You met Siegfried once and know all this about him? Fantastic, I worked with him for four years and know less about how his thoughts evolved. BTW he was a career electronic engineer in R&D of Microwave and RF Test Equipment at SAD division of Hewlett Packard. Yes we all knew about the microchip world because we had to develop some requirements to perform the control functions in the test equipment, but we hardly design any microchips, that was left to the foundries and their scientists.
P.S. Jan Didden interviewed him, he probably got a lot more real information about what his thoughts were at the time.
P.S. Jan Didden interviewed him, he probably got a lot more real information about what his thoughts were at the time.
for the dual drive configuration shown in the 1st post here is the simple reasoning. First, acoustics follows linear theory. Basically that means if you know the solution to problem A and the solution to problem B, then A+B is the solution to solving both at the same time. So looking at the original configuration let A = the SPL from the front surface of the front driver, B the SPL from the rear of the front driver, C the SPL from the front of the rear driver and D the SPL from the back of the rear driver. The far field response is just A+B+C+D where you have to add considering phase and delays and each source is a monopole in the far field. So adding them and looking at the various sums it looks like this on axis:
A is just the flat monople response. A+B the dipole from the front and rear of the front driver. A+B+C the front driver dipole plus the front monople response of the rear driver, and A+B+C+D, all 4 sources.
Looking at it another way, adding the front of the front driver and rear of the rear driver, you have a long axis dipole, A+D. Then adding the rear of the front driver and the front of the rear driver you have a short axis dipole, B+C. Then adding them together you are back to A+B+C+D. But noted that adding the short axis dipole to the long axis dipole actually reduces the SPL. It would be better the have the two drivers mounted on the front and back side of an enclosure.
Of course, the net result will depend on what the separation between the drivers is. But I hope the concept is clear. Just adding the SPL of each source as you would in any multi source system.
A is just the flat monople response. A+B the dipole from the front and rear of the front driver. A+B+C the front driver dipole plus the front monople response of the rear driver, and A+B+C+D, all 4 sources.
Looking at it another way, adding the front of the front driver and rear of the rear driver, you have a long axis dipole, A+D. Then adding the rear of the front driver and the front of the rear driver you have a short axis dipole, B+C. Then adding them together you are back to A+B+C+D. But noted that adding the short axis dipole to the long axis dipole actually reduces the SPL. It would be better the have the two drivers mounted on the front and back side of an enclosure.
Of course, the net result will depend on what the separation between the drivers is. But I hope the concept is clear. Just adding the SPL of each source as you would in any multi source system.
This speaker design is way over my head. I cannot imagine doing all this research and then bolt two speaker together calling it a miracle.
Respectfully. I asked a simple direct question when I started this thread. How to wire the drivers given the amps I have available. I got an answer.
I never called the speakers a miracle. I had the components. I played with them. I liked what I got enough to play more. The tread got taken over with frankly no further contributions to my original question. I don't think it validated my question or the direct answer I got to the question.
Well great, I tried it today as I said last night using 2 x B19 woofers. I hung them from the door post since I did not want to spend time using making a stand thing, just tried to find how it worked. Then I hung a single speaker in the same position, they did sound different. Then I was advised to study Siegfrieds paper on the matter and proceeded to confirm the maths and I come to the same conclusion about the ineffectiveness of the application. The cone distance Siegfried referred to is for two flat cones hence me using the B139. Is there anything I am missing? If the cones were rigidly coupled, which I did, the bass extended a few Hz lower since the mass sort of doubled, the compliance remained the same. The input power was twice that as it would be for one speaker. What would be different than using a driver with two voice coils?
I am not shooting the idea down, it was novel in a way. I remember a similar system proposed in Wireless World some time in the sixties/early seventies that a concealed subwoofer in a box acting as side table was designed using two B139 drivers this way coupled face to face. I actually heard one that a fried at the time built. It did hit very very low notes. I speak under correction but I think the author was Arthur Bailey.
Last edited:
Hanging from a door post implies the speaker's radiation was distributed to two rooms.
If the speakers were hung perpendicular to the door the null pattern was distributed to each room.
If the speakers were hung parallel to the door, the null pattern would be parallel to the walls of both rooms.
In either case the reflections from another room would "sound different".
In either case, your listening results would be different than from a typical dipole set up in a single room, away from walls.
A single driver with two voice coils only has a single cone, two drivers have twice (+3dB) the radiating surface.
Face to face, two speakers in a box radiate as a monopole, that radiation is not similar to a dipole's radiation.
Weltersys,
The door is an arc 18 ft wide, there is an identical arc next to the first about 2 ft away. It is a large room and arcs are in the center and holds the wooden ceiling in place. Overall the room is about 42w x 60l x 14h ft. Drivers hung at approx halfway to ceiling about 7 ft of the carpeted floor.
These doors/openings are in the center of the lounge each side of the lounge is completely furnished one half for TV and the other for my hi-fi. I only had 2 hours to play while wife is shopping and I could only listen and fool around for a while. The speakers were suspended by a nylon rope from a wall light fitting above the arc only. So basically in free air.
I have main speakers disconnected to see how these perform alone - I worried about destroying them because cone movement was severe to obtain reasonable SPL. Oh yes, distance to closest wall is about 11 ft. It seemed impractical to use in my environment. Wife will disembowel me. Was something different though and worth the effort.
I will return the drivers into the H frame they were in (courtesy Siegfried). Strangely enough in the H frame the excursion is less and SPL higher.
EDIT: Facing the drivers edge on caused a null but turning one facing you, the other facing away was a peak if that says much. Maybe I should have them closer to a wall so the back wave reflects forward. That is for tomorrow.
The door is an arc 18 ft wide, there is an identical arc next to the first about 2 ft away. It is a large room and arcs are in the center and holds the wooden ceiling in place. Overall the room is about 42w x 60l x 14h ft. Drivers hung at approx halfway to ceiling about 7 ft of the carpeted floor.
These doors/openings are in the center of the lounge each side of the lounge is completely furnished one half for TV and the other for my hi-fi. I only had 2 hours to play while wife is shopping and I could only listen and fool around for a while. The speakers were suspended by a nylon rope from a wall light fitting above the arc only. So basically in free air.
I have main speakers disconnected to see how these perform alone - I worried about destroying them because cone movement was severe to obtain reasonable SPL. Oh yes, distance to closest wall is about 11 ft. It seemed impractical to use in my environment. Wife will disembowel me. Was something different though and worth the effort.
I will return the drivers into the H frame they were in (courtesy Siegfried). Strangely enough in the H frame the excursion is less and SPL higher.
EDIT: Facing the drivers edge on caused a null but turning one facing you, the other facing away was a peak if that says much. Maybe I should have them closer to a wall so the back wave reflects forward. That is for tomorrow.
Last edited:
Siegfried's model predicted the compound dipole woofer pair's combined output sum would be equal to a single driver in an H-frame, so the results don't seem strange to me.
https://www.linkwitzlab.com/models.htm#B1
At any rate, these arrangements are not about getting the most output from woofers, they are about the radiation pattern.
Last edited by a moderator:
Linkwitz's (and all other) calculation/simulation presume planar radiating discs. Most woofers are cones with large magnet and spider structure. That's why their backside radiation/spl response is not same as frontside. I'll soon make tests with off-axis measurements and most likely create a new thread for them. My guts say that with compound we loose most of the added cone area effect.
SL started in 1978 with big closed box subs and active delay control https://www.linkwitzlab.com/Removed pages/x-sb80-3wy.htm
Then in 2002 he introduced 3-way Orion with single or twin H-frame woofers https://www.linkwitzlab.com/Removed pages/x-sb80-3wy.htm
In 2012 LX521 he used 90deg W frame for twin woofers and narrow baffle mids and was 4-way
Kreskovsky used U-frame single woofer for NaO and slotted H frame with two opposed woofers in NaO Note
SL started in 1978 with big closed box subs and active delay control https://www.linkwitzlab.com/Removed pages/x-sb80-3wy.htm
Then in 2002 he introduced 3-way Orion with single or twin H-frame woofers https://www.linkwitzlab.com/Removed pages/x-sb80-3wy.htm
In 2012 LX521 he used 90deg W frame for twin woofers and narrow baffle mids and was 4-way
Kreskovsky used U-frame single woofer for NaO and slotted H frame with two opposed woofers in NaO Note
I'm not sure that magnets, frames etc will be much of a barrier to the long wavelengths sub 100 Hz. At higher frequencies they certainly affect the frequency response.
The Orion wasn't Siegfried's first OB speaker, incidentally. He was making dipole speakers in the late 80s, although with sealed woofers. See https://www.linkwitzlab.com/AES'89/AES'89.htm
His Audio Artistry designs had W frame dipole woofers, I think, and date from the mid 1990s - see https://www.stereophile.com/floorloudspeakers/328/index.html
Also his Phoenix design, with W frame woofers, came around 1999/2000: https://www.linkwitzlab.com/builtown.htm
The Orion wasn't Siegfried's first OB speaker, incidentally. He was making dipole speakers in the late 80s, although with sealed woofers. See https://www.linkwitzlab.com/AES'89/AES'89.htm
His Audio Artistry designs had W frame dipole woofers, I think, and date from the mid 1990s - see https://www.stereophile.com/floorloudspeakers/328/index.html
Also his Phoenix design, with W frame woofers, came around 1999/2000: https://www.linkwitzlab.com/builtown.htm
Last edited:
Thanks, it's diffult to compress an active designers work! Woofers and mid-T modules evolved independently and several combinations were tried and published. Pluto and LXmini use monopole midbass. Room acoustics adaptation and control were also important issues he was interested in and studied.
SL's last interview is humble
gainphile has earlier lectures here
SL's last interview is humble
gainphile has earlier lectures here
Nude15” dipole = brute force = what a waste of woofer just o avoid buiklding a proper baffle.
The models provided by members show the expected 300-400 Hz OB roll-off and the need for huge EQ and amplifier resources.
If i was doing an OB woofer (i likely wouldn’t be) i’d be more inclined in this direction — 12-16 15” woofers in a Heil configuration with a decent size baffle.
dave
The models provided by members show the expected 300-400 Hz OB roll-off and the need for huge EQ and amplifier resources.
If i was doing an OB woofer (i likely wouldn’t be) i’d be more inclined in this direction — 12-16 15” woofers in a Heil configuration with a decent size baffle.
dave
Do you actually have an argument against using nude drivers, or are you just going to call them names?
Yes, the fact that there is so much brute force required to get flat FR plus people forgetting that the reason is that one side of the driver is screwing with the phase of the other side. Thelower the dipole cancelation effect the better AFAIC.
dave
dave
Sorry, repeating the phrase 'brute force' isn't an argument.
Sorry?
But if you use a large baffle to lower the cancellation you lower the dipole peak and you also mess with the directivity above the dipole peak. So, depending on the passband you want, there is an argument for using large drivers and small baffles.
If this particular set of compromises doesn't appeal to you, that's OK. No-one's forcing anyone else to use it. Live and let live.
What do you think casues the dipole roll-off.
Every loudspeaker has many compromises. Instead of fighting the room i start with a really good room and embracece the room.
The dipolecancelation is the cause of the same flaw i have heard in every one of the many io have heard. embraceer for me Easte with toiler compromises.
dave
Just wanted to comment that I recently threw together a "compound woofer" test. Using a second Eminence Beta 15A I had on hand, on a temporary baffle, behind the one mounted on my flat OB panels. Magnet to cone configuration (similar to the Legacy Whisper) and I liked the improvement in sound. Driving both with a plate amp, wired in parallel. No measurements yet, but the added output and extra null on the sides of the speakers were obvious. Seemed to be an improvement in my room, the OBs are 24" off the side walls. Also tried magnet to magnet, preferred the former. I am going forward with a more finished inner baffle that will allow a closer distance to the front driver.
https://imagizer.imageshack.com/img924/3153/G0A6tw.jpg
https://imagizer.imageshack.com/img922/3537/6zEEGU.jpg
https://imagizer.imageshack.com/img923/1774/DhGMHL.jpg
https://imagizer.imageshack.com/img924/3153/G0A6tw.jpg
https://imagizer.imageshack.com/img922/3537/6zEEGU.jpg
https://imagizer.imageshack.com/img923/1774/DhGMHL.jpg