The 30 (half case) Misco RDC3T-A showed up today. What a fantastic little driver. The workmanship is very good. Oh this is going to be good. Once back from spring vacation I'll get started on turning my Akabak script into real life. I'm thinking I'll start a new thread in about a month. I don't think I'll have much to share until then. Still need to order the measurement equipment and learn how to use it.
Phase curve?
Hi JLH,
Could you please explain to me why the phase curve in your akabak sim looks so awfully good? If you look at the synergy you see +- 180 degree in the higher operating range, which one would expect (which still is very good for a threeway) But is you look at your sim, it only has +- 60 degree phase shift from 150Hz upwards, which is a little too good.. Are you using linear phase x-overs in this sim?
As I understand it, the synergy is three way, the bandpass chambers of the mids and the lows would cause 180 degrees, and an electronically 2nd order x-over would cause another 180 degrees (per x-over). I understand that you can gain something because the lower drivers are placed ahead in time physically, but I can't imagine how you could get the phase so good. Am I missing something here? Please explain cause I am very interested in your design and if you really can accomplish this I will certainly want to build them too!
Thanks in advance,
Kessito
Hi JLH,
Could you please explain to me why the phase curve in your akabak sim looks so awfully good? If you look at the synergy you see +- 180 degree in the higher operating range, which one would expect (which still is very good for a threeway) But is you look at your sim, it only has +- 60 degree phase shift from 150Hz upwards, which is a little too good.. Are you using linear phase x-overs in this sim?
As I understand it, the synergy is three way, the bandpass chambers of the mids and the lows would cause 180 degrees, and an electronically 2nd order x-over would cause another 180 degrees (per x-over). I understand that you can gain something because the lower drivers are placed ahead in time physically, but I can't imagine how you could get the phase so good. Am I missing something here? Please explain cause I am very interested in your design and if you really can accomplish this I will certainly want to build them too!
Thanks in advance,
Kessito
Once I got the compression drivers and mids in I discovered that what I had modeled in Akabak was not physically possible. I’ve had little time to rework the script to reflect the real world measurements of the components. I hope to get moving as my work settles down again. I’ve been on reassignment for two months because another department can’t seem to get their crap together. I still plan to have a working Synergy horn by winter.
Because I designed to be that way.
Its all about looking at the phase relation between the drivers due to their physical spacing while taking into consideration the phase rotation you will get after you add the crossovers. Its a balancing act and takes a lot of simulation time.You are thinking in too hard of terms. The phase is always changing with frequency. Its not strictly just 180 degrees or this or that - its continiously changing. There is a far amount of overlap in my crossovers. Once I get the crossover points and physical position of the drivers phase coherent, I then work on making the crossovers as phase linear (minimal phase) as possible without messing up the frequency response. Its is not an easy task and takes lots of time and experence. That's as good as I can explain it - this subject is far too complex to explain on a forum.
Thanks for the update JLH.
JLH or others,
just as an experiment... How large would it be possible to make a Synergy horn before the highs would begin to suffer? I mean the hard part of the Synergy horn is the design, entry holes, back chambers crossovers etc. If you've got the room, there would be no problem to make a large conical horn. You would gain directivety control and horn loading further down in frequency, but would it be a downside to it?
JLH or others,
just as an experiment... How large would it be possible to make a Synergy horn before the highs would begin to suffer? I mean the hard part of the Synergy horn is the design, entry holes, back chambers crossovers etc. If you've got the room, there would be no problem to make a large conical horn. You would gain directivety control and horn loading further down in frequency, but would it be a downside to it?
A few months back I bought a cheap set of JBL speakers at Costco, mostly out of curiosity. I ran them for a few months, then switched back to my reference. (Gedlee Summas.) While everything about the Summa is better, there's one thing the JBL does that I miss - the top octave is better. The compression driver in the Summa has virtually no output above sixteen kilohertz.
I know this is a non-issue for most; but I can easily hear 20khz. (I've run tone tests on headphones, it's quite audible for me.)
So I was a bit interested in this speaker from Peavey:
It's no Unity horn, but it is similar. A unity horn is a three way line array that's been wrapped around a conical horn in three dimensions. The Peavey design is a three way line array that's been wrapped around a horn in *two* dimensions.
It's interesting that they're using ribbons at the apex.
A few years ago I tried a ribbon on a horn, and it worked very nice. While ribbons are quite fragile, their excellent cumulative spectral delay is a nice match for a horn or a waveguide, because the horn increases their durability by reducing excursion. And a ribbon can easily play to 30khz on a horn, while a compression driver will have a tough time.
Here's the description from the manual. BTW they're $8000 a pair at Parts Express.
"The Versarray 212 Ribbon Tweeter Line Source Array module consists of 2 dual 12” Neo Black Widow woofers combined with a neodymium-based, Peavey-exclusive midrange line array and planar ribbon tweeter line source in a cabinet with a highly flexible rigging system. Designed to provide modular coverage of medium to large venues and intended for use with the companion Versarray Sub models, the Versarray 212 offers extreme versatility and high performance.
The three-way system consists of the following driver components: two 12” Black Widow Neo series woofers with neodymium magnet structure. The woofers are capable of over 500W of continuous power handling (AES Std 2-1984) each.
The midrange is handled by ten 2.5” Neo magnet midrange drivers, providing a total of 200W of continuous power handling, at a high sensitivity of 101 dB. The high frequencies are handled by four Peavey-exclusive planar ribbon tweeters utilizing a neodymium magnet system, firing into a manifold and creating a line source mounted to a low distortion waveguide. Capable of 200W continuous power handling, the tweeter line source array provides a crystal-clear high end."
http://www.parts-express.com/pe/showdetl.cfm?partnumber=248-8106&utm_source=googleps
I know this is a non-issue for most; but I can easily hear 20khz. (I've run tone tests on headphones, it's quite audible for me.)
So I was a bit interested in this speaker from Peavey:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
It's no Unity horn, but it is similar. A unity horn is a three way line array that's been wrapped around a conical horn in three dimensions. The Peavey design is a three way line array that's been wrapped around a horn in *two* dimensions.
It's interesting that they're using ribbons at the apex.
A few years ago I tried a ribbon on a horn, and it worked very nice. While ribbons are quite fragile, their excellent cumulative spectral delay is a nice match for a horn or a waveguide, because the horn increases their durability by reducing excursion. And a ribbon can easily play to 30khz on a horn, while a compression driver will have a tough time.
Here's the description from the manual. BTW they're $8000 a pair at Parts Express.
"The Versarray 212 Ribbon Tweeter Line Source Array module consists of 2 dual 12” Neo Black Widow woofers combined with a neodymium-based, Peavey-exclusive midrange line array and planar ribbon tweeter line source in a cabinet with a highly flexible rigging system. Designed to provide modular coverage of medium to large venues and intended for use with the companion Versarray Sub models, the Versarray 212 offers extreme versatility and high performance.
The three-way system consists of the following driver components: two 12” Black Widow Neo series woofers with neodymium magnet structure. The woofers are capable of over 500W of continuous power handling (AES Std 2-1984) each.
The midrange is handled by ten 2.5” Neo magnet midrange drivers, providing a total of 200W of continuous power handling, at a high sensitivity of 101 dB. The high frequencies are handled by four Peavey-exclusive planar ribbon tweeters utilizing a neodymium magnet system, firing into a manifold and creating a line source mounted to a low distortion waveguide. Capable of 200W continuous power handling, the tweeter line source array provides a crystal-clear high end."
http://www.parts-express.com/pe/showdetl.cfm?partnumber=248-8106&utm_source=googleps
Last edited:
An Unity cabinet with a ribbon (or a planar, for that matter) and cone midranges in a vertical array would probably gie the best results with a modern horn profile that can manage the transition from the rectangle area of the throat to a elliptical rounded termination.
There are a number of such designs, derived by jzagaza from leCleach, Geddes or Iwatta profiles.
Yeah, definitely interesting.. There could be a problem though. I've heared at some point that horns have a natural tendency to roll back the HF as a function of their length. Is this true?
There are a number of such designs, derived by jzagaza from leCleach, Geddes or Iwatta profiles.
Yeah, definitely interesting.. There could be a problem though. I've heared at some point that horns have a natural tendency to roll back the HF as a function of their length. Is this true?
Yes, all horns act as an acoustic low pass filter.
As I see it, a ribbon has two advantages in a horn. First, a ribbon can reproduce ultra high frequencies, and compression drivers struggle to do this. Second, ribbons have very good CSD, due to their very low mass.
Last night I watched some movies on my cheap JBLs and it reminded me of how good their upper frequency extremes are. It's noticeably better than the top octave of my reference. (Keep in mind, my reference excels in all other respects, except for cost )
The downside of a ribbon is that it might produce a lot of HOMs. Based on what I've read from Geddes, this is because the shape of the diaphragm isn't a good match for the waveguide. If I'm not mistaken, a vertical array complicates this, because a vertical array is theoretically infinite in height. (And of course, this is a falsehood in the real world because it's an invalid assumption.)
So a ribbon may be a viable alternative, or it may not. I think the jury is still out.
Another alternative to a ribbon is to use a smaller compression driver. This extends the high frequency response and improves CSD. I think a smaller compression driver is a viable alternative, and that's exactly what I've done in my designs.
The use of a smaller compression driver creates a problem in a Unity horn, because it makes it very difficult to get the midranges to play high enough to meet the compression driver. You wind up with a gap between the midrange and the tweeter.
A solution to *that* problem is to use smaller midranges, which is what I've done in my designs.
Another option is to use a coaxial at the throat of the horn. Which is exactly whats done in all the synergy horns!
I'd love to hear the Danley SM-60f. IMHO, the SM-60F paired with a set of subs is one of about five or ten speakers that are great candidates for listening to music and home theater at home. (I think many of the larger Unity and Synergy horns are impractically large for a home.)
SM-60F - note how small it is
SH-60 - note that a set of these are almost the size of a big screen TV. Ugh.
As I see it, a ribbon has two advantages in a horn. First, a ribbon can reproduce ultra high frequencies, and compression drivers struggle to do this. Second, ribbons have very good CSD, due to their very low mass.
Last night I watched some movies on my cheap JBLs and it reminded me of how good their upper frequency extremes are. It's noticeably better than the top octave of my reference. (Keep in mind, my reference excels in all other respects, except for cost
)The downside of a ribbon is that it might produce a lot of HOMs. Based on what I've read from Geddes, this is because the shape of the diaphragm isn't a good match for the waveguide. If I'm not mistaken, a vertical array complicates this, because a vertical array is theoretically infinite in height. (And of course, this is a falsehood in the real world because it's an invalid assumption.)
So a ribbon may be a viable alternative, or it may not. I think the jury is still out.
Another alternative to a ribbon is to use a smaller compression driver. This extends the high frequency response and improves CSD. I think a smaller compression driver is a viable alternative, and that's exactly what I've done in my designs.
The use of a smaller compression driver creates a problem in a Unity horn, because it makes it very difficult to get the midranges to play high enough to meet the compression driver. You wind up with a gap between the midrange and the tweeter.
A solution to *that* problem is to use smaller midranges, which is what I've done in my designs.
Another option is to use a coaxial at the throat of the horn. Which is exactly whats done in all the synergy horns!
I'd love to hear the Danley SM-60f. IMHO, the SM-60F paired with a set of subs is one of about five or ten speakers that are great candidates for listening to music and home theater at home. (I think many of the larger Unity and Synergy horns are impractically large for a home.)
SM-60F - note how small it is
SH-60 - note that a set of these are almost the size of a big screen TV. Ugh.
There is no 'money seat' with a good constant directivity design. Stand up, sit down, walk around, it all sounds the same.
It's kind of eerie really, with my Summas you wouldn't even be able to pinpoint the speaker if you couldn't see it.
William Cowan put his Unitys into his wall, which seems like a great solution.
It's kind of eerie really, with my Summas you wouldn't even be able to pinpoint the speaker if you couldn't see it.
William Cowan put his Unitys into his wall, which seems like a great solution.
My recollection of hearing Williams were that they were incredibly uniform in tonal spectrum no matter where in the room you were (and it's larger than it looks in the pics).
Yes, all horns act as an acoustic low pass filter.
As I see it, a ribbon has two advantages in a horn. First, a ribbon can reproduce ultra high frequencies, and compression drivers struggle to do this. Second, ribbons have very good CSD, due to their very low mass.
Last night I watched some movies on my cheap JBLs and it reminded me of how good their upper frequency extremes are. It's noticeably better than the top octave of my reference. (Keep in mind, my reference excels in all other respects, except for cost
The downside of a ribbon is that it might produce a lot of HOMs. Based on what I've read from Geddes, this is because the shape of the diaphragm isn't a good match for the waveguide. If I'm not mistaken, a vertical array complicates this, because a vertical array is theoretically infinite in height. (And of course, this is a falsehood in the real world because it's an invalid assumption.)
So a ribbon may be a viable alternative, or it may not. I think the jury is still out.
Another possible problem with a ribbon or planar is how it would handle the sound pressure from the mids and bass drivers loading the same horn.
The SH100 uses a B&C 8" cone/1" cd coax, the SM60 a BMS 5CN140 5" cone/1" cd coax, and the Jericho horns the BMS 459x family cd/cd coax. The other Synergy horns use ordinary (not coax) 1" or 1,4" cd's.
Eerie timing on this 'ribbon as the tweeter in a Unity' discussion, since just last night I threw out the test mule I used on a prototype along these lines some years ago. I used planars rather than ribbons - in this case some pulls from Monsoon planar desktop speakers - and built a 2-d segmented approximation to an OS profile. I think I posted some measurements over at HTGuide - I'll have to try to search for the thread.
My experiment worked really well, aside from the fact that they dropped like a rock at 10k. Patricks concern about HOM is - I think - not likely to be a huge problem with planar drivers. If I remember correctly from Geddes book, a diaphragm that is clamped at the side-wall of the horn is actually a GOOD match for an OS horn profile. So, assuming you could find a planar driver of the correct dimensions, it may actually be pretty close to ideal. Even the B&G Neo3 might be too wide for this, but a line of them would be an interesting place to start. The Neo8 is roughly the same size as the Monsoons I used, and probably would have top-octave problems due to the width of the throat - might be a great option if you don't mind a non-concentric super-tweeter to fill in.
I have a pair of 60" Carver planar drivers which appear to be unsellable, and in the back of my mind building a Unity style setup into the corners of a room is something I think would be a very cool experiment. (leaving aside the fact that I have zero time for experiments). If it worked, it would have close to zero effective footprint in the room, but should have nice even coverage.
This is in fact my big concern with the whole idea. Remember that the acoustic output from the mids/woofs will travel back to the ribbon/planar at the apex. The impedance mismatch coupled with the limited aperature should offer some protection, but I still can't see how the tweeter will be fully protected from loading at mid/bass frequencies. For true ribbons, I think this would be a deal-breaker; for planars, they may be robust enough.
My experiment worked really well, aside from the fact that they dropped like a rock at 10k. Patricks concern about HOM is - I think - not likely to be a huge problem with planar drivers. If I remember correctly from Geddes book, a diaphragm that is clamped at the side-wall of the horn is actually a GOOD match for an OS horn profile. So, assuming you could find a planar driver of the correct dimensions, it may actually be pretty close to ideal. Even the B&G Neo3 might be too wide for this, but a line of them would be an interesting place to start. The Neo8 is roughly the same size as the Monsoons I used, and probably would have top-octave problems due to the width of the throat - might be a great option if you don't mind a non-concentric super-tweeter to fill in.
I have a pair of 60" Carver planar drivers which appear to be unsellable, and in the back of my mind building a Unity style setup into the corners of a room is something I think would be a very cool experiment. (leaving aside the fact that I have zero time for experiments). If it worked, it would have close to zero effective footprint in the room, but should have nice even coverage.
This is in fact my big concern with the whole idea. Remember that the acoustic output from the mids/woofs will travel back to the ribbon/planar at the apex. The impedance mismatch coupled with the limited aperature should offer some protection, but I still can't see how the tweeter will be fully protected from loading at mid/bass frequencies. For true ribbons, I think this would be a deal-breaker; for planars, they may be robust enough.
I did a two-driver unity horn with a ribbon a couple of years ago. This is by far my crudest unity horn - I basically used stuff sitting on my shelf to see if it would work. It's the horn above, but with a ribbon at the apex instead of a compression driver. The pic above is using a peerless mid that is not attached.
Some thouhghts:
- This was by far the easiest horn to assemble. The reason is that you have a ton of leeway with a ribbon, because the diaphragm is IN the throat. With a compression driver your diaphragm is a good three to four inches further back, and that distance dramatically complicates everything. Basically with a ribbon you can leisurely move the woofers forward or back until the response flattens out. With a compression driver you do not have that luxury, as it's often a struggle to get the midranges close enough. Ideally, our midrange in a unity horn would be about a quarter wavelength away (this ignores the effect of the time delay which is caused by the bandpass enclosure, which is substantical.) For instance, if you use a crossover point of 1000hz, the gap from midrange to tweeters is less than three and a half inches.... That rules out a huge swath of midranges.
- The flat impedance of a ribbon is a godsend if you're doing passive crossovers
- I was amazed that the ribbon went all the way to 20khz WITH NO PHASE PLUG. This made no sense to me at all. I lterally bolted the square ribbon to a round throat and it just worked. If I hadn't seen the measurements I wouldn't have believed this was possible. If I had to hazard a guess, I would say that this trick worked because the cavity in front of the ribbon is so shallow. It's about 1/4" deep, and at that depth, standing waves will not develop until 13,500hz. Also, I used a foam plug in the horn which may have also soaked up standing waves. (I EQ'd the whole thing to flat after the passive crossover.)
The improvement in CSD was instantly obvious. Compression drivers clearly have some issues in the top octave, and you can see it in their CSD plots. The flat diaphragm may have some inherent advantage on a horn, also. The BMS compression drivers favored by Danley are much closer to the shape of a ribbon than the shape of a conventional dome. I have made something like six unity horns, and I would say this one was the second best. It came together in literally two hours. (I have a full measurement setup in the garage, so I can adjust xovers and EQ very quickly.) It's depressing that this two hour project sounded superior to some of the unitys that I have invested weeks in building.
Alas, the experiment was short lived. I got greedy with the ribbon, and lowered it's crossover to fill in the gap, instead of finding a proper midrange. (The Peerless midrange that I used was simply unsuitable to play beyond 800hz or so.)
When I got greedy to fill in that gap in the response, the ribbon detonated.
Last edited: