Hi Guys
A few thoughts;
Sheldon, the corner mounting works fine, the important part is having the entry holes in or near the corners. About 14 years ago now when I made the very first Unity horn, I mounted the driver in the horn, while that illustrated the theory I had about the position dependent expansion rate, it made a mess of the hf response. That lead to the holes.
The smaller you can make the holes, the smaller the effect is on the hf driver and if they are small and in the corners, they have close to no effect on the hf. As opposed to some speculation I have read, one can simply measure the hf drivers response and impedance before and after drilling the holes / mounting the mids and see.
The holes (the port it becomes) and the trapped volume under the cone comprise an acoustic low pass filter, the smaller you make the holes, the lower that corner F is, the shorter you can make the duct part of that hole, the higher the corner F is.
While attenuating the natural distortion the driver would produce (falling above the cutoff) a second hf limitation exists for side mounted drivers which is caused by a series of cancellation notches which occur when the distance of the holes to the acoustic ‘end” of the horn is a quarter wl away (the sound reflects back and arrives delayed two quarter wl’s or 180 degrees. This also acts like a low pass filter so one must consider these the upper limit for that driver. On the other hand, the harmonic distortion can be made very low.
Patrick “Mark Seaton's advice from five years ago was ignored by yours truly. He basically said "find any ol' small closed back midrange."”
Mark was not steering you wrong at least for a first shot. You want the Fb to be around the middle of the operating band.
“I should stop doing so many sims and start making more sawdust “
If you want to understand how the Unity / Synergy horn works, you will have to do some “what if” experiments where you change only one thing and consider the measured results.
I do simulate things, but when working on something new where there are no rules and no easy simulation, I figure out “how it works” by thinking, building, measuring in a loop until some “rules” surface or I give up..
So far as the crossover, the idea even back with the Unity was to make the sources add into one acoustic source, with the latest generation the Synergy horn, it is often possible to do the latter but also reduce or eliminate the phase shift that normally accompanies higher order crossovers and that results in what looks like just one wide band driver.
I hoped we could have had a “hifi” speaker by now but being able to combine multiple drivers into one source has turned out to be valuable in large scale sound and these larger sources have occupied all my time. By fall there will be 14 stadiums using these larger Synergy horns.
The break though here was in finding a way to combine multiple hf drivers into the radiation shape of a single source (sort of a lens). I have one I am working on now that has 42 drivers.
If you go to football or baseball games, you might be able to hear some of these larger systems, I got back from Lambau field in Green Bay a couple hours ago.
The cool part is that the large conventional arrays that are so popular normally radiate an interference pattern so if the wind blows, the sound changes from all the combs.
With these large synergy horns, the drivers still add coherently, there is no interference pattern and so they sound fine in a strong wind or if you move around. With the large arrays, the frequency also response changes significantly with distance, with a single point source, the response doesn’t change at all except for air absorption at very large distances. A speaker like the JH-90 can produce measured 102dBA slow weighted at 700 feet playing “hifi” music.
Here is a camcorder recording which with headphones illustrates (sort of) the single source effect at a large distance (measured the next day to be 700 feet at the farthest point)..
Danley Sound Labs Jericho Horn playing Jennifer Warnes track.MOV - YouTube
And a couple places where they are already installed;
ONE DANLEY JERICHO HORN COVERS 30,000 FANS AT TROY UNIVERSITY | BriefingRoom on MixOnline
THIS IS NOT YOUR FATHER’S ANEMIC BOWL STADIUM SOUND. DANLEY JERICHO HORNS AND SUBS DELIVER FULL FIDELITY TO BRIGHAM YOUNG UNIVERSITY FOOTBALL FANS | BriefingRoom on MixOnline
Two Danley Jericho Horns Cover 47,000 Seats at Northwestern University - ET Now USA
I wish there was a way to let you guys hear what the latest designs sounded like and not the old Unity horn from 10 years ago. I had been working towards a kit of some sort but the Neo magnet crisis and driver availability with the growth of the large scale stuff put the kabosh on that for the time being.
Best,
Tom Danley
Danley Sound Labs
Hi GM, John
A few thoughts;
Sheldon, the corner mounting works fine, the important part is having the entry holes in or near the corners. About 14 years ago now when I made the very first Unity horn, I mounted the driver in the horn, while that illustrated the theory I had about the position dependent expansion rate, it made a mess of the hf response. That lead to the holes.
The smaller you can make the holes, the smaller the effect is on the hf driver and if they are small and in the corners, they have close to no effect on the hf. As opposed to some speculation I have read, one can simply measure the hf drivers response and impedance before and after drilling the holes / mounting the mids and see.
The holes (the port it becomes) and the trapped volume under the cone comprise an acoustic low pass filter, the smaller you make the holes, the lower that corner F is, the shorter you can make the duct part of that hole, the higher the corner F is.
While attenuating the natural distortion the driver would produce (falling above the cutoff) a second hf limitation exists for side mounted drivers which is caused by a series of cancellation notches which occur when the distance of the holes to the acoustic ‘end” of the horn is a quarter wl away (the sound reflects back and arrives delayed two quarter wl’s or 180 degrees. This also acts like a low pass filter so one must consider these the upper limit for that driver. On the other hand, the harmonic distortion can be made very low.
Patrick “Mark Seaton's advice from five years ago was ignored by yours truly. He basically said "find any ol' small closed back midrange."”
Mark was not steering you wrong at least for a first shot. You want the Fb to be around the middle of the operating band.
“I should stop doing so many sims and start making more sawdust “
If you want to understand how the Unity / Synergy horn works, you will have to do some “what if” experiments where you change only one thing and consider the measured results.
I do simulate things, but when working on something new where there are no rules and no easy simulation, I figure out “how it works” by thinking, building, measuring in a loop until some “rules” surface or I give up..
So far as the crossover, the idea even back with the Unity was to make the sources add into one acoustic source, with the latest generation the Synergy horn, it is often possible to do the latter but also reduce or eliminate the phase shift that normally accompanies higher order crossovers and that results in what looks like just one wide band driver.
I hoped we could have had a “hifi” speaker by now but being able to combine multiple drivers into one source has turned out to be valuable in large scale sound and these larger sources have occupied all my time. By fall there will be 14 stadiums using these larger Synergy horns.
The break though here was in finding a way to combine multiple hf drivers into the radiation shape of a single source (sort of a lens). I have one I am working on now that has 42 drivers.
If you go to football or baseball games, you might be able to hear some of these larger systems, I got back from Lambau field in Green Bay a couple hours ago.
The cool part is that the large conventional arrays that are so popular normally radiate an interference pattern so if the wind blows, the sound changes from all the combs.
With these large synergy horns, the drivers still add coherently, there is no interference pattern and so they sound fine in a strong wind or if you move around. With the large arrays, the frequency also response changes significantly with distance, with a single point source, the response doesn’t change at all except for air absorption at very large distances. A speaker like the JH-90 can produce measured 102dBA slow weighted at 700 feet playing “hifi” music.
Here is a camcorder recording which with headphones illustrates (sort of) the single source effect at a large distance (measured the next day to be 700 feet at the farthest point)..
Danley Sound Labs Jericho Horn playing Jennifer Warnes track.MOV - YouTube
And a couple places where they are already installed;
ONE DANLEY JERICHO HORN COVERS 30,000 FANS AT TROY UNIVERSITY | BriefingRoom on MixOnline
THIS IS NOT YOUR FATHER’S ANEMIC BOWL STADIUM SOUND. DANLEY JERICHO HORNS AND SUBS DELIVER FULL FIDELITY TO BRIGHAM YOUNG UNIVERSITY FOOTBALL FANS | BriefingRoom on MixOnline
Two Danley Jericho Horns Cover 47,000 Seats at Northwestern University - ET Now USA
I wish there was a way to let you guys hear what the latest designs sounded like and not the old Unity horn from 10 years ago. I had been working towards a kit of some sort but the Neo magnet crisis and driver availability with the growth of the large scale stuff put the kabosh on that for the time being.
Best,
Tom Danley
Danley Sound Labs
Hi GM, John
Its nowhere as long lasting or reliable as treated cloth surrounds. Why build something as difficult as a Unity/Synergy if the surrounds rot in 5-10 years? Perfect example is the Bose 901 speakers. The ones with cloth surrounds are still going after 30 years. The ones with foam surrounds can normally be found by the curb on trash day.
Great post. Thanks Tom. That the first time I've heard the distance issue stated so intuitively. Both the issues above would favor corner mounting over side mounting - I think: The duct can be made with essentially zero depth, and moved closer to the apex.
However, I'd love to scotch that idea in favor of the new generation. I don't mind buying that proprietary info either (assuming it's in a state to be applied).
Sheldon
Tom - Is there an innate difference between the physical configuration of the Unity and Synergy horns that prevents the Unity from implementing a phase linear crossover? Would it be possible, with correct driver selection and crossover design, to achieve a phase linear response using the smaller Unity horn format?
Ever since I heard the SH50 in the demo room at InfoComm 3 years ago, I have been fascinated by your designs. I went on to sell and install 8 SH50's in two schools in my area and everyone who hears them is amazed that a sound system can sound so good in a gymnasium. I'm out of contracting and now am the TD for a large church. I am convinced that the Synergy horn concept is the closest anybody has yet to come to the ideal acoustic point source. I would kill for a pair in my living room or my recording studio, but they're just out of reach. Which is why I've been testing the DIY route. Thanks so much for taking time to contribute to the DIY community.
-Justin
Mine is just a quick approximation of an offset driver 250 Hz conical horn since it doesn't need to be exact to compare driver specs. Anyway, I have a bit different view about certain design details from what you used that apparently has been incorporated into the Synergy concept, so until TD either gets a patent or it becomes obvious he won't I prefer to leave it to those folks sufficiently 'skilled in the art' to figure it out as they [use to?] commonly say in patents.
GM
Sounds like he was quoting TD from a HE forum thread somewhat earlier when folks kept pestering him for the original Unity design details, i.e. if its Fs is around 500 Hz in a 5" sealed back mids, odds are the rest of the specs will be 'close enough' to adapt it.
FWIW, the POC I cobbled together used some old RS ~90 Hz Fs drivers with the baskets all cardboard/duct taped up and piezo tweeter horn in a foam core board horn and was rather surprised by its raw potential, so made me a convert long before I was able to get a Unity audition, proving to my satisfaction that the concept is a viable one even if done crudely and on the 'cheap'; though of course I don't recommend doing it as crudely/cheaply as I did except for experimental purposes as it took significant digital processing to tie it all together 'good enough' and couldn't play loud due to the super lossy horn construction.
GM
Yes the more expensive well respected names have solved most of the foam rot problems. However, I have doutbs that the products in the price range of $5 to $10 a driver. I have a few Tang Band 4" full range speakers that have rotted and they are only 6 years old. I'm sure Tang Band has fixed this because they are now one of the better speaker manufacturers.
Rgs, JLH
Is anyone planning on keeping one diy speaker for long enough for the surrounds to rot? Long before mine rot I probably will have finished, tweaked, enjoyed and then sold them. I'd be more concerned about half roll surrounds being less than ideal for midrange due to less control of the cone edge. Can't think of a reason for such a surround in a speaker with virtually no xmax. Still the Hornresp sims look good. I did a comparison on my blog:
Red Spade Audio: Synergy horn driver selection
I don't know the xmax of the Pyle driver, so used a value the same as the Misco. I'm also not sure of the inductance of the Misco driver, doesn't seem to be included on the data sheet. A lower value can bring the sensitivity up at 1k by about 1db. The Misco has about 2db more output, but either way I can hit THX reference levels @ 3m with 10 or 12 db of headroom. I think that's enough!
Red Spade Audio: Synergy horn driver selection
I don't know the xmax of the Pyle driver, so used a value the same as the Misco. I'm also not sure of the inductance of the Misco driver, doesn't seem to be included on the data sheet. A lower value can bring the sensitivity up at 1k by about 1db. The Misco has about 2db more output, but either way I can hit THX reference levels @ 3m with 10 or 12 db of headroom. I think that's enough!
Sawdust
I started making sawdust this weekend. I ordered eight of the GRS 5SMB-8 mids and two Selenium D220Ti. I chose the GRS mids for a few good reasons: they have a lower inductance and lower QTS than both the Pyle and Misco, and they model real well... oh, and they're crazy cheap! Even if they don't work I'm only out $40 for all eight drivers. The only parameter missing was the VAS and I modeled it within what I thought was a realistic range and it only changes the low cutoff by about 50hz as well as the shape of the roll off.
In the true DIY spirit I dove in head first. I haven't decided yet if I want to just do an active or passive crossover. I've never designed a passive, but I've got the measurement equipment to do it. I'll build one of the horns completely, test and tweak it until I think it's right and then build the other one out. I'll probably start a build thread once things get rolling.
I started making sawdust this weekend. I ordered eight of the GRS 5SMB-8 mids and two Selenium D220Ti. I chose the GRS mids for a few good reasons: they have a lower inductance and lower QTS than both the Pyle and Misco, and they model real well... oh, and they're crazy cheap! Even if they don't work I'm only out $40 for all eight drivers. The only parameter missing was the VAS and I modeled it within what I thought was a realistic range and it only changes the low cutoff by about 50hz as well as the shape of the roll off.
In the true DIY spirit I dove in head first. I haven't decided yet if I want to just do an active or passive crossover. I've never designed a passive, but I've got the measurement equipment to do it. I'll build one of the horns completely, test and tweak it until I think it's right and then build the other one out. I'll probably start a build thread once things get rolling.
Attachments
GM
What do you think about using perforated metal (grill type) with suitably sized holes as a port screen for the midranges that would form part of the horn wall (flush) so that I could use a larger than normal slot port. Would this introduce a ragged response from the high frequency waves of the compression driver as they travelled across its surface? Keep in mind I'm only using mid drivers on the top and bottom horn walls.
What do you think about using perforated metal (grill type) with suitably sized holes as a port screen for the midranges that would form part of the horn wall (flush) so that I could use a larger than normal slot port. Would this introduce a ragged response from the high frequency waves of the compression driver as they travelled across its surface? Keep in mind I'm only using mid drivers on the top and bottom horn walls.
The coupling chamber isn't just there to increase efficiency and roll off the distortion. It also delays the output of the midranges.
This is basically required for the design to work, otherwise the output from the midranges will 'lead' the tweeter by quite a bit, since the gap is a good six inches.
If you simply bolt the drivers onto the horn, with nothing more than a perforated grill, you won't get that delay.
You can model this in bandpass boxmodel (free, google it) or hornresp. Bandpass boxmodel is dated, but easier to use IMHO. (Not that hornresp is bad, it's just not for everyone, because the learning curve is steeper than a conventional loudspeaker modeling program.)
Ok , so I wasn't clear enough
, I've got the idea of the band pass and the port for it, and how it interacts in the horn.... but because I want to use a larger( 8") sealed back mid which will inherently need a larger port/(slot), this scenario dictates that I put my large port within 2 1/4" of the compression driver.
This also will cause me the problem of having a large divot right @ the throat which of course will kill the HF response , therefore my idea of using perforated grill @ the point where the bandpass port meets the horn came to be. I just would like someones educated thoughts on how the grill will affect the HF waves as they traverse it.
, I've got the idea of the band pass and the port for it, and how it interacts in the horn.... but because I want to use a larger( 8") sealed back mid which will inherently need a larger port/(slot), this scenario dictates that I put my large port within 2 1/4" of the compression driver.This also will cause me the problem of having a large divot right @ the throat which of course will kill the HF response , therefore my idea of using perforated grill @ the point where the bandpass port meets the horn came to be. I just would like someones educated thoughts on how the grill will affect the HF waves as they traverse it.
So you want the grille to be flush with the surface of the horn, to appear to the HF as a closer approximation of the holes not being there, yet transparent to the mids. Interesting idea, don't know if it would really work, but perhaps you could try it. You simply try one thing at a time and measure it. Grille, no grille, different grilles. Foam might be another thing to try.
There would be two effects - one would be due to diffraction off the holes which I imagine would improve if you used a perforated metal plate with enough surface area. The other effect would be on the acoustic notch filter of the holes and midrange front chambers in parallel with the horn's impedance from the holes out to the mouth. The screen would put a resistance in series with the hole and front chamber. Then you'd need to look at the same system from the point of view of the midranges to see what the smaller hole size did to air velocity and what the resistance did to the response.
I pondered this about a decade ago with an ultra-fi Unity concept that a [mostly] lurker was going to build. He got the basic horn done, but then scored a set of OEM mids, so what I wanted him to build/test didn't happen. Anyway, from the experiments I'd done WRT high performance acoustic horns and intake/exhaust systems, among other things I'd concluded that there was no such thing as too smooth a bore and that any discontinuity caused multiple disruptions [reflections], so the Unity's deep holes didn't appeal to me even stuck in the corners away from the main 'thrust' of the horn's acoustic 'piston'.
With the driver I had in mind, I felt I needed a super thin slot vent for a certain amount of resistive loading, so my plan was to use thin metal inserts that I could hone to a knife edge with whatever the fine screen mesh used in the Altec large format drivers as a bug screen tightly stretched over them to further increase it to a ~aperiodic loading and act as a sue-do horn wall 'patch'. Of course these must be super flush with the horn surface.
My feeling is that the HF will 'feel' it as more like a fine layer of damping material rather than a true discontinuity, at relative to what's emanating from it figuring a worst case scenario was the HF would have to be run a little 'hot' to make up for any extreme HF losses, then damped with a foam throat insert same as my Altec 511 horns need for other reasons.
GM
Yes Paul that's exactly it, and thanks GM I will build mine with that in mind, maybe I will try different types of perf grill with different hole sizing. I guess I will have to calculate the free open space of all the little holes and make sure they total up to the original calculated port opening size.