Patrick Bateman said:
I spent some more time trying to optimize the holes.
In my last unity, i had a peak and a huuuuge null in the midranges. If anyone's curious, I can explain what I did wrong there. I figured it out yesterday, but haven't posted it yet.
Nonetheless, take a look at the attached pic, where I've tried to optimize the enclosure for the Tangbands. It looks like the peak is caused by using a compression ratio which is too agressive. When the throat (aka the holes) is big, the peak is reduced. Of course as the throat gets bigger, it begins to screw with the response of the compression driver.
I'm thinking it may be possible to have a relatively large peak in the midrange response, then null it out by placing the holes one quarter wavelength from the apex of the horn.
Attachments
Last time I tried to build a Unity, I found that there aren't many drivers which are appropriate for a Unity horn. With some input from GM, I created a spreadsheet which tells you which midranges are appropriate. The optimum midrange is based on some specs from John Sheerin's webpage. There are a handful of midranges which are very VERY close.
The new Peerless 2 inch is almost perfect, if the published specs can be trusted. The B&C 6MDN44 is also very good, albeit expensive.
The "real" Unity uses a compression driver which is much larger than the one I like. Because of this, I prefer drivers which have a higher FS than Sheerin's ideal Unity.
Members of this club include the Ciare 6.38ndmr from Assistance Audio and the P-Audio/Winner WN-520N.
I'm happy to report that I've found a mid which is damn near perfect for my Unity project. A few years back I purchased a set of TangBand W2-852SC drivers. This is a two inch midrange. Darren Kuzma published a speaker that uses it's cousin, which has a poly cone and a phase plug. The W2-852SC is identical, except it uses a paper cone. Because it has a lighter cone with the same motor, it has a lower QMS and a higher FS, which is exactly what we want.
Unfortunately, the W2-852SC was discontinued.
Luckily, I ordered a batch of it's replacement, and they're even better! The W2-852SH looks similar to the 852SC, but uses a cast plastic frame. The cone must be lighter too, because the FS is much higher. But this is good news, it makes it an even better candidate.
I personally measured the T/S values of a batch; here are the averages:
TangBand W2-852SH
FS 192hz
QMS 4.28
QES 0.54
QTS 0.48
The new Peerless 2 inch is almost perfect, if the published specs can be trusted. The B&C 6MDN44 is also very good, albeit expensive.
The "real" Unity uses a compression driver which is much larger than the one I like. Because of this, I prefer drivers which have a higher FS than Sheerin's ideal Unity.
Members of this club include the Ciare 6.38ndmr from Assistance Audio and the P-Audio/Winner WN-520N.
I'm happy to report that I've found a mid which is damn near perfect for my Unity project. A few years back I purchased a set of TangBand W2-852SC drivers. This is a two inch midrange. Darren Kuzma published a speaker that uses it's cousin, which has a poly cone and a phase plug. The W2-852SC is identical, except it uses a paper cone. Because it has a lighter cone with the same motor, it has a lower QMS and a higher FS, which is exactly what we want.
Unfortunately, the W2-852SC was discontinued.
Luckily, I ordered a batch of it's replacement, and they're even better! The W2-852SH looks similar to the 852SC, but uses a cast plastic frame. The cone must be lighter too, because the FS is much higher. But this is good news, it makes it an even better candidate.
I personally measured the T/S values of a batch; here are the averages:
TangBand W2-852SH
FS 192hz
QMS 4.28
QES 0.54
QTS 0.48
Years ago John Sheering evaluated the Galaxy Audio midrange for use on a Unity, but dismissed it in favor of the MCM 55-1595. Since that time they've offered a new model with a neo motor. After reading good things I decided to evaluate one for my Unity.
It's specs are quite good for a Unity, but I prefer the tangband for my application.
Here are the specs which I personally measured:
Galaxy Audio S5N-8
FS: 141hz (130 published)
QMS: 1.01 (1.14 published)
QES: 0.7 (0.66 published)
QTS: 0.41 (0.42 published)
As you can see, the published specs are quite close to reality, which is more than you can say for Tang Bands!
It's specs are quite good for a Unity, but I prefer the tangband for my application.
Here are the specs which I personally measured:
Galaxy Audio S5N-8
FS: 141hz (130 published)
QMS: 1.01 (1.14 published)
QES: 0.7 (0.66 published)
QTS: 0.41 (0.42 published)
As you can see, the published specs are quite close to reality, which is more than you can say for Tang Bands!
GM said:
Nah, I didn't bother with VAS. My plan is to use the smallest enclosure humanly possible, than resize it until the impedance curve is correct.
Mark Seaton, who used to work with Danley, said Unity midranges need an extraordinarily small enclosure.
This is *another* mistake I made with the last Unity. The fundamental screwup was that I had the midrange enclosures tuned too low. So the enclosure was too big, the ports were too long, etc...
Today I went to Home Depot and bought some parts for the horn mold.
Well actually it's a waveguide.
I don't know why Danley uses a conical horn with a square mouth.
Conical isn't so bad, but square mouths are a bad idea.
Oblate spherodial is definitely an improvement, and that is what I'll use, just like last time.
Ironically, the Yorkville Unity uses a speherical mouth, and is designed by Danley.
If you crazy kids want to build a Unity too, here's the data you need for a sixty degree waveguide. Just take "x", which is the distance down the throat, and size the throat to be equivalent to the diameter.
I uploaded a spreadsheet to Google docs too:
60x60 oblate spheroidal spreadsheet
x Radius Diameter index throat angle divisions per inch
0.000 0.500 1.000 0.000 1.000 60.000 8.000
0.125 0.505 1.010 1.000
0.250 0.520 1.041 2.000
0.375 0.545 1.090 3.000
0.500 0.577 1.155 4.000
0.625 0.617 1.233 5.000
0.750 0.661 1.323 6.000
0.875 0.711 1.422 7.000
1.000 0.764 1.528 8.000
1.125 0.820 1.639 9.000
1.250 0.878 1.756 10.000
1.375 0.938 1.876 11.000
1.500 1.000 2.000 12.000
1.625 1.063 2.126 13.000
1.750 1.127 2.255 14.000
1.875 1.192 2.385 15.000
2.000 1.258 2.517 16.000
2.125 1.325 2.650 17.000
2.250 1.392 2.784 18.000
2.375 1.460 2.919 19.000
2.500 1.528 3.055 20.000
2.625 1.596 3.192 21.000
2.750 1.665 3.329 22.000
2.875 1.734 3.467 23.000
3.000 1.803 3.606 24.000
3.125 1.872 3.744 25.000
3.250 1.942 3.884 26.000
3.375 2.012 4.023 27.000
Well actually it's a waveguide.
I don't know why Danley uses a conical horn with a square mouth.
Conical isn't so bad, but square mouths are a bad idea.
Oblate spherodial is definitely an improvement, and that is what I'll use, just like last time.
Ironically, the Yorkville Unity uses a speherical mouth, and is designed by Danley.
If you crazy kids want to build a Unity too, here's the data you need for a sixty degree waveguide. Just take "x", which is the distance down the throat, and size the throat to be equivalent to the diameter.
I uploaded a spreadsheet to Google docs too:
60x60 oblate spheroidal spreadsheet
x Radius Diameter index throat angle divisions per inch
0.000 0.500 1.000 0.000 1.000 60.000 8.000
0.125 0.505 1.010 1.000
0.250 0.520 1.041 2.000
0.375 0.545 1.090 3.000
0.500 0.577 1.155 4.000
0.625 0.617 1.233 5.000
0.750 0.661 1.323 6.000
0.875 0.711 1.422 7.000
1.000 0.764 1.528 8.000
1.125 0.820 1.639 9.000
1.250 0.878 1.756 10.000
1.375 0.938 1.876 11.000
1.500 1.000 2.000 12.000
1.625 1.063 2.126 13.000
1.750 1.127 2.255 14.000
1.875 1.192 2.385 15.000
2.000 1.258 2.517 16.000
2.125 1.325 2.650 17.000
2.250 1.392 2.784 18.000
2.375 1.460 2.919 19.000
2.500 1.528 3.055 20.000
2.625 1.596 3.192 21.000
2.750 1.665 3.329 22.000
2.875 1.734 3.467 23.000
3.000 1.803 3.606 24.000
3.125 1.872 3.744 25.000
3.250 1.942 3.884 26.000
3.375 2.012 4.023 27.000
Alright, it took almost two hours, but I've finished the mold.
First thing you'll notice is that it's stupid small. No joke, it's smaller than a soda can.
When I make a Unity, I make it SMALL.
This one is going right back in my Accord, just like the last one.
Don't give up on this thread though; you can use this for a full size unity. All you have to do is extend the curve all the way to a 30 inch mouth.
I can get away with a dramatically tiny waveguide because I'm placing this in the corners of my dash, where the windshield and the driver's window extend the mouth out to the proper size.
Corner loading rules.
Check out the pix.
First thing you'll notice is that it's stupid small. No joke, it's smaller than a soda can.
When I make a Unity, I make it SMALL.
This one is going right back in my Accord, just like the last one.
Don't give up on this thread though; you can use this for a full size unity. All you have to do is extend the curve all the way to a 30 inch mouth.
I can get away with a dramatically tiny waveguide because I'm placing this in the corners of my dash, where the windshield and the driver's window extend the mouth out to the proper size.
Corner loading rules.
Check out the pix.
Attachments
I've spent the better part of two hours trying to make an enclosure small enough for these TB mids.
Back in '06 I made the enclosures waaaay too big for my Aurasound mids.
This time around I'm carefully measuring the impedance curve to insure the enclosure is sized properly before I bolt anything to the waveguide.
Only problem is that you need a preposterously small enclosure.
First I tried a piece of 2" PVC pipe which is just big enough to contain the mid. That was far too big; the FB of the enclosure was much too low.
Then I found a 2" plastic sphere at a crafts store. If you cut it in half, it's big enough to contain the 2" midrange. That was better, but still too big.
Then I said phukit, and used clay to cover all the holes in the frame. When you do this, the mid is basically in a sealed enclosure defined by the air between the frame and cone.
That was a little too small, but a big improvement over the other two.
Attached is a pic.
As we speak I'm listening to the midrange play in this tiny enclosure, and it doesn't sound too bad. So this may be my best option.
I'll re-model it in hornresp and see how it looks.
Back in '06 I made the enclosures waaaay too big for my Aurasound mids.
This time around I'm carefully measuring the impedance curve to insure the enclosure is sized properly before I bolt anything to the waveguide.
Only problem is that you need a preposterously small enclosure.
First I tried a piece of 2" PVC pipe which is just big enough to contain the mid. That was far too big; the FB of the enclosure was much too low.
Then I found a 2" plastic sphere at a crafts store. If you cut it in half, it's big enough to contain the 2" midrange. That was better, but still too big.
Then I said phukit, and used clay to cover all the holes in the frame. When you do this, the mid is basically in a sealed enclosure defined by the air between the frame and cone.
That was a little too small, but a big improvement over the other two.
Attached is a pic.
As we speak I'm listening to the midrange play in this tiny enclosure, and it doesn't sound too bad. So this may be my best option.
I'll re-model it in hornresp and see how it looks.
Attachments
With a bit of ingenuity, I was able to create an enclosure that's small enough for the woofers, but not too small. The enclosure is half of a 2in plastic sphere which you can buy at a craft store. That size is fractionally too small, so I've added a ring of 2in PVC that's just 1/4in thick so that the sphere clears the motor of the driver.
I'll post some pics later.
Using an enclosure this size, I'm able to get the FB to 450hz, which is where I want it. If hornresp is correct, I may be able to get a couple octaves bandwidth out of the Tangbands. That will cover 350hz to 1.4khz, where the BMS will take over.
One benefit of such a tiny enclosure is the power handling. I'm astonished at how much power the Tangbands can handle in such a tiny box. As I type this I'm listening to a single TB mid, driven by a 50 watt amp. It's able to fill the whole room with music, with no xover, and little strain. Much of this has to due with the dramatically higher sensitivity of the paper cone mid.
I'll post some pics later.
Using an enclosure this size, I'm able to get the FB to 450hz, which is where I want it. If hornresp is correct, I may be able to get a couple octaves bandwidth out of the Tangbands. That will cover 350hz to 1.4khz, where the BMS will take over.
One benefit of such a tiny enclosure is the power handling. I'm astonished at how much power the Tangbands can handle in such a tiny box. As I type this I'm listening to a single TB mid, driven by a 50 watt amp. It's able to fill the whole room with music, with no xover, and little strain. Much of this has to due with the dramatically higher sensitivity of the paper cone mid.
Nice job. I used those TB drivers in an early Unity prototype and that was exactly the problem - I couldn't get the enclosure size right. It was frustrating, as their parameters otherwise looked really good.
The next challenge will be mounting to the OS flare - the problem that I seemed to hit was that the ideal location for the entry holes was along a curved part of the flare which will make either the port too long, or the front chamber too large. (or require machining of the flare to inset the drivers)
The next challenge will be mounting to the OS flare - the problem that I seemed to hit was that the ideal location for the entry holes was along a curved part of the flare which will make either the port too long, or the front chamber too large. (or require machining of the flare to inset the drivers)
Here are some updates.
- I've recanted on using a 60x60 OS waveguide. I originally went that route because I wanted to make this Unity as small as humanly possible. Optimally I wanted it to fit in a 4" x 4" cube! Last night I was looking at my new mold, and my old mold is just ten times better. I probably invested 40-60 hours on the last mold, and I just don't have time to invest in a new mold. It just blows me away that even if you had all the money in the world, you can't BUY a good OS waveguide over the counter. It's ridiculous. JBL or 18 Sound or SOMEONE needs to get with the program. There's a 90 degree waveguide that's close, which you can buy from Assistance Audio, but they never returned my emails when I tried to buy it. I also called their phone number, and the mail box was full. It's frustrating.
If anyone is following along with this project, you'll probably need me to glass up some waveguides for you, because creating the mold is not a trivial project. I've published instructions on how I make waveguides here about two years back.
- Yesterday I ordered eight more midranges. My plan is to build three unities. One will be a test mule, the other two will be the ones I listen to. Last time around it was a total p.i.t.a. to move the Unities from lab to listening environment all the time.
- I may have to shelve $250 worth of compression drivers. I'd intended to use the 16ohm version of the BMS 4540nd, but I haven't been successful in finding a third.
- I've recanted on using a 60x60 OS waveguide. I originally went that route because I wanted to make this Unity as small as humanly possible. Optimally I wanted it to fit in a 4" x 4" cube! Last night I was looking at my new mold, and my old mold is just ten times better. I probably invested 40-60 hours on the last mold, and I just don't have time to invest in a new mold. It just blows me away that even if you had all the money in the world, you can't BUY a good OS waveguide over the counter. It's ridiculous. JBL or 18 Sound or SOMEONE needs to get with the program. There's a 90 degree waveguide that's close, which you can buy from Assistance Audio, but they never returned my emails when I tried to buy it. I also called their phone number, and the mail box was full. It's frustrating.
If anyone is following along with this project, you'll probably need me to glass up some waveguides for you, because creating the mold is not a trivial project. I've published instructions on how I make waveguides here about two years back.
- Yesterday I ordered eight more midranges. My plan is to build three unities. One will be a test mule, the other two will be the ones I listen to. Last time around it was a total p.i.t.a. to move the Unities from lab to listening environment all the time.
- I may have to shelve $250 worth of compression drivers. I'd intended to use the 16ohm version of the BMS 4540nd, but I haven't been successful in finding a third.
dwk123 said:
I think I should be alright. It would be easier to deal with this if I used a 60 degree spherical waveguide, but I've reverted back to my previous waveguide, which is something like 80x100 degrees.
If I wanted the midrange cancellation notch to show up at 2khz, then I want the holes to be 1.725" inches from the throat right? That should be doable.
dwk123 said:
As an aside, you were using a Dayton RS52 dome mid at the apex of your unity horn weren't you? I prefer a compression driver, but those Dayton mids are actually good candidates for midrange duty in a Unity. Can you imagine how great four of those would sound? That's one thing that's a bummer about the Unity, it's not really practical to use cutting edge drivers for the mids, since you need eight of them. The Dayton has a shorting ring, which is unusual at that price level.
Of course Danley might argue that a shorting ring is unnecessary, since we're using them in a passband where the driver's inductance isn't a big issue. And any distortion reducing mechanism is almost overkill when your efficiency is so high.
Based on my examination of the Dayton, you'd have to modify it's cup to raise it's FB though.
http://www.speakerdesign.net/dayton_rs/midrange/rs52/dayton_rs52.html
No joke, I spent close to three weeks trying to find an enclosure for the midranges. I was using a plastic sphere which I purchased at a craft store about three years back. The sphere is 2" in diameter. For the life of me, I couldn't find any more! I tried other enclosures made from plumbing parts you can buy at Home Depot, and they just weren't as good.
After digging around for a few weeks, I found a craft store that still had them. Score! I even looked online and couldn't find these.
After digging around for a few weeks, I found a craft store that still had them. Score! I even looked online and couldn't find these.
Hi Patrick,
I also plan to give a try to a UNITY horn.
I would like to get quite high efficiency, so I would like to use for example 8x 2" as mid and 2x8" as low beside the BMS 1"
Did you do experiment with configurations for the mid playing high to 2kHz?
Also did you try to build low extension like in the SH50? I mean with two speakers into the horn thru holes and the back chamber also goes into the same horn at a different location...
Thanks,
Tamas
I also plan to give a try to a UNITY horn.
I would like to get quite high efficiency, so I would like to use for example 8x 2" as mid and 2x8" as low beside the BMS 1"
Did you do experiment with configurations for the mid playing high to 2kHz?
Also did you try to build low extension like in the SH50? I mean with two speakers into the horn thru holes and the back chamber also goes into the same horn at a different location...
Thanks,
Tamas
Patrick Bateman said:
Probably an engineering compromise for manufacturing reasons. Much cheaper to use flat panels, unless the numbers justify a mold. Also easier to mount the drivers and the horn to the bass enclosure. A spherical horn can only really mount three mids. But that may be enough. Member mefestofelez, had machined the entry section for some round cones. He showed them to me some months ago. I don't know how much further he's gotten. He borrowed my LsCad and loaned me an O-scope. Haven't heard from him lately. I may try an OS horn at some point.
Sheldon
In addition to manufacturing considerations.
He's selling to pro sound market. In terms of efficient use of space, arraying square or rectangular speakers is far easier than other shapes, especially curved. The square/rectangular cabinet front of a synergy speaker is the whole radiating area.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.