AVS Forum banner

1 - 20 of 62 Posts

·
Registered
Joined
·
137 Posts
Discussion Starter #1
I've not posted much here before, but I see great interest in higher sensitivity designs using Pro drivers, so I thought I would start a thread here. I heard this open baffle design by Jeff Bagby/Jim Salk at the inDIYana event in April. I thought they were perhaps the best speaker I've heard, stunningly clean, smooth, balanced and unrestrained with a great soundstage. They were very much live sounding. I was so impressed with them that I've been inspired to try and recreate my own version of it. I'm hoping for sensitivity in the 92-93db range. I believe Jeff said his design was about 93db.


Here is what Jeff is using in the Salk design, according to what he told us InDIYana:

12 Lambda TD12H ($259ea.), crossed at around 450hz.

6.5 PHL 6.5" mid (1120 is the available equivalent) midrange ($159ea.), crossed at 2.7khz

Morel MDT33 tweeter ($149ea.)
http://aespeakers.com/drivers.php?driver_id=6
http://estore.websitepros.com/1736754/Detail.bok?no=48
http://www.madisound.com/catalog/pro...roducts_id=570


For those who have not seen it, here is the Bagby/Salk design.



Goals:

My goal was to create a little smaller speaker (higher SAF, you know
). My concept is about 16 wide at the base, tapering to 7 wide at the top and is about 42-43" tall. I'm thinking of doing the woofer bafflesout of laminated solid hardwood, 1.5" thick. The idea is to laminate them from 1"x2" hardwood stock laid flat and stacked vertically. I will mix a second hardwood as shown strictly for aesthetics. The open baffle section will be a mix of 1/2" Lexan with a contunaction of the same laminated hardwood as a bacer for mounting the drivers.


I'm also considering mounting the woofer, midrange and possibly the tweeter from the rear, leaving 1/2" wood between the driver and the baffle face. I would use a 1/2" round-over on the edge. The tweeter,in effect, would have a shallow, 1/2" deep waveguide. This serves to boost the low end output by about 3db over a broad range centered around about 3kHz. When equalized out in the crossover, it will allow the tweeter more headroom and lower distortion at the low end and also help in countering the rising response of the Peerless HDS at the upper end. Mounting the woofer from behind lessens the visual impact it has on the design. Of course, I will be doing some measurements to see just what effect this will have on the drivers natural frequency response before making a decision to use this kind of mount with any of the drivers. Below is a mock-up of what it is planned to look like.




Drivers:

The drivers of the Salk/Bagby design run in excess of $1100/pair. I would like to reduce that where possible, while losing little in either sensitivity or sound quality. Here are the drivers I've come up with to use, with an all passive crossovers at about 450 and 2500hz.


Woofer:

Lambda TD12H 12, 93db: $468/pair (link above)


Midrange:

B&C 6MD38, 96db: $92 ea.
http://www.partsexpress.com/pe/showd...number=294-651


Tweeter:

Peerless HDS 810921, 93db: $78
http://www.madisound.com/catalog/pro...oducts_id=1744

I pcked up two of these for $55 before the price went up.


The B&C 6MD38 seems to be well thought of, has very similar specs and published FR. Zaph has recently tested it and "it looks like a drop in replacement for the PHL" (Jeff's words, not Zaph's). The total driver cost will have been reduced to about $762/pair, and though still very expensive, is about a $340 savings.


The TD12H is in a 55L box. I'm using a slot port, exiting at the bottom rear of the box. Tuning will be such that the F3 will be about 35hz. It will put out over 110db at 100watts and not be even close to Xmax. I'm trying a new box deisgn that uses red oak ribs dadoed into the walls combined with simple cross braces for panel resonance control. According to my calculations, it will use about 17% less space than an equivalent window-brace type design.


I've done some diffraction modelling in BDS. Here are the results for the two drivers:



In addition, I've done some diffraction simulation and dipole simulations in John k.'s ABCDipole and in the Edge. Attached are the diffraction and then dipole simulations from ABCDipole:





Below is a graph from the Edge model showing dipole and diffraction results.



Here are some results on woofer/midrange floor bounce nulls.



Construction of a prototype is ongoing and I will post some photo's next.
 

·
Registered
Joined
·
137 Posts
Discussion Starter #3
Here are more photos. The slot port will be removable until a final version of the designis complete so adjustments can be made. Also, I will be able to pull it out and finish it separately, or possibly make it out of hardwood.



 

·
Registered
Joined
·
137 Posts
Discussion Starter #5
Here are some more construction pics. FYI, the baffles are test baffles. The finished ones will be laminated hardwood and Lexan as planned. The box has been glued up (except the top). The woofer baffle will be removable. Here you can see the cross bracing implemented, including a "filler brace". This brace will be glued to front of the cross brace and have speaker gasket material on its face. It will physically connect to the back of the woofer. This will essentially connect the woofer to all four sides, to distribute vibrations and reduce motor movement. Nxt up are to add a small dowel, top to bottom, through the slot port (to control panel resonance in the top slot port wall), add front face cross braces that will be used for mounting threaded inserts for the baffle removal and pack the space between the ribs with insulation.


When the top and baffle are clamped on, the box does indeed sound quite dead, from a knuckle rap perspective. I have no other way to test it. Of course it will be even better when the top is glued on, the filler brace fully implemented and the final version of the port glued in.


You can see the woofer mounted from the back, 1/2" inset, with 1/2" roundover. I plan to cut a steeper chamfer on the top of the woofer baffle, probably 30º or less. That's all for now!



 

·
Registered
Joined
·
3,475 Posts
Very nice work indeed! I like the oak splines to give the panels stiffness... without taking away from the volume much at all. It looks like you randomized the placement of them so that any (?) resonances are pushed over a larger area.


I also like the fact that the shape will also reduce the chance for a standing wave to set up side to side and front to back. Like I said... fantastic work so far!
 

·
Registered
Joined
·
137 Posts
Discussion Starter #7
Thanks, Michael.


The rib placement is generally at golden mean distances on each panel, starting with one about centered where the most support is needed. The cross-braces are offset similarly, with the exception of having to work around the driver.


There are certainly benefits to the bass-bin shape, but ease of construction is not one of them.
 

·
Registered
Joined
·
137 Posts
Discussion Starter #9
I've got a prototype of my BaSSlines together, for testing purposes, though I don't yet have my midrange drivers yet, since they are on backorder until October 10th. Instead I popped in some used Audax Pro mids into the open baffle.


I started with Jeff's original crossover for his Salk speaker and just made some adjustments by ear and also knowing that the tweeter would have a boost at the low end due to the 1/2" deep waveguide mount. Based on my raw listening impressions, I'll say is I think these could be very, very good. The TD12H is one clean woofer. I think the Peerless HDS tweeter in the ½ waveguide is going to work out very well also, based on this test.


Here are a couple of photos of what I'm listening to as I build the second box and wait for the mids to get into stock. I should be starting on some measurements soon on the TD12H and the HDS. I have the Lexan for the open baffle section, but see no reason to mess with it until the design is nearly finalized.


 

·
Registered
Joined
·
3,880 Posts

Quote:
I started with Jeff’s original crossover for his Salk speaker and just made some adjustments by ear.

You rebel
 

·
Registered
Joined
·
137 Posts
Discussion Starter #11
Not really. I will do complete measurements for the final crossover design work.
This was just to give me an idea of the potential and something to listen too while completing the other box.
 

·
Registered
Joined
·
3,451 Posts
dlneubec,


Outstanding work ! I like your cabinet design and construction. I will keep this thread a must check.
 

·
Registered
Joined
·
137 Posts
Discussion Starter #13
While waiting for my B&C6md38 mids to arrive, (or the Lambda 6.5" ids to be ready) I thought I'd start on some measurements and xover design simulations. However, I will use the Audax Pro170 mid in the meantime.


I wanted to test the Peerless HDS tweeter in a 1/2" deep waveguide mount, a standard flush mount, and also wanted to see what would happen if I created a 3/4"+ waveguide by mounting the tweeter to the back of 3/4" stock, not removing the faceplate of the tweeter. The latter two I accomplished by making one baffle with a 4-1/4" through hole and a 1/2" rebate and several wood inserts that fit in that hole and rebate. I then installed the tweeter in those inserts. Sort of the same idea that Zaph uses for his infinite baffle testing. The 1/2" waveguide mount was already constructed on a separate test baffle.


I have completed some preliminary testing of the flush mount, 1/2" deep waveguide and 3/4" deep waveguide, both on and off axis. The 3/4" deep wavguide was done as follows, since I don't have a 3/4" roundover bit at the moment. First I rounded it over with a 1/2" bit. I actually tested it this way outside. Then I used a 1/2" chamfer bit to open up the hole a bit more. Then I sanded the sharper edges by hand to round them over. This is what I tested with indside. I'd still like to test that with the tweeter face plate removed, as it is in the 1/2" waveguide. I would guess having the faceplate on adds at least 1/8" to the depth of the 3/4" waveguide, making it closer to 7/8"-1" deep. The face plate itself is also recessed, which I filled by using speaker gasket tape.


Here are some measurements of the three. The flush mount and first 3/4" waveguide measurements were done outside, at 103" and only 0,22.5 and 45 degrees. The 1/2" waveguide and second 3/4" waveguide were done inside, at 1 meter and a 4ms window, with the tweeter up about 6' off the floor, at 0, 15, 30, and 45 degrees. The relative spl levels between inside and outside measurements are not comparable due to different levels when importing the impulse response.


The 1/2" waveguide provides a subtle boost that flattens out the rising response of the HDS tweeter from 4khz to 20khz on axis. For example, from 2khz to 17khz there is a 5db rise in the flush mount FR. With the 1/2" waveguide, that is reduced to about 1/2 of that.


The 3/4" waveguide provides a much larger boost, centered around 2.7khz. The modified 3/4" waveguide is not too much different, so opening it up with the 1/2" chamfer did not seem to have too much effect. I will probably test it with the tweeter face plate removed, but it looks like the 1/2" waveguide will solve the main issue, by boosting the low end FR just enough to flatten out the rising responce of the HDS tweeter at the top end.


Peerless HDS Tweeter tests:


Flush Mount: outside at 103", 0, 22.5 and 45 degrees horizontal off axis:



1/2" deep waveguide: inside at 103", 0, 15, 30 and 45 degrees horizontal off axis:



3/4" waveguide: outside at 103", 0, 22.5 and 45 degrees horizontal off axis:



Modified 3/4" waveguide: inside at 1m, 0, 15, 30 and 45 degrees horizontal off axis:
 

·
Registered
Joined
·
137 Posts
Discussion Starter #14
Here are some Lambda TD12H and Audax Pro170 mid measurements. The Audax is just temporarily being used while I'm waiting for the B&C 6md38's to come in. Both of these were tested on axis, mounted from the rear with a 1/2" roundover. The TD12H was measured outside and includes a far field and almost nearfield (about 10") merged.


Quite a response for a 12" woofer!



TD12H:



Audax mid:
 

·
Registered
Joined
·
137 Posts
Discussion Starter #15
Here is a preliminary passive xover simulation, currently using the temporary Audax mid, wth crossover points around 450hz and 2500hz.

 

·
Registered
Joined
·
137 Posts
Discussion Starter #16
Here is what I'm currently thinking on the finish. Padouk and Maple or Sycamore. This one is with maple.

 

·
Registered
Joined
·
26,476 Posts

Quote:
wth crossover points around 450hz and 2500hz.

Why not cross over the TD12 at 700 or 800 Hz?
 

·
Registered
Joined
·
137 Posts
Discussion Starter #19
Exo/Penn,


I wanted these to run full range and the slot ported TD12H's here have about an F3 of 35hz in this bass bin. Also, for reasons having to do with the open bafle dipole mid design, a crossover in the 400-500hz range is ideal. Since I wanted to use a dome tweeter, I didn't need the extra sensitivity of the other TD12 line. In addition, the Bagby/Salk speaker I heard back in April used the TD12H, though a different midrange and tweeter, and it was simply a fantastic speaker. The TD12H is probably the cleanest, most dynamic woofer I've heard. There is no question in my mind that I will be very happy with them.


I have two 12" subs that are sealed with Linkwitz transform circuits, flat to about 20hz, that pick up the very low end and these will cross to them, when used for HT, at 40hz. The subs are typically not used for music, though blend very well with my current mains which have an F3 of 29hz, for music or HT, when crossed at 40hz.
 

·
Registered
Joined
·
2,575 Posts
Since I got my speakers I no longer use the sub with music those Lambda's work just fine. I will one day move my Corwn XTi1000 to the mains when I get another amp for the sub.
 
1 - 20 of 62 Posts
Top