges.  The main point here is that all of these things ring like crazy, so I prefer to use isolation.

 

  

      

 

500 Hz - 300 kHz  Cone top, disc bottom

 

300 kHz - 600 kHz

 

600 kHz - 900 kHz

 

900 kHz - 1 MHz

 

Great information Solstice.  I assume similar results, but different frequencies would result from the Rollerblock ball and cup type approach depending on materials used for the ball and cup.  As we discussed at Axpona last weekend, the metal and other hard materials create a resonance vs dampening or isolating materials help to reduce the resonance.  Is that a correct assumption based on your results?

 

It was nice meeting you Saturday night.  I look forward to next time.

[Solstice380]

1 hour ago, Johnseye said:

As we discussed at Axpona last weekend, the metal and other hard materials create a resonance vs dampening or isolating materials help to reduce the resonance.

 

That’s right.   When I’m done with the current evaluation I’m doing I will be able to break down this fixture and do some more testing on fun stuff.  I have Si3N4 balls (1 1/8” diam), other cones, the glass lenses and the Herbie’s material and pucks.   I’ll do the free resonance spectra like above and, for the appropriate things like Herbie’s layered puck, I’ll do through excitation so we can see how well stuff works at knocking down the transmission of (bad) vibes.

 

Was definitely great to meet up with you guys - almost walked right past you in the bar!  Bet you all had fun on Sunday messing around.  We’ll coordinate in the future for one of my Chi-town trips to get together.  

[Johnseye]

8 minutes ago, Solstice380 said:

 

That’s right.   When I’m done with the current evaluation I’m doing I will be able to break down this fixture and do some more testing on fun stuff.  I have Si3N4 balls (1 1/8” diam), other cones, the glass lenses and the Herbie’s material and pucks.   I’ll do the free resonance spectra like above and, for the appropriate things like Herbie’s layered puck, I’ll do through excitation so we can see how well stuff works at knocking down the transmission of (bad) vibes.

 

Was definitely great to meet up with you guys - almost walked right past you in the bar!  Bet you all had fun on Sunday messing around.  We’ll coordinate in the future for one of my Chi-town trips to get together.  

 

Definitely let me know when you're in town again.  Looking forward to your further tests.

[Ralf11]

just to prevent this thread from falling into the uninterpretable pit like the red/blue pill one has...

 

Si3N4 is a silicon nitride material that is very hard (8+ mohs; harder than most metals) and dense - often used for orthopedics and aerospace - not sure about cars

 

I'm not sure how it would "ring"

[Ralf11]

Did you guys do any listening tests when you all were together?

[Johnseye]

9 minutes ago, Ralf11 said:

Did you guys do any listening tests when you all were together?

 

Yes, but unfortunately Solstice wasn't there.

 

 

[Ralf11]

ok - any results you guys came up with would be interesting (on whatever thread is appropriate)

[Johnseye]

2 minutes ago, Ralf11 said:

ok - any results you guys came up with would be interesting (on whatever thread is appropriate)

 

I detailed the results in the other thread linked.  They are interesting.  It was a good session.

[Ralf11]

This manf. really thinks they have something special:

 

 

 

Note the "matrix material" - is that inside the body of the block?

 

The small ones are $200 for 3 (for one piece of gear).

 

 

[jabbr]

Well let's see ... thats what I got made. Tungsten Carbide Grade 10 is common. I paid significantly less than $100 but they need to make a profit. The big cost is the setup for the CNC machine -- the model is simple. and then they need to be polished. My guy would want a bunch made to do a run -- need to buy a 2" cylinder of alu 7075.

[Ralf11]

you used solid blocks of Al, right?

 

his description makes it sound there is an acoustic foam (is there any other kind?) inside the block ("matrix")

 

could buy some and take a bandsaw to them...

 

I dunno if the cup for the balls to rest in is hardened but if it is just Al, then those are trivial to mill - even I could do it (except I no longer have access to a Bridgeport)

[jabbr]

10 hours ago, Ralf11 said:

you used solid blocks of Al, right?

 

his description makes it sound there is an acoustic foam (is there any other kind?) inside the block ("matrix")

 

could buy some and take a bandsaw to them...

 

I dunno if the cup for the balls to rest in is hardened but if it is just Al, then those are trivial to mill - even I could do it (except I no longer have access to a Bridgeport)

 

In Barry Diaments version the Alu bowls sit on 3/4” ply which itself is on inner tube — on top of the balls is a slab of “dull” marble

 

You can easily use a 2” piece of acoustic foam under the bowls. 

 

Dont get me wrong, the Symposium is a very fine product — just that I decided to put roller bowls under all my equipment and so would have been cost prohibitive. 

 

A few  years ago a bunch of folks wanted to do a group buy, but ppl spent endless time debating what to do and there was no agreement to get to a good volume. Ultimately one prototype didn’t have smooth enough polish but we found a local shop which cut a 2” dia sphere into a 2” x 3/4” disc producing a 1” dia bowl with a mirror polish.

[Ralf11]

polishing costs $$

 

I've seen Barry's setup on his website, but my question above is specifically about the Symposium product and the internal construction of the Al block.  If they did use a hollow "case" construction then I am curious as to why.  I could speculate, but our vibrationologist is better suited for that.

 

I also have to wonder... why doesn't a $5,000 pre-amp have these?  ARC does use extra-thick sheet metal, and got rid of the metal top cover in favor of polycarbonate - so they are not unaware of vibration effects.

[brucew]

6 hours ago, Ralf11 said:

...about the Symposium product and the internal construction of the Al block.  If they did use a hollow "case" construction then I am curious as to why. 

Symposium provides a description of what they did with the honeycomb matrix and what it accomplished for them. www.symposiumusa.com/rollerblocks.html

[Ralf11]

Their verbiage makes no sense to me

 

I found a pic of the bottom - the holes are uniform in layout, not random

https://www.google.com/search?q=bottom+of+rollerblocks&source=lnms&tbm=isch&sa=X&ved=0ahUKEwic-KbE7NHaAhWDKGMKHeW7BLEQ_AUICigB&biw=1231&bih=709#imgrc=B-gdvQfde-ZnbM:

 

[brucew]

19 minutes ago, Ralf11 said:

Their verbiage makes no sense to me

 

I found a pic of the bottom - the holes are uniform in layout, not random

Can't help you with the verbiage making sense.

 

The pic was at the top of the page I linked to. Their description does not say the holes are randomly spaced but are irreguarly shaped, which implies they are not drilled and either have multiple tapers or the holes are not straight.

 

The former is more likely, because I don't imagine one can 3D print aluminum. So the holes have to be accomplished by CNC tools accessing from outside. I guess I'm not really bothered about trying to figure it, let alone duplicate it. Whether it's worth doing or not... someone should buy a set a compare to an Ingress Engineering set or similar and see if there is a significant difference to warrant the cost/design.

[Johnseye]

29 minutes ago, brucew said:

Can't help you with the verbiage making sense.

 

The pic was at the top of the page I linked to. Their description does not say the holes are randomly spaced but are irreguarly shaped, which implies they are not drilled and either have multiple tapers or the holes are not straight.

 

The former is more likely, because I don't imagine one can 3D print aluminum. So the holes have to be accomplished by CNC tools accessing from outside. I guess I'm not really bothered about trying to figure it, let alone duplicate it. Whether it's worth doing or not... someone should buy a set a compare to an Ingress Engineering set or similar and see if there is a significant difference to warrant the cost/design.

 

You can 3D print metals now, but it's fairly recent by my understanding.  I don't think Rollerblocks were made with 3D.

[Solstice380]

Some of the Symposium verbiage is isn't really wrong.  They are talking about draining vibrations from the component through the ball/cup and into their "engineered" block.  The issue I still see with the whole ball and cup is that in no way will "slip" of the ball in the cup stop vibrations from traveling through into whatever they are mounted in.  The ball and cup have to be excited by the frequency of vibration or they will damp out those that are not close to one of the cup's or ball's resonances where it acts like an amplifier.  Oh, they sure don't know what they are talking about when they start comparing metal and ceramic.  There is not, to my knowledge, a direct correlation between thermal conductivity and transmission of sound.  In genera it may be true when you talk about wood-plastic-metal/ceramic but there are always exceptions.   Also, one of the least resonating (lowest Qm - Mechanical Quality Factor) metals is grey iron.  They use it for diesel engine blocks because it cuts down on the sound.  High thermal conductivity.  A problem that the Symposium Rollerblocks have is that their block still provides a connected, solid aluminum path between shelf and component.  If they had put the mushy stuff in layers between aluminum sheet in a stack I would think better of the idea. 

 

I spoke at length at AXPONA with the Bel Canto guy about the construction of their new Black series.  (I also noted that many mfrs are now using a single layer layout for their products;  bigger square/rectangular cases that are only good for 1 board high and that they are separating the power, input, processing and output sections.)  It's about 2" high and about 21" x 17" and weighs over 30 pounds.  When you rap your knuckle anywhere on it it just goes thud.  He said they use a constrained layer damping bottom plate and that some of the boards are "potted" in a semi-viscous material to isolate them from vibration.  He wouldn't let me open it up unless I bought it!  

 

 

[Solstice380]

14 hours ago, Johnseye said:

 

You can 3D print metals now, but it's fairly recent by my understanding.  I don't think Rollerblocks were made with 3D.

 

There are actually a couple Additive Manufactured components already certified for aerospace use.

[Solstice380]

Here are a couple shots of the rollerblocks that are under my Lampizator Golden Atlantic.  They mount the top cups to the chassis.  They are also smaller than we have been considering/using at about 1" diameter overall and probably a 1/2" or 5/8" ball.  I put the bottom cups on a Herbie's Big Fat (relative) Dots and they are on a 2" end grain Maple chopping block on more Herbie's then the LDF shelf!

 

 

   

 

A couple of the 3" Maple block amp stand with my McIntosh MC452 on it.  I sunk threaded inserts into the block and screwed in cones and they are on Herbie's brass/dBN pucks.  Cylindrical spikes would be better due to a simpler resonance spectrum, so I think I will try to find some.  Note in the cone vs. disc spectra posted before that there are a lot more high amplitude resonances for the cone. 

   

[Ralf11]

a stack would make sense to me

 

Thinking more about this... it seems like the first thing to do is to make some measurements to get an idea of actual vibrations in the (digital) listening environment. 

 

Then think about what components might be most susceptible, or most susceptible at various freq.s 

 

(tube gear would be my bet)