Can you smell what Brando's cooking!?

[Spanky Ham]

The BB v2 looks really nice. The Fountek driver looks good in that small enclosure. How does it sound without a tweeter?

[Idefixes]

Hi brandon,

can we have more details on fountek system? I am searching on some so systeme due my lack of place. i have a subwoofer and HT reciever to go with.

Marc

[Zilch]

Dennis> be nice don't need to crank Zilch up already, it'll just derail the thread. Well, not here, Tom is a good mod, but it will over there

Nah, crank me up...!

[augerpro]

The DCX2496 can load transfer functions via PC interface correct? Once I get this design completed I may post the transfer functions as text files if someone wants to try it out on their DCX2496.

[Dennis H]

The DCX2496 can load transfer functions via PC interface correct? Once I get this design completed I may post the transfer functions as text files if someone wants to try it out on their DCX2496.

I don't think the DCX can automatically turn a transfer function into filter settings. Having the transfer functions would still be useful though to help people set the curves manually.

[augerpro]

Huh. Is it the DEQ2496 or am I just "misremembering"?

[Paul W]

It's been 3-4 years since I've played with it so memory is foggy (and maybe faulty) but there is a PC interface on the DCX. Don't think you can upload transfer functions but, as I remember, you can adjust the DCX with a PC.

SoundEasy can export to the DCX but SE is advertised to support only 1.14 DCX firmware which was buggy.

Since it only takes a few minutes to set one up, I'd just publish the settings.
Paul

[ThomasW]

IIRC none of the Behringer devices will import transfer functions and convert them to filter settings.

[augerpro]

So cabinet work is done for the BBV2 and I'm finishing the No Quarter cabs now. Paint is done and it looks beautiful! I'm also finishing up the extendable stand with lazy susan top. Haven't decided how fine I'll mark the angles yet. But this will help me take the design portion to a new level as I can do quite good polars now of complete speakers. Crossovers will be developed soon now that woodwork is done.

When I get to the point of publishing the design I'll post a build thread here and link it in this thread. I'm also going to post the design on my Soma Sonus site. I started a business by that name so I can get wholesale accounts and write off the expenses from taxes. So if you order the parts from PE in the "where did you here from us?" drop down I think you can hit other and type in Soma Sonus. I would appreciate since that will help me negotiate with them to stock the DE160 and maybe even the QSC horn if they see there is a popular project with them that is making sales. It would be nice if people only had to go to one spot to get all this stuff.

[Amphiprion]

If you guys send me transfer functions in text format (CSV or tab delimited or similar) I can turn them into filter coefficients using Matlab for you.

[augerpro]

Thanks for the offer Mark, sounds like that may work. I imagine it can't do any notches though?

[Amphiprion]

You'd be amazed what it can do

[JoshK]

I'd love to see the code if your sharing. I'm a matlab user at work. I've just never sat down to look at the math behind filter theory. Its on my to-do list, but it is a little too close to work.

[augerpro]

So what the heck is going here? I'm measuring the BBV2's with them mounted to my big test baffle to simulate the on-wall response. There will be a dip associated with the delayed signal from the big baffle due to the speaker baffle being about 3" (TM) or 3.5" (MTM) forward of the big baffle surface. This is what I get:



That dip at 675hz is NOT the dip I was expecting, it is far too low. This the TM shown, the deeper MTM is about 600hz. So it is not a defective driver. I messed with the gate also thinking maybe it was slightly too long and allowing a room reflection. Not that. It also shows up off axis. The boxes are stuffed heavily and even though the internal dimensions are far too small for a 675hz dip I took out the stuffing to see what would happen. All it did was cause a wiggle around the expected 2khz area. The wavelength is 20" and 22.5" and no dimension matches that, or 1/2 or 1/4 that I can see. Overall dimensions of the boxes are 4 3/8" x 7" (TM) and 4 3/8" x 10 1/2" (MTM).

What do you guys think?

[Dennis H]

Hey Brandon,

I think the math works for those cancellations. I didn't run the numbers other than roughly in my head but you'd add

average dimension, center of driver to edge of baffle
+ baffle to wall
+ wall back to baffle

and those three add up to 1/2 wavelength at the center of the dip. The MTM has a bigger baffle so the frequency of the dip is lower. Working backwards from your wavelengths, average center-edge of the MT would be 4" and the MTM would be 5.25".

[augerpro]

By god your right Dennis! I hadn't really considered the driver center to edge.

Just measured the fullrange which is 4" deep, 4 3/8" x 5 1/2" baffle and the dip is around 550hz. Doing the math it should be around 520hz so that's pretty close. I'm going to compare the Q of each visually since the FR should be the worst with dimensions of the side being so close and the MTM the best as it should spread out the reflections more. Interestingly as you move off axis the dip gets shallower and higher in frequency until about 40 degrees where it separates into two dips with one going lower and the other higher.

I'm glad I measured this way instead of just simulating IB response. Now the question is to what extent-if any-do I attempt to correct for it? Since it is basically there at all angles it is correctable. But I would putting energy into a null, so it may be something that you have to live with.

[Dennis H]

It's kinda ugly but I think Zaph experimented with putting absorbers around on-walls and it fixed a lot of the reflection problems.

[augerpro]

That was what I was going to try while I was troubleshooting this earlier!

Well the big test baffle is coming down so I can start measuring and developing these two projects. Can't believe it took a whole summer to pull this off, and I still felt rushed the whole time! Bad thing is there are some drivers I'm still curious about, including Zaph's new one so I might have another round this fall. I have two AE IB15's sitting in my basement so I'm thinking matching a pair of Zaph's woofer and maybe a SEAS 29TAFW would make for a cool and low-ish cost open baffle. OF course that would be after the other two projects after these two that I'm already building. So looks like summer 2010 is gonna be busy again

Built a cool stand for doing the normal speaker measurements too. It's basically an extendable stand that can be raised in one inch increments from 23" at the bottom to 65" fully extended. It can measure everything from a full tower speaker to a little bitty single driver speaker. It also has a lazy susan turntable with 10 degree markings for full polar measurements.

[Amphiprion]

For those wondering about the Matlab code, it's nothing fancy. Just the invfreqz() function built into the signal processing toolbox. I used that function in the all-pass listening tests I posted some months ago.

http://www.mathworks.com/access/helpdesk/help/toolbox/signal/index.html?/access/helpdesk/help/toolbox/signal/invfreqz.html&http://www.google.com/search?q=invfreqz%28%29&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a

It seems to crap out if you make it do anything over 10th order (data precision problems maybe?) and you sometimes have to play with the error weighting vector to make it do what you want.

[JoshK]

Thanks Mark. I don't have the signal processing toolbox. I'll have to look up what it does and then see if there is some R/S-plus code or free code that I can transcode.

[Dennis H]

Brandon, you may have been thinking about REW. It can make a stab at generating inverse filters from the frequency response.

[augerpro]

Here are the cabinet dimensions. These are the outside dimensions that assume a certain thickness of wood used. The BBV2 use 1/2" MDF. The No Quarter is 3/4" (or 23mm) Birch ply, but the baffle is 1 1/2" thick and made from one layer of 3/4" Birch ply and one layer of 3/4" MDF. MDF to the outside. It is VERY IMPORTANT to chamfer the backside of the baffle for the BBV2. I just used a Dremel with the little sanding drum. Drill the mounting holes first so you know where to leave some "meat" then chamfer the area between them so the driver flange is sitting an 1/8" landing. On the BBV2 all corners are rounded over with a 1/2" bit. The No Quarter uses a 1 1/4" roundover on the baffle and a 1" everywhere else. I think I bought the big bit from MLCS for a reasonable price. Note this is not a definitive build thread so I'm leaving out some detail, but there are a number fo people with the parts that want to start building so here it is. First the BBV2 backside baffle, TM, and MTM:








And the No Quarter:

[JonP]

Whoops, Brandon! Got a little typo in there, in that the MT is the same height as the MTM...

[augerpro]

Thank Jon, fixed! BTW some construction pics are dumped here: Link Broken

[JoshK]

One word: badass!

[augerpro]

Cabinet

[JonMarsh]

Looking very nice, Brandon! Are they ready to play? Plots and graphs? :W

[augerpro]

Nope, nope and nope But they will be operational by RMAF! BSC is going to be a bit of head scratcher since the baffle is so wide by the time the response rolls off boundary effects of the walls should be rolling. At least so I've read. What do you think Jon?

Bi-ampable with some nice parts from PE:

[Bear]

The No Quarter is 3/4" (or 23mm) Birch ply, but the baffle is 1 1/2" thick and made from one layer of 3/4" Birch ply and one layer of 3/4" MDF.

FYI, 18mm Baltic Birch is usually quoted as being 3/4" equivalent. Since 1 inch is 25.4 mm, your 23mm Baltic Birch is really closer to being 1" equivalent.

Beyond that, these look really, really good. :T

[augerpro]

Ya know what...it isn't 23mm. It's actually a hair less than 3/4" so it must be like 19mm or something. Good thing this isn't the build thread!! So if anyone is building this USE THE 3/4" EQUIVALENT BIRCH PLY OR ACCOUNT FOR USING THICKER PLY!!!

[Hdale85]

Wow those do look really nice.

[JoshK]

yeah they do

[JoshK]

Can I ask what finish did you use on the baffles? I like the look.

Also, it looks like you veneered, then painted. Did you just tape off the veneer? I was afraid of bleeding into the veneer.

[JonMarsh]

BSC is going to be a bit of head scratcher since the baffle is so wide by the time the response rolls off boundary effects of the walls should be rolling. At least so I've read. What do you think Jon?

Your dimensions will change the transition frequency but won't remove the need for BSC- I think once you measure the cabinets, you'll have a clear picture of what to do. Might be a good idea to "measure in situ", that is, placed in a manner relative to boundaries similar to how you want to use them normally.

[augerpro]

Jon> I'll be "looking" at these speakers from a number of perspectives, so yeah, the answer may be clearer when I do the measurements.

Josh> I posted a couple more pics today to try to show the veneer better. Wouldn't you know it but it is still washed out in the pics. Luckily the second speaker behind the first is in shadow and better shows what the finish is like. I'll have to get some close-ups of the horn/roundover transition. Almost a continous curve.

I used Sam Mahoff poly/oil from Rockler for the finish. I used the palm sander to 200 grit, then hand at 400 grit. One tip I can give you use sand as smooth as you can. I had raw unbacked veneer and wish I would have used the palm sander at 400 grit. The smoother the finish the better the poly/oil works. I then applied some Transtint Red dye mixed with alcohol/laquer thinner. After that you just wipe on a coat of the poly/oil and wipe away the excess. I did four coats with a day of dry time in between like the directions say. Really idiot proof and I can't reccomend it enough. I also used the Renassiance wax at the end, although I'm not sure that is needed. There is some Sam Mahoof polish also so maybe that would be better. Just make sure you sand very smooth.

[augerpro]

BTW I was trying to finish up measurements so I can start on crossover modeling but three crickets in my garage made today a total wash

[mikela]

Are the chirpers gone?

[Curly Woods]

BTW I was trying to finish up measurements so I can start on crossover modeling but three crickets in my garage made today a total wash

RAID -"It kills bugs dead"

[CllessuR]

Thank Jon, fixed! BTW some construction pics are dumped here: Link Broken

Say, on picture number 9, which shows the router in the slot being cut for the "horn hole", I notice you already have the inset for the horn cut. It looks kind of like a moat around the area where the hole is being cut.

Question: How was that "moat" cut? I'm a router newbe, so it just wasnt obvious to me how that was made.

Thanks, Cllessur

[augerpro]

Man I did hunting with the RAID!! Little buggers hide pretty though. Measurements have started again, I'll be all done by this weekend. I'll also probably have a first swipe at the No Quarter crossover that I can take the transfer function for and send to Mark H (if he is still available) to generate the filters it would take for active XO users to try out. Not sure if anyone who is building these is that far, and I think both that I know of are doing passive versions, but I would like some helpers with active XO to help me settle on a decent BSC value since these may be trickier than a much narrower cabinet.

[augerpro]

Clessur I used a 1/2" straight bit to do the initial routing of the horn "landing". I probably did 1" width, jsut enough for the flat part of the landing. You don't have to do the full 1 1/2" since the inside 3/4" is going to be rounded off anyway.

[JoshK]

Brandon,

I was actually asking what black finish you used. Is that one of the Rustoleum textured paints?

[augerpro]

My favorite: Duplicolor Truck Bed Coating in the blue spray can. One lesson learned: take the tape off immediately after the last coat of paint is shot or it hardens and doesn't always leave a clean line when the tape is removed. This paint sets up pretty much instantly so you don't have to worry about runs.

[CllessuR]

Clessur I used a 1/2" straight bit to do the initial routing of the horn "landing". I probably did 1" width, jsut enough for the flat part of the landing. You don't have to do the full 1 1/2" since the inside 3/4" is going to be rounded off anyway.

I see, just straight forward plunge routing with a 1/2" straight bit, Thanks for the tips and advice.

CllessuR

[looneybomber]

Brandon, I see you're selling some stuff including the Beyma CD. The Beyma is a pretty nice driver. What made you chose the DE160 over the CP385ND?