I put together a spreadsheet with eleven midranges for a Unity horn / waveguide.
It's on Google Docs, so you guys are welcome to add to it.
So far the two best candidates are the LA6-MR from Eminence and the JBL 400GTI.
I came up with a scoring system too. Basically 532 is a perfect score, and anything above or below that isn't quite perfect. John Sheerin's driver gets a perfect score, bracketed by the JBL 400GTI and the LA6-MR. The Beyma 5G40ND and the Ciare 6.38 NdMR look sharp too.
Here it is:
Unity Midrange Spreadsheet
woofer fs qes qms qts (2 * fs)/qes cost Unity Index Buy It! Specs
MCM 55-1595 100 0.88 2.44 0.65 227 $12 -1.30 http://www.mcminone.com/product.asp...ts&product_id=55-1595&info=details [/url]
Eminence NH2008 47 0.29 6.5 0.28 324 $67 -0.61 http://www.partsexpress.com/pe/psho....partsexpress.com/pdf/290-892s.pdf [/url]
Aurasound NS2 164 0.85 4.83 0.72 386 $16 -0.36 www.madisound.com [url]www.madisound.com [/url]
P-Audio SN-8MB 62 0.32 4.67 0.30 388 $79 -0.35 http://www.loudspeakersplus.com/htm...ustic.com/pdf/SN-SERIES/SN-8MB.pdf [/url]
Aurasound NS3-194-8E 125 0.6 7.1 0.55 417 $11 -0.26 http://www.madisound.com/cgi-bin/in...om/pdf/aurasound/NS3-194-8E v2.pdf [/url]
JBL 400GTI 187 0.85 7.8 0.77 440 $150 -0.19 ebay
the john sheerin driver 81 0.31 4.2 0.29 523 $999 0.00 http://ldsg.snippets.org/HORNS/unity2.html
Eminence LA6-MR 90 0.28 2.92 0.26 643 $50 0.19 http://www.usspeaker.com/LA6-MR-1.htm [url]http://www.usspeaker.com/LA6-MR-1.htm [/url]
Beyma 5G40ND 110 0.33 4.9 0.31 667 $110 0.22 http://www.usspeaker.com/beyma 5g40...w.usspeaker.com/beyma 5g40nd-1.htm [/url]
Ciare 6.38 NdMR 197 0.36 1.47 0.29 1094 $60 0.52 http://www.assistanceaudio.com/08_C.../www.ciare.com/oem/pdf/638NDMR.pdf [/url]
P-Audio WN-520N 173 0.256 0.912 0.20 1352 $100 0.61 www.loudspeakersplus.com [url]http://www.paacoustic.com/pdf/WINNER-SERIES/WN-520N.pdf[/url]
It's on Google Docs, so you guys are welcome to add to it.
So far the two best candidates are the LA6-MR from Eminence and the JBL 400GTI.
I came up with a scoring system too. Basically 532 is a perfect score, and anything above or below that isn't quite perfect. John Sheerin's driver gets a perfect score, bracketed by the JBL 400GTI and the LA6-MR. The Beyma 5G40ND and the Ciare 6.38 NdMR look sharp too.
Here it is:
Unity Midrange Spreadsheet
woofer fs qes qms qts (2 * fs)/qes cost Unity Index Buy It! Specs
MCM 55-1595 100 0.88 2.44 0.65 227 $12 -1.30 http://www.mcminone.com/product.asp...ts&product_id=55-1595&info=details [/url]
Eminence NH2008 47 0.29 6.5 0.28 324 $67 -0.61 http://www.partsexpress.com/pe/psho....partsexpress.com/pdf/290-892s.pdf [/url]
Aurasound NS2 164 0.85 4.83 0.72 386 $16 -0.36 www.madisound.com [url]www.madisound.com [/url]
P-Audio SN-8MB 62 0.32 4.67 0.30 388 $79 -0.35 http://www.loudspeakersplus.com/htm...ustic.com/pdf/SN-SERIES/SN-8MB.pdf [/url]
Aurasound NS3-194-8E 125 0.6 7.1 0.55 417 $11 -0.26 http://www.madisound.com/cgi-bin/in...om/pdf/aurasound/NS3-194-8E v2.pdf [/url]
JBL 400GTI 187 0.85 7.8 0.77 440 $150 -0.19 ebay
the john sheerin driver 81 0.31 4.2 0.29 523 $999 0.00 http://ldsg.snippets.org/HORNS/unity2.html
Eminence LA6-MR 90 0.28 2.92 0.26 643 $50 0.19 http://www.usspeaker.com/LA6-MR-1.htm [url]http://www.usspeaker.com/LA6-MR-1.htm [/url]
Beyma 5G40ND 110 0.33 4.9 0.31 667 $110 0.22 http://www.usspeaker.com/beyma 5g40...w.usspeaker.com/beyma 5g40nd-1.htm [/url]
Ciare 6.38 NdMR 197 0.36 1.47 0.29 1094 $60 0.52 http://www.assistanceaudio.com/08_C.../www.ciare.com/oem/pdf/638NDMR.pdf [/url]
P-Audio WN-520N 173 0.256 0.912 0.20 1352 $100 0.61 www.loudspeakersplus.com [url]http://www.paacoustic.com/pdf/WINNER-SERIES/WN-520N.pdf[/url]
omarmipi said:
Nice one! I'll have to look that up.
I found another good & cheap candidate. The TangBand W2-880 scores an 889. It costs $12 IIRC. I have a few at home, but I didn't measure them yet. I'm using the data that John Krutke measured. Another plus is that the W2-880 has much lower distortion than the Aura NS2s that I'm using.
For a "real" unity the TBs are rather puny, but for a car, the small size is a bonus.
Here's the data:
http://www.zaphaudio.com/minitest/
Greets!
It would be tiny, ~2x the front chamber Vb with it mounted flat against the horn. For open back, you'd need a ~500 Hz Fs driver with a mass corner slightly >1 kHz, ergo a large front chamber Vb with large vents, so not a viable option IMO since you'd need a Qes = ~0.05 to get the front Vb/vents down to ~where they need to be.
GM
It would be tiny, ~2x the front chamber Vb with it mounted flat against the horn. For open back, you'd need a ~500 Hz Fs driver with a mass corner slightly >1 kHz, ergo a large front chamber Vb with large vents, so not a viable option IMO since you'd need a Qes = ~0.05 to get the front Vb/vents down to ~where they need to be.
GM
Ok open back it is...
The Fostex fe126e is close in specs and a little cheaper would this work also? I would worry that with the few very inexpensive drivers that you might end up having to buy a bunch to obtain drivers with the parameters needed so not much if any saving in the end. Regards Moray James.
The Fostex fe126e is close in specs and a little cheaper would this work also? I would worry that with the few very inexpensive drivers that you might end up having to buy a bunch to obtain drivers with the parameters needed so not much if any saving in the end. Regards Moray James.
GM,GM said:
Every Unity will have a different target. To keep things simple, I decided to assign a "bullseye target" in my spreadsheet. It corresponds to John Sheerin's "perfect unity woofer", that's all.
For example, if you went with a BMS4540nd, you'd want a target that's an octave higher (1064hz). This would allow you to extend the high frequency response of the mid, taking some stress off the BMS. Of course it would also limit your low frequency extension!
If you used a monster compression driver like one of the TAD 2001s, you can get away with a target that's a full octave lower (266hz). This is because the TAD can live with a lower xover point. Of course, the $95 BMS is hard to beat in the top octave!
In a nutshell, your "target" will depend on your desired SPL level and your compression driver.
Aren't compromises a b1tch? This demonstrates that even if you spend $1000 for a compression driver, there will still be compromises. I hate to gush, but I think Danley has come up with one of the most elegant ways to mask these compromises. IMHO, a lot of newbies think that you can just throw money at the problem by purchasing the best drivers. This shows that driver selection is more important than the cost of those drivers.
:: PB ::
An externally hosted image should be here but it was not working when we last tested it.
Dear Greg:
The specifications you are mentioning were for the Audax, not the Real McCoy, correct?
I am working towards a goal of a conically-shaped horn, but my efforts are being slowed by almost constant travel. Did you see my latest attempt: http://people.qualcomm.com/kalousek/index.htm ?
Dear John,
Since I have been working on different implementations of the Unity concept for a while, I could not help myself from thinking about Tom's thought process while he was designing the first Unity. And, it seems to me, that there are not so many degrees of freedom as it appears.
First, the choice of a cross-over frequency is limited - at the low frequency range by the loading of the compression driver, and at the high frequency range by the physical limitations of the horn structure (e.g., the wall angle) and the dimensions of the mid-frequency drivers (one needs to be at about 1/4 wavelength), as well as the highest frequency that the mid-frequency drivers can reach on the horn.
Additionally, the bandwidth of the mid-frequency drivers is limited, I am afraid that 3 octaves is the maximum, again depending on the horn structure that will affect e.g., loading. This further influences the cross-over frequency choice if one wants a certain bandwidth with only a compression driver and mid-frequency drivers.
Furthermore, the mechanical design presents additional limitation. Not everyone would accept long ports necessitating less compact design.
Please, do not misunderstand me; I think this is an ingenious concept. I just wonder if when Tom decided on performance goals the design did not just logically follow. However, perhaps I am just fooling myself and see the logic using hindsight and experiences of the many smart people who understood and discussed the Unity concept.
Kindest regards,
M
The specifications you are mentioning were for the Audax, not the Real McCoy, correct?
I am working towards a goal of a conically-shaped horn, but my efforts are being slowed by almost constant travel. Did you see my latest attempt: http://people.qualcomm.com/kalousek/index.htm ?
Dear John,
Since I have been working on different implementations of the Unity concept for a while, I could not help myself from thinking about Tom's thought process while he was designing the first Unity. And, it seems to me, that there are not so many degrees of freedom as it appears.
First, the choice of a cross-over frequency is limited - at the low frequency range by the loading of the compression driver, and at the high frequency range by the physical limitations of the horn structure (e.g., the wall angle) and the dimensions of the mid-frequency drivers (one needs to be at about 1/4 wavelength), as well as the highest frequency that the mid-frequency drivers can reach on the horn.
Additionally, the bandwidth of the mid-frequency drivers is limited, I am afraid that 3 octaves is the maximum, again depending on the horn structure that will affect e.g., loading. This further influences the cross-over frequency choice if one wants a certain bandwidth with only a compression driver and mid-frequency drivers.
Furthermore, the mechanical design presents additional limitation. Not everyone would accept long ports necessitating less compact design.
Please, do not misunderstand me; I think this is an ingenious concept. I just wonder if when Tom decided on performance goals the design did not just logically follow. However, perhaps I am just fooling myself and see the logic using hindsight and experiences of the many smart people who understood and discussed the Unity concept.
Kindest regards,
M
The B&C driver will probably take a lot of modifications to hide in the dash but it is the driver closest to the 529Hz value.
Woofer fs qes qms qts (2 * fs)/qes cost
B&C 6MD38 130 0.49 3.70 0.44 530 $78.45
http://www.partsexpress.com/pe/showdetl.cfm?&Partnumber=294-651
Woofer fs qes qms qts (2 * fs)/qes cost
B&C 6MD38 130 0.49 3.70 0.44 530 $78.45
http://www.partsexpress.com/pe/showdetl.cfm?&Partnumber=294-651
The bad news is that the driver I used in my Unity clone, the aurasound nsw2 "whisper" is no longer available.
The good news is that Peerless is selling a two inch driver with thiele small parameters which are virtually identical, but a LITTLE bit better. John Sheerin and I both had a b1tch of a time getting the midranges to play high enough, and this little Peerless driver is a compelling candidate. The Aurasound Whisper gave up at about 772hz, but the Peerless should be good for 888hz.
It's so close that it should practically be a drop-in replacement in my design.
I put the thiele small parameters online in my spreadsheet:
http://spreadsheets.google.com/ccc?key=p4iluSrYmufFENZeZHPciOw&pli=1
I took the parameters from John Krutke. His measurements are usually more reliable than the manufacturers, who often candy-coat the data:
http://www.zaphaudio.com/tidbits/Peerless-830970-TS.gif
You can buy it here:
http://www.madisound.com/catalog/product_info.php?products_id=8267
The good news is that Peerless is selling a two inch driver with thiele small parameters which are virtually identical, but a LITTLE bit better. John Sheerin and I both had a b1tch of a time getting the midranges to play high enough, and this little Peerless driver is a compelling candidate. The Aurasound Whisper gave up at about 772hz, but the Peerless should be good for 888hz.
It's so close that it should practically be a drop-in replacement in my design.
I put the thiele small parameters online in my spreadsheet:
http://spreadsheets.google.com/ccc?key=p4iluSrYmufFENZeZHPciOw&pli=1
I took the parameters from John Krutke. His measurements are usually more reliable than the manufacturers, who often candy-coat the data:
http://www.zaphaudio.com/tidbits/Peerless-830970-TS.gif
You can buy it here:
http://www.madisound.com/catalog/product_info.php?products_id=8267
Awesome news! Thanks for the update. I'm thinking of changing my system again. I have the Peerless 4" PPB cones in 1.4 Liter enclosures on my dash with the LPG neo Aluminums and JL Audio 8IB4s as midbass right now. The time alignment is easily dealt with (for one passenger) but the imaging is killing me. The imaging is similar to that of headphones... none.
Are you still using the BMS compression driver or have you tried the Alpine XT19?
I'm thinking of buying a pair of 18sound 6ND430 drivers because of my failed attempt to fabricate perfect door panels. Fitting 8" woofers with a 3.75" depth is almost impossible in the Subaru Impreza platform front doors.
Are you still using the BMS compression driver or have you tried the Alpine XT19?
I'm thinking of buying a pair of 18sound 6ND430 drivers because of my failed attempt to fabricate perfect door panels. Fitting 8" woofers with a 3.75" depth is almost impossible in the Subaru Impreza platform front doors.
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.
An externally hosted image should be here but it was not working when we last tested it.
omarmipi said:
Did you have any luck with this project?
You may have noticed that I re-vived my Unity project. I'm about to purchase a few more midranges for testing purposes, and the Audax which you are using is one of the candidates. The supply is nearly exhausted, but Madisound still has it on their sale page at the moment.
http://www.madisound.com/catalog/product_info.php?products_id=101
I've been messing with Unity horns for close to three years now, and I've managed to purchase a pile of midrange drivers that work on Unity horns. I have the Misco drivers which I believe were used in the Lambda Unity horns, as well as a newer driver from Misco which is as good or better.
Misco JC5RTF-B, KCN5FD and TB W4-1805S
One thing that we have ignored in this discussion is frequency response and distortion. The woofers in a Unity horn require a very VERY specific set of Thiele Small parameters. Basically it needs a VERY high FS, and a VERY low QES.
What you end up with is a woofer that has an extreeeeeemely light cone, and an accordian surround. If you add even a fraction of a gram to the cone or the surround, it drives down the FS too low, and then it doesn't work for a Unity horn! It's as simple as that - if the specs on the midrange are wrong, your Unity horn isn't going to work.
Because the Unity mids require a cone which is astonishingly light, and an accordian surround, we end up with a woofer that looks a lot like something you'd find on an old CRT TV. Basically a giant version of one of those cone tweeters you see in speakers from the 70s.
One interesting alternative is the Tang Band W4-1805S. Tang Band makes a lot of great drivers, but only TWO of them will work on a Unity horn. The reason that all the others won't work is that their cones are too heavy, which drives down the FS too low, and the QES is too high. This particular woofer is an exception. It has the right specs for a Unity horn, AND it has an underhung motor with very low distortion. I haven't measured it's distortion yet, but it's audibly cleaner than the Misco.
(That isn't a knock on the Misco; it's a MUCH less expensive driver. Misco offers exceptional performance, and the TB gives you all that and more, for an additional cost.)
Parts-Express.com:*Tang Band W4-1805S 4" Neodymium Midrange Driver | Tang Band W4-1805S 4" midrange tb speakers neodymium driver line array point source midrange home theater computer speaker center channel set tube amplifier cone mid
Misco JC5RTF-B, KCN5FD and TB W4-1805S
One thing that we have ignored in this discussion is frequency response and distortion. The woofers in a Unity horn require a very VERY specific set of Thiele Small parameters. Basically it needs a VERY high FS, and a VERY low QES.
What you end up with is a woofer that has an extreeeeeemely light cone, and an accordian surround. If you add even a fraction of a gram to the cone or the surround, it drives down the FS too low, and then it doesn't work for a Unity horn! It's as simple as that - if the specs on the midrange are wrong, your Unity horn isn't going to work.
Because the Unity mids require a cone which is astonishingly light, and an accordian surround, we end up with a woofer that looks a lot like something you'd find on an old CRT TV. Basically a giant version of one of those cone tweeters you see in speakers from the 70s.
One interesting alternative is the Tang Band W4-1805S. Tang Band makes a lot of great drivers, but only TWO of them will work on a Unity horn. The reason that all the others won't work is that their cones are too heavy, which drives down the FS too low, and the QES is too high. This particular woofer is an exception. It has the right specs for a Unity horn, AND it has an underhung motor with very low distortion. I haven't measured it's distortion yet, but it's audibly cleaner than the Misco.
(That isn't a knock on the Misco; it's a MUCH less expensive driver. Misco offers exceptional performance, and the TB gives you all that and more, for an additional cost.)
Parts-Express.com:*Tang Band W4-1805S 4" Neodymium Midrange Driver | Tang Band W4-1805S 4" midrange tb speakers neodymium driver line array point source midrange home theater computer speaker center channel set tube amplifier cone mid
You say you want low low Qes and very high Fs. So you are asking for, basically whats known as Efficiency Bandwidth Product? Is that correct??? Fs/Qes=EBP
But is BL/MMS possibly the better figure of merit? Or is this all the same as EBP except for L of the coil overhang? And you don't want any overhang...
What I mean, if the motor can handle the extra weight, does it matter Fs is a bit low? And if the motor ain't powerful enough to move the cone's weight, does it really help that Fs may be sprung moderately high?
What do you think of the Galaxy Hotspot and the Neodymium version of the same? Fs a little low (but higher than your Tang Band, and half the price). But then again, backs of either driver aren't yet cupped, so how does one compare??? Cupping them should raise Qms, Fs, and EBP. But no change whatsoever to Qes or BL/MMS.
Galaxy 5" drivers should perhaps not be overlooked? Galaxy Audio S5C160-8 5" Full Range Driver
DeltaPro8 also has murderously high EBP (308) for a driver of its size, and will only go up when cupped. But perhaps bigger than you were looking on a unity. Unfortunately, its not cheap.
But is BL/MMS possibly the better figure of merit? Or is this all the same as EBP except for L of the coil overhang? And you don't want any overhang...
What I mean, if the motor can handle the extra weight, does it matter Fs is a bit low? And if the motor ain't powerful enough to move the cone's weight, does it really help that Fs may be sprung moderately high?
What do you think of the Galaxy Hotspot and the Neodymium version of the same? Fs a little low (but higher than your Tang Band, and half the price). But then again, backs of either driver aren't yet cupped, so how does one compare??? Cupping them should raise Qms, Fs, and EBP. But no change whatsoever to Qes or BL/MMS.
Galaxy 5" drivers should perhaps not be overlooked? Galaxy Audio S5C160-8 5" Full Range Driver
DeltaPro8 also has murderously high EBP (308) for a driver of its size, and will only go up when cupped. But perhaps bigger than you were looking on a unity. Unfortunately, its not cheap.
Last edited:
I've studied the Unity horn for close to a decade, and the comments that GM made in this thread were one of the "Eureka" moments when I began to understand how all these pieces fit together in a Unity.
It's not just that we need a high EBP; we need a very SPECIFIC EBP.
As GM noted, here's the formula:
2 * FS / QES = 500
GM or Danley might come in and school me, but here's how I understand this:
The EBP will dictate the upper and the lower limit of our midranges on the horn. It works out to about one octave on either side of (EBP x 2). So if we want to get the mids up to 1khz, our lower limit is about 250hz.
Here's the EBP and upper limits for various woofers:
Here are my thoughts on the ones you suggested. I have the Galaxy Audio, and it's parameters aren't appropriate for a Unity horn. You basically can't get it to play high enough. Getting it past 800hz will be a struggle. John Sheerin evaluated the ferrite version for his Unities, and discarded them also.
The Misco JC5RTF-B is clearly the easiest one to work with, has the right parameters, and you don't have to hassle with sealing the back.
The TangBand W2 sounds a bit "cleaner" to my ears. Is it the underhung motor? I dunno. I haven't measured it's distortion, but it sounds very good. It's efficiency is much higher than you might think too; a pair of the W2s nearly match the efficiency of the Misco 5".
The TangBand W4 is clearly the cleanest sounding of the bunch. It is also the most expensive.
[/font]
It's not just that we need a high EBP; we need a very SPECIFIC EBP.
As GM noted, here's the formula:
2 * FS / QES = 500
GM or Danley might come in and school me, but here's how I understand this:
The EBP will dictate the upper and the lower limit of our midranges on the horn. It works out to about one octave on either side of (EBP x 2). So if we want to get the mids up to 1khz, our lower limit is about 250hz.
Here's the EBP and upper limits for various woofers:
- Misco KCN5FD (only sold in quantity) - 384 / 1538hz
- Tangband W2-852SH ($15) - 356 / 1422hz
- Tang Band W4-1805S ($69) - 275 / 1100hz
[*]Misco JC5RTF-B (only sold in quantity) 250 / 1khz - Galaxy Audio S5N-8 ($49) - 197 / 788hz
Here are my thoughts on the ones you suggested. I have the Galaxy Audio, and it's parameters aren't appropriate for a Unity horn. You basically can't get it to play high enough. Getting it past 800hz will be a struggle. John Sheerin evaluated the ferrite version for his Unities, and discarded them also.
The Misco JC5RTF-B is clearly the easiest one to work with, has the right parameters, and you don't have to hassle with sealing the back.
The TangBand W2 sounds a bit "cleaner" to my ears. Is it the underhung motor? I dunno. I haven't measured it's distortion, but it sounds very good. It's efficiency is much higher than you might think too; a pair of the W2s nearly match the efficiency of the Misco 5".
The TangBand W4 is clearly the cleanest sounding of the bunch. It is also the most expensive.
[/font]
Take a look at the Synergy Horn patent.
Both the Unity horn and the Synergy Horn are arrays. In fact, the lab sub is an array too.
Danley is subdiving a single woofer into multiple unique parts, then arraying them on a horn. The midranges and the woofers on the Synergy, Unity, and Lab horns are arrays.
It isn't a coincidence that Danley's new designs, like the new tapped horns and the Paraline use multiple elements. Danley's commercial offerings are quietly solving the problems with conventional arrays.
