A cabinet out of thin material

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have never been a fan of extreme bracing schemes. I understand the logic of tying all of the panels together with cross bracing, but I have always wondered if the result makes an audible difference. I also have a problem with obtaining decent sheet goods out here in the hinterlands.

Second issue first. I can get decent ¾” cabinet grade plywood. 5-ply with decent birch/oak/maple veneer. What I would like to get is real Finnish 13-ply 18mm plywood, but it simply isn’t available to me in small quantities. A skid? Yes. Half a dozen sheets? No. What is available is “TigerPly”. 11-ply kiri core with a microscopically thin birch veneer. The veneer chips and tears with any machining. See pictures below. My other option is 12mm Russian birch in 5’x5’ panels. This stuff is pretty good and I have used it successfully in smaller projects.

This leads to the first issue. Can I build a fairly large cabinet out of the 12mm Russian birch without suffering unacceptable panel vibrations? I designed a MLTL for the Alpair 12P and here is the build story. I used 12mm “TigerPly” because I had a couple of sheets in the shop. I have a brace above and below the driver, 2”x12mm stiffeners from the second brace to the bottom of the cabinet, a doubler behind the driver, and the top panel is doubled (24mm). It is a happy circumstance that the driver rabbet is 12mm, so rather than glue up a double panel and route out the rabbet, I just cut the appropriate holes and glued the doubler behind the baffle. I lined the entire cabinet with 5/8” fiberglass and screwed on the fourth side. The cabinet, by the way, is 9”x7”x40” i.d.

The driver only have ~20 hours on them, so “how does it sound?” is not yet an issue. I did want to know what kind of noise was passing through the sides of the box. I made and RTA of the driver close mic’d and used a 5.5ms gate. Then I made RTA’s positioned ¼” from the side at the driver level and the back both at the driver level and half way between the second brace and the bottom of the cabinet. Note that I did not change the amp volume between measurements. I did scale all of the plots by the same number so that the driver plot would be somewhere near 0dB. Build pictures and the RTA’s attached.

Enjoy.
Bob

( Upper black=driver Lower black=back Blue=lower back Red=side)
 

Attachments

  • Panel with brace.jpg
    Panel with brace.jpg
    84.4 KB · Views: 718
  • Three sides.jpg
    Three sides.jpg
    82 KB · Views: 705
  • Raw cabinet.jpg
    Raw cabinet.jpg
    79.5 KB · Views: 693
  • M12-A12_Panel_Test.jpg
    M12-A12_Panel_Test.jpg
    84.5 KB · Views: 692
The BBC research into this issue is worth looking at, although their approach was the opposite of a stiff cabinet. However they found that the difference between 9mm and 18mm wall cabinets at certain frequencies was negligible.

You probably already came across it, but the main research paper was this one

http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf

Hope that is of some use.
 
without meaning to be cheeky,

could you possibly do another similar cabinet study but use constrained layer construction with very thin sheets of ply (6mm?), with the 'adhesive' between the two layers consisting of 'green glue' or other similar adhesive with sound dampening properties?

NOISEPROOF YOUR LIFE® | Green Glue

i tried to measure this myself but couldn't actually measure any difference between any panels i used (mdf, osd, various ply). i assume this was just general feklessness.
 
Bob, I see what you mean by thin veneer, that's some nasty chip/tear-out. I'm lucky to have easy access to 13-ply 18mm BB, with a decent veneer on it. Only in 5'x5' though; 4'x8' would be better for certain cabinet dimensions.

gafhenderson: CLD might be interesting to test here. I wonder how construction adhesive (PL Premium) would fare. It cures to a hard-rubbery consistency. It works well to secure bracing pieces along with screws, much like Bob's long and slim brace. Maybe a vacuum press would be useful to produce such panels.

IG
 
Last edited:
Bob, I see what you mean by thin veneer, that's some nasty chip/tear-out. I'm lucky to have easy access to 13-ply 18mm BB, with a decent veneer on it. Only in 5'x5' though; 4'x8' would be better for certain cabinet dimensions.

gafhenderson: CLD might be interesting to test here. I wonder how construction adhesive (PL Premium) would fare. It cures to a hard-rubbery consistency. It works well to secure bracing pieces along with screws, much like Bob's long and slim brace. Maybe a vacuum press would be useful to produce such panels.

IG


Where exactly in Canuckia are you located? Without prejudging your resourcefulness, there could be some supply chains of better grade 4x8 plywood nearer than you think - although it may mean tracking down distributors to commercial millwork and cabinet trade, rather than retail dealers.

Funny that you should mention CLD & vacuum bagging in the same post - this is a process we frequently employ to fabricate curved panels using "rubberply" (not kerfcore) and yellow PVA glue. While this method approaches the CLD goal from the opposite direction (ie. the material is very soft, as each layer pf "ply" has 2 interfaces of wood to "rubber", and the glue provides the stiffening) 3 or 4 layers will make for a surprising rigid panel.

They will of course need a structural support framework, and the quality of surface makes the cheapest Meranti or Kiri plywood look like A1 architectural grade maple veneer Apple ply. In commercial work the curved panels will generally be finished with a plastic laminate or phenolic backed real wood veneer.

I'm working on something with this technique myself, but it'll take a while to finish with everything else that I've got going on at present.
 
Where exactly in Canuckia are you located? Without prejudging your resourcefulness, there could be some supply chains of better grade 4x8 plywood nearer than you think - although it may mean tracking down distributors to commercial millwork and cabinet trade, rather than retail dealers.

Funny that you should mention CLD & vacuum bagging in the same post - this is a process we frequently employ to fabricate curved panels using "rubberply" (not kerfcore) and yellow PVA glue. While this method approaches the CLD goal from the opposite direction (ie. the material is very soft, as each layer pf "ply" has 2 interfaces of wood to "rubber", and the glue provides the stiffening) 3 or 4 layers will make for a surprising rigid panel.

They will of course need a structural support framework, and the quality of surface makes the cheapest Meranti or Kiri plywood look like A1 architectural grade maple veneer Apple ply. In commercial work the curved panels will generally be finished with a plastic laminate or phenolic backed real wood veneer.

I'm working on something with this technique myself, but it'll take a while to finish with everything else that I've got going on at present.

Québec. I can have that 5'x5' easy local. The big chain stores (Home Depot, Rona, Réno-Depot) don't have 4'x8' of the good stuff last time I checked, though they have 2'x4' and 2'x2' at somewhat ridiculous prices. Maybe on special order? Probably could get some via specialized distributors, but I don't have storage space for lumber, so this would go down to convenience of obtention. 5'x5' is usually fine though, just that sometimes it'd be easier to arrange certain cut layouts on 4'x8' depending on cabinet size, such as one I'm currently working on. :)

IG
 
Last edited:
gafhenderson ,

Not cheeky a all, but that product may be a little hard to source on this sind of the pond. I'll look at its ingredients, as far a one can, and see what I can get here.

I product that I am very interested in is a underlayment for laminate floors. It really deadens the sound of anything dropped on the floor. It looks like a poly ethylene or poly polypropylene foam. Trouble is, those materials are very difficult to bond to anything else. We'll see.

Bob
 
I have put small 1/8 inch dia dots of plumbers putty (every inch or so spacing) between two layers of foam core and compress the sheets to flatten he dots and used that for a baffle. It becomes almost dead sounding compared to plain foam core. There is a lot to be had with layering composites. I have found some interesting materials that house computer hard disk platters: stainless steel sheets with a what appears to be a sorbothane like polymer sandwiched in between. When you tap on it, a dull thud, very little transmitted sound.

I have tested the material called "Noene" for shock absorption and it is an amazing material. It is used primarily as a thin liner in sport shoes and is quite expensive. It is special - if you place it on a hard surface like a granite counter top and drop a steel ball on it you can test the coef of restitution. The ball will bounce quite high (90%) if it hits the bare granite, and maybe 10% high bounce with 2 mm sorbothane (pretty good stuff). With 1 mm Nonene, the ball does not bounce at all - like it hit an adhesive surface - all energy is dissipated. I bet this Noene sandwiched between sheets of metal or plywood would make a very aborbent cabinet.
 
Last edited:
Budget permitting, another one to consider could be raw cork used as underlay for engineered wood & laminate flooring.

This product comes in rolls, which is a PITA to work with; the flat "tiles" of 2x3ft, 6mm (1/4") that would be much easier to work with -they're flat to start with and can be laminated with white wood glue or contact cement - retails hereabouts for $8 per sheet.

A concern with multiple thin layers of such different materials and compliant adhesives, could be structural integrity of joinery if fabricating frameless cases.
 
Fig 12 is really the key for the BBC paper. It claim that above 400 Hz resonances has to be supressed by 30 dB to be below audiability threshold. Below 400 Hz less is needed, by about 9 dB per octave. So thining the walls will bring down the resonances below 400 Hz.

The BBC paper was published when LPs was the primary sound source. The digital recording and playback technique of today may changed these finding or not.

Harbeth still use the thinwalls and have active fora on their website
Ask the designer a technical question ...
 
The BBC research into this issue is worth looking at, although their approach was the opposite of a stiff cabinet. However they found that the difference between 9mm and 18mm wall cabinets at certain frequencies was negligible.

You probably already came across it, but the main research paper was this one

http://downloads.bbc.co.uk/rd/pubs/reports/1977-03.pdf

Hope that is of some use.

very old but trusty study, and the reason I dont brace to uber OCD levels. Use a good material and then use less of it.
Good ply comes in 2 kinds, birch 13ply or phenolic birch.
3 or 5 ply is not good. Id rather use MDF, and play my part in recycling

digital audio has no requirements over those in analogue, as far as im concerned.
 
Last edited:
The BBC paper was published when LPs was the primary sound source. The digital recording and playback technique of today may changed these finding or not.

Harbeth still use the thinwalls and have active fora on their website
Ask the designer a technical question ...

Bear in mind that the research was conducted at the BBC, so their primary sound source was either live music and voice or high quality analogue tape recordings. The research was done at their Kingswood Warren research centre, which had an anechoic chamber. The BBC research in that paper was specifically for loudspeakers used in live audio monitoring.

The only constraint I'd put on the paper is that it pre-dated MDF, which is less stiff than a similar thickness plywood. From browsing the Harbeth forum over the years, I believe the designer stated they no longer use ply. I suspect they use MDF as that would be even easier to push the fundamental resonance down into less audible territory when properly damped.

A final point about the paper - a large number of manufacturers over the years seem to have extracted what they want from it and ignored other bits of information. So we had 18mm chipboard enclosures 'damped' with pads - totally missing the point of the thin-wall technique. I suspect it was for manufacturing expediency, more than anything else. Something we don't have to worry about.

Stirling Broadcast in the UK have gone back to the original BBC findings and their version of the venerable LS3/5a speaker is made of 9mm ply, rather than a mix of 12mm and 9mm used elsewhere.

Apologies if this is derailing Bob's thread.
 
Last edited:
The paper mention the importance of using hardwood fillets not soft wood. If I look at the Spenor BC1 the wood looks like pine to me
Spendor BC1

I also used pine for my BC1 inspired creation
http://www.diyaudio.com/forums/multi-way/228670-one-bc-homage-classic-vintage-design.html

There was a german online article about sound transmission through varius variants of loudspeaker walls but it seems to be gone.
M Colloms High Performance Loudspeaker book show how damping material and wall thickness has to be matched (page 175 on the second issue.)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.