Clear acrylic speakers - any ideas? (Also, how should I make B3s open baffles?)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
To make a long story short, I'm in an art metal and glass class. Most of the class does stained glass; I prefer constructed sculpture. (Read: I get credit for building amplifiers, or so far, trying to. No, really!)

Anyway, someone donated some REALLY nice clear acrylic. There's about 10-20 sheets of 24x36" roughly 1/8" acrylic sheet. I've also found a trick for covering right angles with 90-degree sections of copper tube to make them look better.

Can anyone reccomend speakers for this stuff? OB's have an obvious advantage of being simple and flat, but I'm limited by size. (Multiple panels can be joined in the middle if necessary, though.) Normal speakers or MLTL's might work better still, but they'd need stuffing, which would not look good at all. Horns may also be a possibility; if made of flat panels, they'd be easy to build, but it might be possible to heat-shape the acrylic.

There's also the issue of reinforcement. There's the good old-fashioned I-shaped joiner made of two pieces of flat metal with a rod inbetween them; although normally used for glass, some rivets and/or epoxy will let them work just fine.

Finally, can I get any better acoustic properties for using glass in certian areas?


So, what do I build? I'm within a short distance of Madisound, so I can get the B3S's for 10$ ea. (no shipping!), and I have the following:

2x 40-1041 RadioShack FR drivers. They look like 40-1354's, except less efficient.

2x RFT slightly oblong FR drivers. I've yet to test these.

2x Tesla drivers. These are weirdly long and narrow drivers, roughly 1' long by 3" wide. I've yet to test these, but they have an awesome vintage look.

1x Foster 10T3 driver from a Solo 103 speaker. This thing sounds like crap, and is nothing like a FE103. However, I feel bad not doing anything with it.

SI Soundpads. A friend of mine, after experiencing the excellence of the T-amp, bought a pair. However, he has no idea what to do with them. I was thinking some simple boxes, or perhps horns; these need'nt be truly hi-fi, as they're liable to just be rear-channel or workshop speakers anyway.

Hi-Vi B3S. I just ordered a pair. Obviously, some woofers are necessary.

I'd also like to use some of the following drivers if possible. (3$
each! WOOT!)

http://www.madisound.com/cgi-bin/index.cgi?cart_id=8893731.23023&pid=1902

Obviously, they're going to need a tweeter, but they might make nice OB speakers.

If I can get a good, working design based around the RadioShack, Foster, or Hi-vi drivers, that would be great. (I have a venue for sales).
 
Hi Spasticteapot

There are others in this forum (actually, almost anyone in this forum) who can give you better advice than I about speaker design issues.

I just wanted to say that, since one of the advantages of acrylic sheet is that it's heat-formable, if you go to boxes (as opposed to OBs) you might consider a curved shape. A truncated cone would be almost as easy as a cylinder to bend, would have good acoustic properties, would be much stiffer than a flat panel of 1/8" acrylic, and could look very nice indeed.

Sounds like a great project - I look forward to the pictures.

Aengus
 
Clear speakers?
 

Attachments

  • fhspeaker-full-size-2.jpg
    fhspeaker-full-size-2.jpg
    49.8 KB · Views: 894
Its called a form mold The product is heated to a given temperature and either vacumme molded or press fit in which it tales a male mold and a female mold. Another method is a heat draw in which a center form is made the product is heated to a given temperature and the center is drawn into the shape thats required.

ron
 
There was once a loudspeaker from I believe Acoustic Research that looked a little like a b@stard child of a Dahlquist DQ-10 and the main panel of the Infinity Beta (..without the sloped/tilted profile of either speaker). In fact it looked considerably better then either (taken alone).

Note my poor "drawing" below. The front is covered with "grill" cloth like the Quad electrostats:

http://www.quad-hifi.co.uk/

However, with the exception of the vertical column of drivers and the surrounding frame, the extension panels from the vertical column can be seen through - because they should be from acrylic. This is not dissimilar from the "void" space at the top of the Infinity Beta main panel as viewed here:

http://www.arduman.com/aa/Sayfalar/rob/speak.htm

The bass bin should contain a 12 inch driver.

The basics are:

4 of the Foster "woofers" (not the fullrange drivers) open baffle with a 4th order LR around 3 kHz.

1 HiVi K1 with a first order filter around 3 kHz.. (this will limit spl somewhat and increase distortion). Driver rotated to an inverted position for aiming the polar "tilt" up. This driver should be about 2 feet from the floor. Measured by member Zaph here:

http://www.zaphaudio.com/tweetermishmash/

1 12 inch Powered woofer/subwoofer "box/bin". DVC connected in parallel to an MCM sub plate amp. (prob. bass reflex - need to do some sim's though)
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=295-490
http://www.mcminone.com/product.asp?catalog_name=MCMProducts&product_id=50-6272

In effect the top OB column is two way phase coherent design with a dipole midrange section. With the tweeter below the midbass line (and considerably lower than the listener) but with an upward veritical radiation angle of 15 degrees - gives a good stand-up/sit down even high freq. sp-level (..in other words when you stand-up from the listening position the treble won't dramatically reduced in level). The bottom section is a powered sub.

The top portion will need to be modeled for the crossover and baffle effects. The bottom "sub" will need box modeling to give the desired character.

Cost-wise the loudspeaker can be fairly in-expensive for such a *fullrange* design (particularly because its also powered in the bass).
 

Attachments

  • cheap speaker.gif
    cheap speaker.gif
    6.8 KB · Views: 634
Thank you, Scottmoose. I found them, and the thread on them.

ronc: Have you done such moulding of similarly large parts? I can see that one could make a form from, say, plaster of paris, but I doubt my ability to vacuum-mould something that size in the garage. Tempting idea, though.

Aengus
 
ScottG said:
There was once a loudspeaker from I believe Acoustic Research that looked a little like a b@stard child of a Dahlquist DQ-10 and the main panel of the Infinity Beta (..without the sloped/tilted profile of either speaker). In fact it looked considerably better then either (taken alone).

Note my poor "drawing" below. The front is covered with "grill" cloth like the Quad electrostats:

http://www.quad-hifi.co.uk/

However, with the exception of the vertical column of drivers and the surrounding frame, the extension panels from the vertical column can be seen through - because they should be from acrylic. This is not dissimilar from the "void" space at the top of the Infinity Beta main panel as viewed here:

http://www.arduman.com/aa/Sayfalar/rob/speak.htm

The bass bin should contain a 12 inch driver.

The basics are:

4 of the Foster "woofers" (not the fullrange drivers) open baffle with a 4th order LR around 3 kHz.

1 HiVi K1 with a first order filter around 3 kHz.. (this will limit spl somewhat and increase distortion). Driver rotated to an inverted position for aiming the polar "tilt" up. This driver should be about 2 feet from the floor. Measured by member Zaph here:

http://www.zaphaudio.com/tweetermishmash/

1 12 inch Powered woofer/subwoofer "box/bin". DVC connected in parallel to an MCM sub plate amp. (prob. bass reflex - need to do some sim's though)
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=295-490
http://www.mcminone.com/product.asp?catalog_name=MCMProducts&product_id=50-6272

In effect the top OB column is two way phase coherent design with a dipole midrange section. With the tweeter below the midbass line (and considerably lower than the listener) but with an upward veritical radiation angle of 15 degrees - gives a good stand-up/sit down even high freq. sp-level (..in other words when you stand-up from the listening position the treble won't dramatically reduced in level). The bottom section is a powered sub.

The top portion will need to be modeled for the crossover and baffle effects. The bottom "sub" will need box modeling to give the desired character.

Cost-wise the loudspeaker can be fairly in-expensive for such a *fullrange* design (particularly because its also powered in the bass).

Woah...


Those look AWESOME!

Prices:

8x Fostex Unnamed Woofers: 24$
2x Hi-Vi tweeter: 18$

Total cost of drivers without subwoofer: 32$.

So, how much are you going to ransom the crossover specs for? I'm hardly opposed to bi-amping; would two LM3875 gainclone modules, one for the tweeter and one for the woofers (wired series-parallel) do the trick, or should I use one for each set of woofers wired in parallel?

Also, I now must find some cheap subwoofer drivers.
 
Spasticteapot said:


Woah...


Those look AWESOME!

Prices:

8x Fostex Unnamed Woofers: 24$
2x Hi-Vi tweeter: 18$

Total cost of drivers without subwoofer: 32$.

So, how much are you going to ransom the crossover specs for? I'm hardly opposed to bi-amping; would two LM3875 gainclone modules, one for the tweeter and one for the woofers (wired series-parallel) do the trick, or should I use one for each set of woofers wired in parallel?

Also, I now must find some cheap subwoofer drivers.

Umm, the "drawing" looks pretty retched IMO - but it COULD look Awesome. :D (..and in fact those Acoustic Research speakers DID look awesome - and sorry, I couldn't find a pic of them anywhere.)

For the crossover & box - thats something for you to model. I'd suggest getting familiar with some of the free programs here:

http://www.pvconsultants.com/audio/frdgroup.htm

and this program to show open baffle and diffraction effects:
http://www.tolvan.com/edge/

All are freeware.

Then when you have a better idea of modeling and a better grasp of the "Edge's" interface you can play with Basta for 90 days free (and pay for it if you want to keep it):
http://www.tolvan.com/basta/

The 90 days free for Basta should give you plenty of time to get the model that you want provided that the drivers are similar to what was measured.

If you don't want to "putz" with freeware and just want to spend the cash to get the right software at a good price then I'd suggest SoundEasy (which has both measurement and advanced modeling capability - and with both features integrated it makes it a LOT more convienient and ultimatly more accurate).
http://www.partsexpress.com/webpage.cfm?webpage_id=3&CATID=48&ObjectGroup_ID=360

Note you'll still need the correct hardware to take measurements.. and Zaph goes into that on his site.

Its obviously a better solution - but then it actually costs money.. :D

btw.. you'll have difficulty finding a less expensive subwoofer OR sub amplifer. Now IF you go "active" for everything (modeled in Basta or SoundEasy) then you don't need a sub plate amp and you can just use something like multiple gain clones or digital amps (1 channel of amplification for each speakers midrange section, tweeter section, and woofer/subwoofer section). Additionally, you can use a LR 4th order for the tweeter as well and not only lower distortion but also lower the crossover point for the tweeter to about 2 kHz.

Anyway - up to you, and good luck! :)
 
I looked at MCM and Parts Express for even less expensive woofers.

There weren't any at MCM that I'd recomend for less cash in the 10-12 inch driver categories - all were cheap.

Parts Express came down to various dayton models (the goldwood stuff is cr@p):

$30 price range:
12 inch:
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=295-320
10 inch:
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=295-485
$23 price:
http://www.partsexpress.com/pe/showdetl.cfm?&DID=7&Partnumber=295-315


Notice the weight difference between the drivers generally, and between the DVC and non DVC drivers. Most of the difference between the DVC vs. non DVC drivers can be attributed to the magnet shielding and the additional VC, BUT not all (..and this implies a driver that it is built a bit better). Additionally, the DVC woofers also have greater xmax.
 
ScottG said:


Umm, the "drawing" looks pretty retched IMO - but it COULD look Awesome. :D (..and in fact those Acoustic Research speakers DID look awesome - and sorry, I couldn't find a pic of them anywhere.)

You have


That's the plan - triple gainclones!

10$/ea. + 15$ transformer + 10$ worth of PSU. (In this case, a snubberized chipamp.com-style PSU with 24,000uf of capacitors.)

If I got the math right:
Efficiency for each Foster driver: 84dB/w
Impeadance: 8 ohms in series/parallel
Power each, wired series/parallel for a total of 8 ohms: about 4W (assuming 16w gainclone output).
SPL from each: About 90db/1m
SPL total: 110db/1m.

Efficiency for tweeter: 92db/w
Impeadance: 5 ohms
Power, assuming identical gain into 5 ohms: About 25w
SPL: 107 db/1w, give or take a bit.

Efficiency for sub: 94db/w
Impeadance: 8 ohms
Power, assuming identical gain: 16w
SPL: 96db.

This all seems to work very nicely. I have no idea how on earth those modelling programs work, though.


It's probbably cheaper than a standard crossver, anyway.

EDIT: Page to soon apear in primary loudspeaker forum.
 
Play with the Edge first - I absolutely LOVE the interface to that! (..the interface is SO intuitive that it only takes about 15 minutes to learn most of the features the program offers - and they are VERY sophisticated features).

The Basta doesn't look terribly different, it is however more complex (by neccesity). (Look through the documentation as well for the Basta before installing it.)

Also, don't expect to get anywhere near your max rating for the dipole drivers for their entire bandwidth. If you can get to 105 db at lower freq.s with it and a decent baffle then be very happy. (..and 105 is painfully loud IMO.)

..and yes, an active crossover can be a LOT cheaper in that the parts values are much smaller due to much higher impeadances - smaller parts ='s MUCH lower parts cost (..even for premium parts).

The most costly thing will likely be Power transformer for the amplifiers. Consider building 1 good LARGE power supply for all 3 amplifiers AND the crossover. (i.e. 1 power supply per Loudspeaker to add in noise seperation, physical seperation, and reasonable cost.)

Note that when it comes to Caps - consider Clarity Cap SA grade (..which prob. represent the best value in quality vs. price). See member Geenius's (Tony Gee) "Cap Test" evaluation page and then compare pricing:

http://www.humblehomemadehifi.com/

http://www.e-speakers.com/products/claritycap.htm

Additional suggestions:

1. Use only LR 4th order crossovers (..it and 1st order don't suffer from linear distortion). You can still use a 1st order for the tweeter if you want to though like I originally suggested (..there are some benefits and some detriments to doing so).
2. IF you use a LR4th order for the tweeter then be sure to model it with different placements in the "Line" (..rather than just at the bottom).
3. Definitly get as much help as you can from others - If you are playing with Basta then Svante might help you out (..it would certainly be good advertising).

Again, Good Luck!
 
ScottG said:
Play with the Edge first - I absolutely LOVE the interface to that! (..the interface is SO intuitive that it only takes about 15 minutes to learn most of the features the program offers - and they are VERY sophisticated features).

The Basta doesn't look terribly different, it is however more complex (by neccesity). (Look through the documentation as well for the Basta before installing it.)

Also, don't expect to get anywhere near your max rating for the dipole drivers for their entire bandwidth. If you can get to 105 db at lower freq.s with it and a decent baffle then be very happy. (..and 105 is painfully loud IMO.)

..and yes, an active crossover can be a LOT cheaper in that the parts values are much smaller due to much higher impeadances - smaller parts ='s MUCH lower parts cost (..even for premium parts).

The most costly thing will likely be Power transformer for the amplifiers. Consider building 1 good LARGE power supply for all 3 amplifiers AND the crossover. (i.e. 1 power supply per Loudspeaker to add in noise seperation, physical seperation, and reasonable cost.)

Note that when it comes to Caps - consider Clarity Cap SA grade (..which prob. represent the best value in quality vs. price). See member Geenius's (Tony Gee) "Cap Test" evaluation page and then compare pricing:

http://www.humblehomemadehifi.com/

http://www.e-speakers.com/products/claritycap.htm

Additional suggestions:

1. Use only LR 4th order crossovers (..it and 1st order don't suffer from linear distortion). You can still use a 1st order for the tweeter if you want to though like I originally suggested (..there are some benefits and some detriments to doing so).
2. IF you use a LR4th order for the tweeter then be sure to model it with different placements in the "Line" (..rather than just at the bottom).
3. Definitly get as much help as you can from others - If you are playing with Basta then Svante might help you out (..it would certainly be good advertising).

Again, Good Luck!

I have two 300+ VA toroids, so I think I'm set. Cost me all of 30$!

Can you explain the differences between crossovers work? This is more or less my first attempt (I'm currently tinkering with a simple fullrange and ABS design), and I was hoping for something simpler.

That said, by the time I finish modelling them, I should have no trouble building them. (Copper and acrylic FTW!)
 
Spasticteapot said:


I have two 300+ VA toroids, so I think I'm set. Cost me all of 30$!

Can you explain the differences between crossovers work? This is more or less my first attempt (I'm currently tinkering with a simple fullrange and ABS design), and I was hoping for something simpler.

That said, by the time I finish modelling them, I should have no trouble building them. (Copper and acrylic FTW!)

Sounds good for the power supply! (..may have a little less power on-tap, but in reality it shouldn't be a problem.) ;)

With a LR 4th order for the highpass on the tweeter around 2 kHz you'll be able to play the thing louder with less distortion. It however will not have any polar tilt (where the vertical dispersion effectivly acts as if the driver were aimed down or up by 15 degrees, but instead is on flat baffle). Now because their isn't any polar tilt you'll likely want the driver higher on the baffle. Additionally, with a LR 4th order for the midranges at 2 kHz you can seperate the "Line" with the tweeter in the middle with out substantial "combing". (i.e. a MMTMM design for the "Line")

Now a 1st order highpass tweeter will give you better off-axis dispersion (horizonatally and vertically for the stand-up/sit-down thing I talked about previously), a bit better imaging (because it goes lower in freq.), a little bit better tonal balance (..again because it goes lower in freq.), considerably more distortion (..it goes lower in freq.), and considerably lower MAX spl's (..it goes lower in freq.).

Of course there is nothing (beyond some modest additional expense) that says you can't try both. :)
 
ScottG said:


Sounds good for the power supply! (..may have a little less power on-tap, but in reality it shouldn't be a problem.) ;)

With a LR 4th order for the highpass on the tweeter around 2 kHz you'll be able to play the thing louder with less distortion. It however will not have any polar tilt (where the vertical dispersion effectivly acts as if the driver were aimed down or up by 15 degrees, but instead is on flat baffle). Now because their isn't any polar tilt you'll likely want the driver higher on the baffle. Additionally, with a LR 4th order for the midranges at 2 kHz you can seperate the "Line" with the tweeter in the middle with out substantial "combing". (i.e. a MMTMM design for the "Line")

Now a 1st order highpass tweeter will give you better off-axis dispersion (horizonatally and vertically for the stand-up/sit-down thing I talked about previously), a bit better imaging (because it goes lower in freq.), a little bit better tonal balance (..again because it goes lower in freq.), considerably more distortion (..it goes lower in freq.), and considerably lower MAX spl's (..it goes lower in freq.).

Of course there is nothing (beyond some modest additional expense) that says you can't try both. :)


Dumb question, but could'nt a capacitor be used to block low frequencies, reducing distortion?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.