Combining compression drivers in one horn?

It's been a while, but I've started to get back into thinking about loudspeakers (though I cringe at some of the past conversations I've been involved in here!). I recently stumbled upon a low-midrange compression driver (the Miyako DU-100), and thought about combining it with the Pyle Pro PDS521 I already have.

I have 2 of each, so I could potentially make a stereo pair of...something.

How could I best combine these two compression drivers? I was thinking of aiming for a point-source design, maybe just one horn with a throat combiner, but I'd rather ask than go down too many rabbit holes!
 

Attachments

  • compression-drivers.jpg
    compression-drivers.jpg
    967.1 KB · Views: 384

GM

Member
Joined 2003
Way back when I did ~the latter in a huge WG in that I made the throat as two smaller WGs to load down to 1 kHz. No measurements, just some really good female ears to guide me for the HF with open cell foam to damp reflections back to the CD throats and similar for the WG terminus back to the dual WG terminus.
 
Interesting! I think I have a visual idea of what you'd created in my head - it must have been massive! How complicated a build was it?

The only sane thought I've had apart from the throat combiner was to have the PDS521 run as the HF, and the DU-100 run as the low-mid as a side-tap in the Synergy fashion.

The PDS521 can cross as low as 500Hz (not necessary), and the DU-100 appears to be good from 110Hz - 6kHz, so there's plenty of crossover frequency options.

I'll run the end solution with subs, which I think can crossover at 200Hz.
 
1)The only sane thought I've had apart from the throat combiner was to have the PDS521 run as the HF, and the DU-100 run as the low-mid as a side-tap in the Synergy fashion.
2)The PDS521 can cross as low as 500Hz (not necessary), and the DU-100 appears to be good from 110Hz - 6kHz, so there's plenty of crossover frequency options.
unaHm,

I like the idea of "using what you got", but you really don't have good low-mid potential with what you have.

1)"Synergy" style horns are conical expansion, they provide little low frequency response unless very narrow, (like under 15 degrees) and even then 500 Hz would be pushing the drivers into "quacksville".
If you were to use a more "normal" angle, like 90x40, don't expect much below 1000 Hz from either driver.
Using the PDS521 at the apex with cone mids could do the trick down to 100-200 Hz, but a compression driver like the DU-100 won't.
2) At 200Hz using around a meter long exponential horn the low end of the DU-100 would perhaps be 90dB 1 watt one meter, and would require padding the rest of the range down to that level.

A description of what the intended use and your build complexity comfort level would be helpful to suggest alternatives.

Art
 
How could I best combine these two compression drivers? I was thinking of aiming for a point-source design, maybe just one horn with a throat combiner, but I'd rather ask than go down too many rabbit holes!


For some reason I'm reminded of Hitler's 3rd vengeance weapon, a gun with multiple charges feeding into one very long barrel, precisely timed... And no, it didn't work very well. I fear you may have a slightly analogous problem, but an interesting one. Perhaps a less crazy comparison is with the two lungs feeding the trachea, or perhaps more accurately with the avian syrinx, birds have vocal organs before the airways merge...
 
unaHm,

I like the idea of "using what you got", but you really don't have good low-mid potential with what you have.

1)"Synergy" style horns are conical expansion, they provide little low frequency response unless very narrow, (like under 15 degrees) and even then 500 Hz would be pushing the drivers into "quacksville".
If you were to use a more "normal" angle, like 90x40, don't expect much below 1000 Hz from either driver.
Using the PDS521 at the apex with cone mids could do the trick down to 100-200 Hz, but a compression driver like the DU-100 won't.
2) At 200Hz using around a meter long exponential horn the low end of the DU-100 would perhaps be 90dB 1 watt one meter, and would require padding the rest of the range down to that level.

A description of what the intended use and your build complexity comfort level would be helpful to suggest alternatives.

Art

Thanks for the thoughts Art! Your insight is always valuable, and definitely prompts my need to take a step back and look at what question I'm trying to ask.

For some reason until now, it hadn't really twigged that the low/mid taps in a synergy horn weren't being loaded by the horn...even though it's mentioned a number of times by a number of people (including yourself and bwaslo), so it now makes all the sense in the world why a low-mid compression driver wouldn't work in a similar application.

I like the idea of option 2 that you've mentioned...using a 200Hz horn with that compression driver. I'll be using these devices indoors, so I'm not looking to blast the paint off the walls (of my, er, unfinished basement...maybe deter a few spiders?). I can't say that I've attempted to make a multi-way horn system before. Some of the solutions on this forum are beautiful and fantastic, but I fear that I would get lost in time delay calculations, or just run out of room.

Would it be feasible to stack Smith horns for both drivers? I saw that bwaslo mentioned a number of years ago a passing interest in experimenting with making a Smith horn a synergy, but I have to read up more on how they work.

As for build complexity, I'm not sure - I think I'd like to spend time building a really cool physical enclosure or horn and improving those skills, and then work with a DSP.

For some reason I'm reminded of Hitler's 3rd vengeance weapon, a gun with multiple charges feeding into one very long barrel, precisely timed... And no, it didn't work very well. I fear you may have a slightly analogous problem, but an interesting one. Perhaps a less crazy comparison is with the two lungs feeding the trachea, or perhaps more accurately with the avian syrinx, birds have vocal organs before the airways merge...

Well, that first analogy I'd rather not see pan out :) I think this question stems from seeing those Y adapters that allow you to screw two compression drivers into one throat, but I would imagine that the intended application for those is for two of the same driver, and not differing ones.
 
Last edited:
I like the idea of option 2 that you've mentioned...using a 200Hz horn with that compression driver.

Would it be feasible to stack Smith horns for both drivers?
Regarding the "meter long exponential horn", see the response (and directivity!) of the TOA horn/drivers below.
The TH-650 mouth is a half meter wide, it's folded (re-entrant, reflex) horn length is about 3x394mm, 1.182 meters, about 46 inches. It's sensitivity 1w/1m is about 91 dB at 200 Hz. Note the 1.275 meter TH-660 does only 1 dB more down at 200 Hz.
I couldn't find any meaningful specifications for the DU-100, at best, it might equal that of the TU-631 driver, but could be much (like -3 to -12dB) less.
A (cheap) 6" speaker in a little box would sound far better down that low, and would be around 95dB 1w/1m...

Smith (DSH) horns would work OK, but 500 Hz would be about the bottom for them.

Art
 

Attachments

  • TOA PA Horns.png
    TOA PA Horns.png
    292.9 KB · Views: 230
Last edited:
Agreed - I couldn't find anything meaningful anywhere about them either - even on their website they don't provide a datasheet.

They also appear to make re-entrant horns, so it looks like they're for a specific purpose.

I'll play around with them a bit and see how quacky they get :)

I also have (which might be more interesting) a pair of Yamaha JA-6681B compression drivers, but wanted to be careful with them because I've been told replacement diaphragms can't be found for these any more :)
 
Man, don't bother with those junk drivers when you have the real deal sitting around!
The Smith horn was designed for the type of throat the JA-6681B has, similar to the old JBL, Altec, TAD, etc.

That said, I have a pair of thread on to bolt on adapters you could buy if you want to experiment with the cheapos and change over later without modifying the horn.

Art
 
Sorry to hold out on you Art, that wasn't the intent :)

So with those, if I were to inject some cone drivers into the <insert horn style here>, apart from the crossover and delay compensation, is that pretty much it?

I could see one mounting one or two woofers on the bottom of a Smith horn, using the sealed box also as a stand (I suppose I could try that with the Eminence Alpha 8as I have).

The only concern I have about those compression drivers is that their specs are from 300Hz-12kHz. I'm not sure if that would require another driver to extend up to 20kHz, or whether the horn would provide enough gain to recoup the drop off in response. In my head they didn't quite go low enough, and didn't quite go high enough :)

They have a nice smooth sound to them too. I was also listening to the DU-100s outside of horns yesterday to give them a little break in, and they were quite pleasant to listen to (although harsh to begin with, without EQ) in combination with the Pyle PDS521s.
 
it must have been massive! How complicated a build was it?

The only sane thought.........Synergy fashion.

Quite simple actually, but wouldn't meet the rigorous technical design excellence demanded by today's more avid DIYers.

Right, when Tom hinted a bit about his new [Unity] patent in '98-'99?, my 'one missing piece of the puzzle' WRT to wide BW speaker design was complete.

Anyway, from dim memory and doing a quick sketch to confirm actual inside depth [i.d.]; ~56" W x ~32.5" H x ~28" D, ~500 Hz conical horn to match the dual 15" FLH's terminus/mouth area and loaded with dual 1" CD with the throat divided in two for ~1 kHz conical.

That said, in retrospect it was technically a parabolic flare since one pair of walls [top/bottom] was parallel, though in my defense didn't know this in the early '70s and doesn't make much difference anyway in the scheme of things, just minor frequency response deviations that with no inexpensive way to measure back then, am confident the terminus, throat damping plus CD horn EQ smoothed it right out.
 
So with those, if I were to inject some cone drivers into the <insert horn style here>, apart from the crossover and delay compensation, is that pretty much it?

I could see one mounting one or two woofers on the bottom of a Smith horn, using the sealed box also as a stand (I suppose I could try that with the Eminence Alpha 8as I have).
"Horn gain" won't fix driver deficiencies.

The JA-6681B would go plenty high, the Pyle PDS521 "maybe", the DU-100 not likely, as it probably rolls off with an acoustic band-pass to allow more excursion without the diaphragm hammering the phase plug. You won't want to cross any of them lower than about 800 Hz on a DSH.

The Alpha 8" could be mounted top and bottom on the DSH using multiple entry ports (one per cell), and reach to around 1kHz before the acoustic band-pass cut off. Most all the cone driver horn gain from the DSH will be in the 400-900Hz range, and be peaky, not of much use in a domestic environment.
The injection ports will also have negative impact on HF response in such small exponential horn paths as the DSH has.
Using a DSH for high frequency, you would probably be better served using the Alpha 8 in a standard box. Forget the DU-100, the Alpha 8 will sound better than a compression driver on a DSH (or any horn I can think of) from 100-800Hz.

If you do want to make a MEH (multiple entry horn) you should decide the dispersion you want, then take it from there.

Art
 
"Horn gain" won't fix driver deficiencies.

The JA-6681B would go plenty high, the Pyle PDS521 "maybe", the DU-100 not likely, as it probably rolls off with an acoustic band-pass to allow more excursion without the diaphragm hammering the phase plug. You won't want to cross any of them lower than about 800 Hz on a DSH.

The Alpha 8" could be mounted top and bottom on the DSH using multiple entry ports (one per cell), and reach to around 1kHz before the acoustic band-pass cut off. Most all the cone driver horn gain from the DSH will be in the 400-900Hz range, and be peaky, not of much use in a domestic environment.
The injection ports will also have negative impact on HF response in such small exponential horn paths as the DSH has.
Using a DSH for high frequency, you would probably be better served using the Alpha 8 in a standard box. Forget the DU-100, the Alpha 8 will sound better than a compression driver on a DSH (or any horn I can think of) from 100-800Hz.

If you do want to make a MEH (multiple entry horn) you should decide the dispersion you want, then take it from there.

Art

Hmm. So the scenario I'm thinking of is this:
* Indoor application
* A design mostly for a single listener, so toe-ing in the speakers to converge on a sweet spot is allowed
* Wide dispersion not entirely necessary. Maybe 60x60, 60x40?
* Distance from speaker to listener = 6-8 feet

Thanks for the explanation of why it would be ill-advised to horn-load the cone driver into the DSH. We definitely don't want peaky!

The funny thing is that I used the PDS521 with a narrow 1kHz horn some years ago, and the Alphas in a box (as prescribed by Eminence) below them. They sounded very good, but I felt like I was missing out on a lot of cool experiences with horns :D

Those DU-100s we can't expect them to do any form of high frequency (if we had an application for them). Their frequency response is only up to 6kHz, so they're definitely built for a specific purpose (which may or may not work to our benefit).
 
* A design mostly for a single listener, so toe-ing in the speakers to converge on a sweet spot is allowed
* Wide dispersion not entirely necessary. Maybe 60x60, 60x40?
* Distance from speaker to listener = 6-8 feet

Thanks for the explanation of why it would be ill-advised to horn-load the cone driver into the DSH. We definitely don't want peaky!

The funny thing is that I used the PDS521 with a narrow 1kHz horn some years ago, and the Alphas in a box (as prescribed by Eminence) below them. They sounded very good, but I felt like I was missing out on a lot of cool experiences with horns :D
A "narrow 1kHz horn" does not give enough information for me relate to your "experience", or what you are missing. Response to 1kHz and pattern control to 1kHz are two different considerations.

As a guess, the Pyle PDS521 is probably as peaky as the Eminence PSD2002. If you did not EQ out the peaks, your horn experience would not be very "cool" :cool:

As an example, see the response of a Eminence PSD2002 (16ohm) compared to a BMS 4550 (16ohm) on the same 13x13 degree Maltese conical horn.
Below 1KHz, the response "ripple" from the long horn is evident, but response varies by 9dB at 300 Hz.
Above 1000Hz, 6dB variations at 2kHz, 7kHz, 11kHz, then the PSD2002 takes a bow, while the BMS 4550 continues to "zing" up at 18kHz.
Without EQ, listening to those two drivers would be quite "different experiences";)

The PSD2002 went through a diaphragm redesign over 20 years ago.
On a "90x40" horn (of bastard design..), the two are compared.
"Em1" is the older, but smoother "Series 1" diaphragm, which could be equalized with one filter, while the newer, "brighter, extended response" version would need about 7 parametric filters to tame it's response.

Since you plan to use DSP, are going three way, have the Alpha 8s in boxes, (and a bunch of poly cone "subs"?) just tune up what you have, then tell us what the experience is missing.

Art
 

Attachments

  • PSD2002, BMS 4550.png
    PSD2002, BMS 4550.png
    139.2 KB · Views: 141
  • PSD 2002 series1 &2.png
    PSD 2002 series1 &2.png
    103.6 KB · Views: 139
Since you plan to use DSP, are going three way, have the Alpha 8s in boxes, (and a bunch of poly cone "subs"?) just tune up what you have, then tell us what the experience is missing.

Right now, the only experience I have is a pair of Sennheiser headphones :) We moved house a couple of times, so I suppose I have a clean slate!

Apart from the Yamaha drivers, maybe I should cut my losses and build a pair of CoSynes, or SyntripPs, both designed by some wizards I've heard about here and there ;)
 
Last edited:
I think we can end this thread - given that I have the Yamaha compression drivers, and I can cross between them and the subs with the Alpha 8a speakers, I'll look up some designs that people have already done here to get ideas and go from there. Art, thanks for your help and being the voice of reason :) If I come up with something, I'll post a build thread of that specific project.
 
While I wait for the opportunity to go and get some plywood for the 'real deal', I went out on a whim and had a go at putting together a non-serious design using the Pyle, the Miyako and the Eminence drivers as a 3-way.

I focused on the compression drivers, attaching them to a 500Hz horn. See the attached picture.

Here's a list of things I noticed once I'd got the DSP together:
  • The Pyle driver was definitely more sensitive than the Miyako, which wasn't much of a surprise
  • The Miyako had a wider range than expected, though I found it performed best between 250Hz and 2kHz (it appeared to roll off aggressively after 3kHz in this configuration)
  • I had to turn the Miyako down a little
  • After listening to the Yamaha compression drivers, the Pyles were definitely 'brittle' in their sound by comparison
  • Regardless of whether I used the Eminence or one of the Polycone woofers, it was pretty easy to get a Bass driver to work with them, even without an enclosure for it
  • I 'bifurcated' the Miyako's output by creating a channel leading from the centre of the panel to the sides which opened to half of the exit cross-sectional area on either side (hopefully it can be seen on the bottom panel of the horn). No science (just inspired by the paraline thread, though what I did bears little resemblance)...I just wanted to see what it sounded like. It worked better than I expected, but I wasn't expecting much!

I might build a second one to see how they work in a stereo configuration, as I remember the foam core 2-way horns I made ages ago benefited from being a pair. It was fun to experiment and play around in the physical realm :)
 

Attachments

  • 3-way-nonsense.jpg
    3-way-nonsense.jpg
    234.7 KB · Views: 107