Poor Man's Ardent inspired design and build

[wkhanna]

+1

:agree:

[Renron]

Have a fantastic trip and enjoy your time away from us.
Aloha!
Talk to you when you get back. I've got questions about which beginner DLSR to get too.
Have Fun!
..............................Your not taking E.T. are you?..........................

[xandresen]

Y'all's choice. Jon Marsh has the right idea with the multiple LBL layers. Stiffer and much better damped. Baltic Birch Plywood and MDF have very little internal damping. Even though Baltic Birch Plywood has twice the stiffness of MDF, the baffle cumulative spectrum decay plots for the two look very similar. The Baltic Birch Plywood does not settle any quicker either. The additional hardness of the LBL will also help.

Material............................Stiffness (Modulus of E)...Damping Factor
Laminated Bamboo Lumber....1.5 Mpsi.............................0.2 (good damping)
Baltic Birch Plywood..............1.1 Mpsi.............................0.04 (little damping)
MDF....................................0.45 Mpsi...........................0.02 (little damping)

While I would like to believe there is a stiff but highly damped cellulose/resin composite we could use for speaker cabinets, my many years as an engineer lead me to question the damping coefficient for the lbl.

First it doesn't specify if it is single layer or ply. But more importantly all the other wood products except soft pine are in the range of 0.02 to 0.04. I would expect the resin to be similar to Baltic Birch, stiff but not damping. My guess is that the 0.2 damping value for lbl is a typo from 0.02.

I would suggest that someone expecting high damping in lbl make some test panels and compare to MDF or BB with a knuckle rap. The large difference in damping should be evident if it exists.

Constrained layer damping
I have made constrained layer test panels from MDF and BB with poor results. What has worked well is using Home Depot's cheapest bathroom tile (available in 9.5" x 13" x 0.25") as the constaining layer with a Parts Express damping layer. The test panels use 1/2 or 3/4 MDF or BB layers. The tile adds stiffness. They are completely dead to a knuckle rap. However the 1.7 cubic ft cabinets I made this way are good but not so dead.

An alternative to massive cabinet walls and weight.

[Renron]

xandresen,
That's why I asked about "Hardi backer" concrete board. Very stiff and almost zero resonance with a knuckle rap. Any experience testing concrete boards?

[benthe8track]

Again, the advantage of the bamboo is the high modulus (when compared to other typical speaker building materials). If you are concerned about damping just add mass (Hooke's law) like Jon did with lead...best of both worlds.

[Winter]

First it doesn't specify if it is single layer or ply. But more importantly all the other wood products except soft pine are in the range of 0.02 to 0.04. I would expect the resin to be similar to Baltic Birch, stiff but not damping. My guess is that the 0.2 damping value for lbl is a typo from 0.02.

No typo. Damping factor for LBL is 0.2. Bamboo is a grass, not wood.

http://audioqualia.prophpbb.com/topic20.html

http://forum.ascendacoustics.com/showthread.php?2597-Sierra-The-Cabinet

[---k---]

You know, my garage time is an escape from work. You guys are dragging work into my garage and ruining it! Stop!

[Winter]

Constrained layer damping
I have made constrained layer test panels from MDF and BB with poor results. What has worked well is using Home Depot's cheapest bathroom tile (available in 9.5" x 13" x 0.25") as the constaining layer with a Parts Express damping layer. The test panels use 1/2 or 3/4 MDF or BB layers. The tile adds stiffness. They are completely dead to a knuckle rap. However the 1.7 cubic ft cabinets I made this way are good but not so dead.

Constrained layer damping is a tuned system. Results depend significantly on several factors. Miss one and damping can be greatly reduced.
1. Damping layer must be viscoelastic, not just elastic, with good damping factor. Material damping curve is temperature and frequency dependent. Look for good damping factor, above 0.3 up to around 1.0, near 20 degrees C temperature, with peak of curve in the 200 to 600 Hz range. Damping works in shear.
2. Constrained layer should be stiff. If same material, 1/2 to 1 times thickness of base layer.
3. The optimum damping layer thickness dependent on the shear modulus of the damping layer , stiffness of the constrained layer and panel composite damping effective frequency range. Usual range is slightly less than 1mm to 3mm.

Constrained layer damping with stiff segments is effective too.

https://web.archive.org/web/20160316130404/http://www.earsc.com/HOME/engineering/TechnicalWhitePapers/Contrain/index.asp?SID=60

https://web.archive.org/web/20160412102702/http://www.earsc.com/HOME/engineering/MaterialNomograms/C-1105Reduced/index.asp?SID=52

[Winter]

Again, the advantage of the bamboo is the high modulus (when compared to other typical speaker building materials). If you are concerned about damping just add mass (Hooke's law) like Jon did with lead...best of both worlds.

Both the high modulus and the good damping are assets of LBL. Just adding mass, as with attaching lead directly to a wall, may be fine for a room wall with a very low resonance in the 5 to 20 Hz range, but is not optimum for a loudspeaker wall. For a loudspeaker wall when mainly weight is added as with lead, some resonances go up and some go down.

[benthe8track]

You know, my garage time is an escape from work. You guys are dragging work into my garage and ruining it! Stop!

Ha! At work we call it, "bringing out the math" and it's met with a lot of groans.

Both the high modulus and the good damping are assets of LBL. Just adding mass, as with attaching lead directly to a wall, may be fine for a room wall with a very low resonance in the 5 to 20 Hz range, but is not optimum for a loudspeaker wall. For a loudspeaker wall when mainly weight is added as with lead, some resonances go up and some go down.

Well, we are saying the same thing I think–use LBL if you can.
What is the mechanism by which some resonances go up and some go down? More mass = lower resonant f, no? Lead has about the 10-20x the density of LBL and makes sense as a Cadillac option.

[xandresen]

xandresen,
That's why I asked about "Hardi backer" concrete board. Very stiff and almost zero resonance with a knuckle rap. Any experience testing concrete boards?

I am not familiar with the material so I didn't try it. You could easily make some 8.25 by 10.5 constrained layer test panels. The Parts Express brand viscoelastic damping material is available in single sheets. You can buy a single 6x8 tile at Home Depot and bond (with acrylic based mortar, not something soft!) to the damping layer to see what I feel works well. Compare with a 6x8 piece of your material.

The reason to make the test panels so large is so you can screw them to a rectangular frame of 1x4 wood to simulate a braced portion of a speaker cabinet wall. I use 10 screws.

For a cabinet, tile sounds hard to work with but it can be scored and snapped to size just like glass. Buy cheap tool at the Depot or Lowes.

2. Constrained layer should be stiff. If same material, 1/2 to 1 times thickness of base layer.

Yes I read the same thing. The guide lines I saw were mostly for steel.
I made test panels with 1/2" MDF outer layer, PE viscoelastic damping layer and 1/4" MDF or 1/4" BB constraining layer. Didn't work worth anything.

Tile was an act of desperation but it sure works. Really stiff.

[Winter]

What is the mechanism by which some resonances go up and some go down? More mass = lower resonant f, no? Lead has about the 10-20x the density of LBL and makes sense as a Cadillac option.

I'm not sure of the mechanism. I've seen the results in a couple of tests, where a layer of significantly dense material (and not much stiffness) was added to a baffle. One such article, Jim Moriyasu's "Panel Damping Studies: Reducing Loudspeaker Enclosure Vibrations", AudioXpress, February 2012, tested multiple baffle configurations.

MDF and Baltic birch plywood baffles both give an elevated vibration signature slightly over an octave wide around their natural fundamental frequency, say from around 200 to 400 Hz, higher or lower depending on thickness and bracing of material. Applying a dense layer, such as lead, or heavy polymer sheets, results in an increase in the primary resonance amplitude and a reduction in the amplitudes of the adjacent secondary vibration frequencies. In Moriyasu's study, this was shown when applying 1/16" lead sheet to 3/4" MDF with contact cement, or when applying a Deflex subwoofer damping panel to 3/4" MDF.

Moriyasu's article is available as an archived e:file at the following website:

https://web.archive.org/web/20170529150043/http://www.cc-webshop.com/CD-DVDs_c47.htm

[Winter]

Yes I read the same thing. The guide lines I saw were mostly for steel.
I made test panels with 1/2" MDF outer layer, PE viscoelastic damping layer and 1/4" MDF or 1/4" BB constraining layer. Didn't work worth anything.

Tile was an act of desperation but it sure works. Really stiff.

Appears the damping material was too stiff to damp well with the 1/4" MDF or Baltic birch layers constraining layers, but worked well with the stiffer tile. Good results.

[5th element]

Hurrah for piles of drivers. I am not surprised that the 12MU sounded good like that. That's one of the advantages to well designed soft cones, a general absence of severe resonances and a decently extended upper response, especially with small diameter drivers. I think you'll be quite happy with the combination, the wavecor bass units certainly are quite something. Sometimes they no replacement for a large amount of steel and magnets!

[BobEllis]

Since nobody seems to be playing while Jon is away, I'll just note that I ordered a minidsp 2x4 in a box on February 26 and it was delivered March 2. Haven't played with it yet, but impressive delivery. Warmish weather predicted next week, I may be able to knock up some test baffles...

[kevinm]

You'll enjoy the MiniDSP. I have been very happy with mine. If you end up using it for its PEQ, it's one of the best I've dealt. 6 channels of PEQ on my MiniDSP runs circles around the 11 on my UMC-200.

[BobEllis]

Good to know, Kevin. Thanks.

[Renron]

Ahhhhhhh!
Just got a call from Solen in Canada, they said that the SWbd01 has been discontinued and they cannot fill my order of 1 more driver. Thanks a H3ll of a lot, what do I do with 3 drivers???
Anybody have 1 they want to sell? not likely, but it's worth asking........
Ron

[BobEllis]

Sorry to hear that. Although the numbers look different, Jon did mention that the -02 should work. Hopefully it will be close enough that you can get by with buying no more than 2 drivers. Or use the current ones for a killer center...

[Renron]

OK,
So I contacted Wavecor directly, they have some stock in the BD01 drivers left. Here is the Email from them:

Hello Ron,

Thank you for contacting us.

Since we introduced the new models SW223BD02 and SW223BD03 the SW223BD01 will no longer we a stocked item.
However, as we do have a few of the SW223BD01 in stock we will be able to help you.
The price is US$ 295 plus shipping.

If you let me know the destination post code I will get you a complete price including shipping.

Sincerely

Ed, Wavecor Customer Service

Ouch! I'm still looking around, I think I have found a dealer in the Netherlands with a much better price.
To others who have not purchased drivers yet, I think you better not chance it with the BD01.
I'll find my one that I need somewhere.... I was panicked because Solen Canada told me they were unable to source any more. Not quite true........
Ron

[---k---]

We'll it is cool that they have some stock that they are willing to sell you, but you'd think they'd be more accommodating on the price. I mean, aren't they trying to clear these out? That is probably full retail list price. They probably want to sell it to you at that so they don't piss their dealers off by selling stuff behind their backs and undercutting them. So, I can sort of see it. BUT... Come on. Maybe if you show them the Solen receipt they'd be willing to match. Or maybe they'd be willing to sell it to Solen and then to you or something. I would respond and see what happens.

[dar47]

Hope you got the sale price on the other 3? The original 3 of us paid $270 so 1 for $290 still puts you ahead if you got the sale price. :W

[Renron]

I'm emailing someone in Europe who has quoted me $236 including shipping. I hope they have one in stock. At this price, total driver cost is ~$1900 including shipping. Gotta pay the driver fairy somehow. I'm hoping to keep the total cost for the pair under $3,000. It's looking iffy.

Dar47,
How many watts are you feeding your Ardents? I'm building an amp that is ~180W @ 4 Ohms. I hope mine look 1/2 as good as your's and Benthe8tracks'.
Ron

[dar47]

Ya there is 2 sticker shocks drivers then xovers but once you get over that it's all good.

Right now an Emo xpa-2 rated 350w at 4 ohms but I don't even move the meters. Pretty benign impedance so unless your in a very large room your 150W should do music very loud. I'm waiting to see how Ben makes out with the Ncore build he is doing as I do want to go balanced all the way through the chain. The only time I moved the meters was on Cris Botti in Boston Blue Ray on the last song where the full orchestra was in play. Just a little IB sub in the bottom end and the Ardents had the best control with the most detail I have heard. Your going to love these I'm still smiling.:B

[Renron]

Ya there is 2 sticker shocks drivers then xovers but once you get over that it's all good.

:roflmao:

I knew the job was dangerous when I took it.

Ron

[JonMarsh]

Ahhhhhhh!
Just got a call from Solen in Canada, they said that the SWbd01 has been discontinued and they cannot fill my order of 1 more driver. Thanks a H3ll of a lot, what do I do with 3 drivers???
Anybody have 1 they want to sell? not likely, but it's worth asking........
Ron

Just got back from Antarctica/Argentina. I have a few extra SW223BD01- can sell you one. Don't panic....

[Renron]

Cool Doc, welcome back.
I tried to PM you, but, of course your box if full.
Thank you for the generous offer. I'll try to PM ET. I bet he doesn't get much fan mail. LOL
Arrrrghch (force choke)
Ron

[JonMarsh]

I cleared out some room in my PM box- I keep forgetting that that there's a limit on those that isn't very big- (reminds my of my work email inbox....)

[---k---]

I believe that the PM limit is adjustable by status level. I know one forum I frequent grants paid members bigger inboxes for paid users. Maybe Lex should allow Moderator's to have unlimited in boxes.

[BobEllis]

Bringing this thread back from the dead. Life has interfered with speaker building. I spent the day tuning my saw - mostly removing rust, lubricating and adjusting after many years of disuse. Next weekend I should be ready to start cutting panels for the Poor Man build.

In the meantime, I have received my drivers for the original Wavecor Ardent build, using the 002 woofer. Must resist temptation to start that too soon until my skills have recovered from lack of practice.

[---k---]

Without pictures, I'm not sure we believe you have any Wavecor drivers.

[BobEllis]

OK, I should have expected that... Transducer labs ceramic tweeters and C79s in the background

[JonMarsh]

If there are pictures on the internet, then it is real! :W :B :T

[TEK]

Hmm, you are going for all theoriginal drivers as far as I can tell (except for the newer wavecor)?
As the thread title states"poor mans" and "inspiered" I would have assumed other, less costly drivers. Did something change between the start of your journey to where you are now - or is it just something I don't get?
Disclaimer: I did nor reread the whole thread, so this might have been discussed earlier on...

[BobEllis]

TEK, there will be a poor man's practice build with RS225s, ZA14 and probably the Scan 6600, although I may go with a TL N26C or 27TDFC (I had the RS225s and 27TDFCs on hand when I started this as a way to use them). Last time I made sawdust of any significance was 2009, so I need to buff up my skills before doing the real thing.

I got a serious case of mission creep, and will be building a set of Wavecor Ardents as well. With a little prioritizing I am not as poor as I thought I was. Bamboo lumber went on sale, and if I was going to spend that much money and effort, I should consider a better mid. Jon got me thinking about the 12MU/4731, and from there it wasn't too much of a leap to the C79s. Then surprise money came a long and there were the 6640s and Wavecors. It will take a few months to gather crossover parts, by which time I hope Jon will have updated the XO to reflect the currently available Wavecor woofer.

[TEK]

So, instead of a "poor mans" edition you actually ending up with two sets, one "not quite as expensive" and one full blown Wavecor Ardent. Jup, I would call that a "serious case of mission creep" :rofl:
I like it :-)

[Renron]

Bob, glad to see there is sawdust in your future, I was beginning to wonder about this thread.
If your building an Ardent light and a full Monty Ardent, then, I guess I'll take the middle ground. Wavecor (originals) ScanSpeak 12MU/4731 and the ScanSpeak 6600.
I think I read somewhere that the XO for the 6600 is the same (or very close to) the beryllium driver. Does that sound familiar to anyone?
Acquiring parts for the XO's now.
Anyone in the SF bay area / Sacramento area want to split a sheet of 3/4" bamboo ply?
Table saw and sliding saw blades have been sharpened along with some dull bits, I'm almost ready to go too.
Bob, wish we lived closer, we could share tools.
Ron

[BobEllis]

Yes, I remember Jon saying that the 6600 and 6640 can use identical crossovers

[JonMarsh]

Yup.

BTW, my SW223BD02's arrived today...

[BobEllis]

Finally, some progress. Baffles routed for SS6600, ZA14 and RS225. I glued up inner and outer layers before routing. Still need to chamfer the ZA14 openings, but didn't have a suitable bit with me.
Outdoor work is interesting - focusing muscles aren't the only ones that get slow, weak with age. Going inside to change bits and get dimensions of the next cut took a long time to adjust.

Obviously that's not a 6600 in the tweeter recess, just seeing how a 27TBFCG looks if I decide to stay with things I already have.


I'm glad I went practice route first. This did not go smoothly. I used a 3/4" straight bit to do the recesses so it took several tries to get it right. I didn't have the Jasper Jig formula with me, but I was off about a 16th just subtracting the additional radius of the bit. Then I used a cheap digital caliper to measure the ZA14. Inside and outside measurements differ enough that the driver didn't fit the first attempt. Going the PMazz suggested route and using templates and a pattern bit for the good build.

[JonMarsh]

Going the PMazz suggested route and using templates and a pattern bit for the good build.

That's not a bad idea at all... though I'm reasonably handy with the jasper jig and a caliper, it still comes down to measure three times, route once... if it all goes well!

Well, I'll have some more practice coming up soon, routing test baffles for driver testing. I'd better tally up the number and types of drivers to test- I have three spare front baffles for the 0.75 cu ft PE box, but that may not be enough at this point!

[BobEllis]

I'd have thought I'd be OK with a caliper, too. But, the too small recess still measured "big enough". I guess it depends on the meaning of zero...

As long as I use the standard 1/4" bit the Jasper gives me the size I expect. Maybe a downward spiral for the rim then use the straight bit to cut a flat for the mounting surface. Test cuts and learning coming up.

Looking forward to your test results.

[JonMarsh]

My trick with the Jaspar is NOT to use their calculation, just to use measurements. So, after mounting the router and the bit I want to use, I pick the hole that gives me the needed radius between the jig pin and the outer edge of the bit. Just measure it with a steel rule, and I mark my own radius hole locators, because I'm virtually always using a 3/4", except for the spiral bits- again, just measure from the bit edge you're using to the pin locator that gives the right radius.

Don't need no 'rithmetic... :W

[Renron]

I always get it close enough with jasper jig #s, then test it on scrap mdf. You don't have to plunge more than 1/16" deep to determine the correct diameter. Adjust the pin as needed. If you start out on the small size you can even use the same hole.
Very quick and cheap way to get a perfect size every time.
Especially if your bits have been resharpened.
Ron

[JonMarsh]

OK, have the data now for the ScanSpeak 12MU4731- have FRD and ZMA files, too-


First, a composite set of curves at 0 degrees, 15, and 30:




Here's the impedance curves- point to a fairly clean motor with good inductance control:



It will certainly be easy to work with, as regards crossover filter- no undesirable gyrations.


Here, we have a look at distortion- standard 2.83VRMS drive, for comparison to my other measurements:



A fairly typical SS profile- somewhat higher 2nd order, but pretty good performance on 3rd harmonic. This points to doing a crossover on the low end around 600 or above, which is no problem with the current Ardent design.