Is there a previous thread discussing these?
Are they waveguides like this:
http://aa.peavey.com/downloads/pdf/qwp1.pdf
as discussed here:
http://www.diyaudio.com/forums/showthread.php?threadid=16571&goto=newpost
It has a single compression driver in the horn?
Very impressive.
Mark Cronander
Are they waveguides like this:
http://aa.peavey.com/downloads/pdf/qwp1.pdf
as discussed here:
http://www.diyaudio.com/forums/showthread.php?threadid=16571&goto=newpost
It has a single compression driver in the horn?
Very impressive.
Mark Cronander
Variac said:
I don't know. The horns are Unity horns designed by Tom Danley of Servodrive.
They are conical, but without the foam at the mouth. Since they are 60x60 I don't have to worry about astigmatism like in a MantaRay horn. The flare at the end of the horn accomplishes pretty much the same thing as the foam in the Peavey horn. The foam gets you a smaller horn, but I'm not too worried about that. I prefer the look of wood compared to foam.
There is one TAD TD2001 compression driver and four 5" midrange drivers mounted on the sides of the horn (you can just see the entry holes in the pictures. Tom Danley came up with this rather genius innovation that allows a single horn, which normally is a narrow range device, to be loaded over a wide spectrum. These are two-way horns, but you can scale this up to 3, 4, 5-way etc. to cover 20Hz to 20kHz if you have the room. The drivers combine with absolute coherence, so no lobing effects at all, it's really like it was one driver except that the distortion is way lower than you could get with one driver.
Thanks. I'll give the design credit to Tom Danley, but I have put some sweat into these as well.
John
Yep. that's what I thought- the secondary flare does the same as the foam. In the Peavey white paper they pretty much say they use the foam mostly to avoid making the horn bigger with the secondary flare, but it did help with another aspect too. It is indeed a lot nicer looking without the foam.
I asked about the number of drivers because I saw the ports and recalled the Unity horn, byut couldn't remember what it was called.
Is there DIY info on how to construct them? I think I heard they were tricky to get right.
I asked about the number of drivers because I saw the ports and recalled the Unity horn, byut couldn't remember what it was called.
Is there DIY info on how to construct them? I think I heard they were tricky to get right.
roddyama said:
The tweeter/mid crossover is an analog filter at 1.5kHz. It is an asymmetric filter so that the delay in the lowpass is compensated by moving the midranges forward of the tweeter and everything is pretty much time-aligned--not quite linear phase, but close. Group delay is way less than anything else I've ever seen that doesn't use FIR filters and that is on AND off axis.
The crossover between the horn and the dipoles is a delay-derived linear phase digital crossover with an asymptotic slope of 60dB per octave and crossover frequency of 350Hz--note that besides not introducing phase distortion, this crossover also achieves a high slope without ringing.
I'm also doing inverse filtering on everything to correct magnitude to +/-0.2 dB and phase is absolutely linear. I realize there is considerable debate about whether phase is audible or not. With the DSP I've done some listening with the phase correct and with it screwed up. I don't notice any difference at all on sustained notes, like strings, but for percussive sounds like drums, I really notice a difference.
The filtering is done on a PC with a program called wavewarp (www.soundslogical.com) and the DA is an M-Audio Delta 1010.
The "bass" module is an M-baffle dipole with 4 ScanSpeak 10" drivers. They will eventually be augmented with subwoofers (I've got plenty of subs, but a shortage of amps and output channels on the DA right now). Note that the crossover frequency is well into the 1/4 wave transmission line resonance. There is no reason to fear this resonance. You can filter it with a notch filter just like any other resonance or if the resonance is at the crossover frequency, you can just put the f3 of the low-pass filter lower than usual. The nice thing about this resonance in the M-baffle is that if the length and width of the baffle are approximately equal, you get a nice boost to the directionality--it becomes more unidirectional--and mates very well to the coverage patten of the horn. You have to put the side of the dipole with two holes in front and the side with one hole in the middle in back.
I had 2 prototypes up and running in MDF. They went through various iterations. Now I'm keeping the wife happy by making them look nice. I made them modular so I can still play around, though. I'm building a total of 7. Three are almost done, but the workshop I go to is closed for the rest of the month :-(
For me these are the ultimate speakers, absolutely flat magnitude and linear phase on axis. The directionality index is around 12dB from 500Hz up and steadily widens from 12dB at 500Hz to around 8dB at 250Hz. That's pretty darn flat. If you've ever looked at power response measurements before, you know what I mean. I had some second order gradients running for the bass that had absolutely ruler flat power response (second order gradients have a 12dB directionality index when placed on the floor), but they were too expensive to make 7 of them.
Best of all, though, the distortion is extremely low (OK will be once I get the subs going below 80Hz). The horn loading dramatically reduces excursion to begin with, but then I've got 4 midranges to boot!!!! These things are seriously understressed in my living room.
OK, that's enough showing off for now. Thanks for the kind words everyone. It's nice to be able to share these with someone. I got my wife to admit that my system sounds better than her JVC mini-stereo system, but that's about as much as I'll ever get out of her...
John
Hi John, all
Thanks for the picture, looks interresting
I am working on something similar, except
I have two 15" woofers a side in a flat open baffle
for the lower frequencies.
Can I ask you two questions:
What are you using for midrange drivers?
Why the big gap between the Unity and your M-baffle?
Thanks
Thanks for the picture, looks interresting
I am working on something similar, except
I have two 15" woofers a side in a flat open baffle
for the lower frequencies.
Can I ask you two questions:
What are you using for midrange drivers?
Why the big gap between the Unity and your M-baffle?
Thanks
slowmotion said:Hi John, all
Can I ask you two questions:
What are you using for midrange drivers?
Why the big gap between the Unity and your M-baffle?
Thanks
No problem...
The midrange drivers are custom made for Sound Physics and the only way to get them now (unfortunately) is to buy a Unity from Sound Physics. There were some drivers built by Audax that were pretty close, but I noticed that they seem to have discontinued these. I don't know what to recommend now.
The gap is there because I wanted the dipole on the floor to take advantage of the 1/2 space radiation and I wanted the horn at ear level. The crossover frequency is 350Hz, so there are no lobing issues to worry about with this separation.
How about telling us a bit more about your project?
John
hancock said:
Hi John, all
Well, I have a junk prototype running at the moment
in the right channel , a 1" comp-driver and a 6" RCF midrange
on a 60x60x60 cm conical horn,with 2 15" woofers
on a 120 by 120 cm flat baffle on the bottom.
Left channel is same speakers on an open baffle except for
a small exponential horn on the compressiondriver.
It's all very rough at the moment, with only one midrange driver,
and the crossover needs a lot of work, but still I can hear the potential, so this is probably the way to go.
I am bit uncertain about the midrange drivers,
Galaxy makes a 5" that may be usable , and Seas have some
4" and 5" speakers that probably will work. Maybe.
This will be tri-amped , I just have to make some more amps first.
cheers
slowmotion said:
I am bit uncertain about the midrange drivers,
Galaxy makes a 5" that may be usable , and Seas have some
4" and 5" speakers that probably will work. Maybe.
This will be tri-amped , I just have to make some more amps first.
cheers
I've heard not so nice things about the Galaxies, ie. the actual parameters are not even close to the advertised ones and vary quite a bit from unit to unit. Which Seas drivers are you looking at? Another guy I know was asking about possible midrange drivers for horn loading and I'd like to pass on the info.
I know what you mean about having to build more amps. I decided if I was actually going to get this 7-channel-all-biamped-plus-2-subwoofer system going some time this century, I would have to break down and buy some amps, yeach...
John
hancock said:
Hi
Seas CA11RCY H149 is one , a 4" papercone,
Seas MP14RCY H0522 is another, that's a 5" poly.
They are both pretty low sensitivity,
that might not matter so much, since there is
4 of them pr horn.
I haven't a clue if they will work ok, tho
I have just started looking for possible drivers,
so other posibilitys might pop up
Yeah, I'm lucky , I only have two channels
cheers
slowmotion said:
They are both pretty low sensitivity,
that might not matter so much, since there is
4 of them pr horn.
I haven't a clue if they will work ok, tho
I have just started looking for possible drivers,
so other posibilitys might pop up
Yeah, the 50% efficiency of the Unity mid drivers is a bit wasted because they have to be padded to match the tweeter. Please be a gent and let me know if you find anything else promising.
Thanks, John
Here is the electronics setup (for two speakers). I am now running the digital crossover using a plugin in Foobar2000. I love this convenience. I loaded all 800 (or so) CDs onto a Network Attached Storage unit located in another room (so I don't have to listen to it). and run everything from foobar--no more searching for that damn CD--if you haven't done this, you don't know what you are missing. The crossover plugin is a modified version of the convolve plugin. The original convolve just handles stereo in stereo out. My edited version handles stereo in, six channels out. I'll post the plugin later on. I just have the non-optimized Microsoft C++ .NET 2003 compiler. If anyone out there has the pro compiler and wants to speed this plugin up, please let me know. Warning, though...the plugin works for me, but no guarantees for anyone else--careful about hooking your tweeter up directly to the outputs.
Attachments
Here is a closeup of the modification I did of the holes. The holes on the unmodified horn are much bigger than needed for home use. The issue with the holes is that you need to keep air velocity through the holes at less than around 5% of the speed of sound. Above this point you lose laminar flow and start getting turbulance. The smaller the holes, though, the faster the velocity through the hoels for a given SPL. However, you can make the holes around 8mm in diameter and still get plenty of output before hitting that 5% point. I forget exactly how much output--I did the calculations a while ago. However, I seem to recall 115dB. When you make the holes smaller, you also have make them shallower to keep the impedance of the holes the same at the crossover point. I think 3mm was the number (you'll have to play around with this for whatever drivers you are using). The end result of this is that impedance of the holes at 4kHz is raised by an order of magnitude. The reflection at 4kHz that showed up on Nick's original horns (which wasn't a big issue to begin with) is now a non-issue. Tom has a more elegant way of addressing this, but this will work too.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.