Midbass
I have used an 80 Hz Edgarhorn (straight horn) with D130F drivers (old JBL labeled for Fender) for midbass duties. These units proved to be fairly intense sounding, but very musical. The diffuculty was actually sufficient room size. To get decent integration I needed to either move well back from my normal listening position, or somehow adjust the propagation of the midbass.
I have plans for folded 40 Hz horns which I intend to build sometime soon, in the hopes of spanning the large gap up to 450-500 Hz, though unlikely...
I have also built a bass horn for duties down to 30 Hz which uses a pair of 10" drivers. Weighs about 140 kilos at ~ 122 cm square x 46 cm (48" x 48" x 18"), so I really don't need the 40 Hz horns, but I believe that a really great 3-way system could be managed with a midbass that would span 50 Hz up to 500 Hz as the foundation. Stay tuned...
Electrovoice or EV built some good midbass drivers for years, and a friend is using a pair of 12" in his 100 Hz midbass horn and they span up to 1000 Hz.
The EV 640 is also a very good horn for a 2" driver, particularily the TAD 4001.
I have heard the TAD 1201 and it does provide some pop. Same is true of the JBL D123, but to a slightly lesser extent.
The 1" TAD 2001 will play from ~ 600 Hz up to about 16K on a good horn. Lively two way is possible with, say, a pair of 12" and the TAD.
I have heard several systems with the TAD 1601 as a bass driver in a direct radiator with horns on top and they do a good job.
I like dipole bass too -- less room interaction, uncolored, natural sounding bass. The midbass from a horn, however, will pin you to your chair with, for example, kick drum, with no sense of strain or distortion from the dynamic peaks.
Probably missed a few points...
Tim
I have used an 80 Hz Edgarhorn (straight horn) with D130F drivers (old JBL labeled for Fender) for midbass duties. These units proved to be fairly intense sounding, but very musical. The diffuculty was actually sufficient room size. To get decent integration I needed to either move well back from my normal listening position, or somehow adjust the propagation of the midbass.
I have plans for folded 40 Hz horns which I intend to build sometime soon, in the hopes of spanning the large gap up to 450-500 Hz, though unlikely...
I have also built a bass horn for duties down to 30 Hz which uses a pair of 10" drivers. Weighs about 140 kilos at ~ 122 cm square x 46 cm (48" x 48" x 18"), so I really don't need the 40 Hz horns, but I believe that a really great 3-way system could be managed with a midbass that would span 50 Hz up to 500 Hz as the foundation. Stay tuned...
Electrovoice or EV built some good midbass drivers for years, and a friend is using a pair of 12" in his 100 Hz midbass horn and they span up to 1000 Hz.
The EV 640 is also a very good horn for a 2" driver, particularily the TAD 4001.
I have heard the TAD 1201 and it does provide some pop. Same is true of the JBL D123, but to a slightly lesser extent.
The 1" TAD 2001 will play from ~ 600 Hz up to about 16K on a good horn. Lively two way is possible with, say, a pair of 12" and the TAD.
I have heard several systems with the TAD 1601 as a bass driver in a direct radiator with horns on top and they do a good job.
I like dipole bass too -- less room interaction, uncolored, natural sounding bass. The midbass from a horn, however, will pin you to your chair with, for example, kick drum, with no sense of strain or distortion from the dynamic peaks.
Probably missed a few points...
Tim
Re: Midbass
I have been thinking very much that I might have a problem with this also, which is why I remain undecided on a 135 hz down solution. I wil probably go with whatever puts me the closest to matching the dispersion of the Azurahorns.
Soooooooooooo, since the sealed subs sound passable, I decided to just go ahead with the upper ranges and see how that worked out. If I wind up with something really nice from there on up, by that time I will have a solution in mind.
If I'm not happy with my results and wind up starting over then I'll probably be looking for a different solution in the midbass area anyhow _grin_
Regards
Ken L
Tim Moorman said:
I have been thinking very much that I might have a problem with this also, which is why I remain undecided on a 135 hz down solution. I wil probably go with whatever puts me the closest to matching the dispersion of the Azurahorns.
Soooooooooooo, since the sealed subs sound passable, I decided to just go ahead with the upper ranges and see how that worked out. If I wind up with something really nice from there on up, by that time I will have a solution in mind.
If I'm not happy with my results and wind up starting over then I'll probably be looking for a different solution in the midbass area anyhow _grin_
Regards
Ken L
2206
Rocky,
The JBL 2206 looks to be a typically fine JBL driver, and would probably work very well in a horn. I don't know of any used in commercial systems, but that doesn't mean much.
Any driver will benefit from horn loading, to some extent. You will get some on-axis efficiency gains just by limiting directivity. But for the response to remain smooth and meet your design goals, then some number crunching and/or horn design software work will be necessary.
My guess is that you could use this down to around 35 or 40 Hz in a horn optimized for low frequency. Probably wouldn't get much beyond 350-400 Hz on the upper end though.
On the high end, the horn folding may roll off some of the extension, if any folding is used, plus just horn loading the unit will narrow the passband and lower the top end. Still, with this driver you have a lot to give away, and it might work very well from 100 - 600 Hz.
Tim
Rocky,
The JBL 2206 looks to be a typically fine JBL driver, and would probably work very well in a horn. I don't know of any used in commercial systems, but that doesn't mean much.
Any driver will benefit from horn loading, to some extent. You will get some on-axis efficiency gains just by limiting directivity. But for the response to remain smooth and meet your design goals, then some number crunching and/or horn design software work will be necessary.
My guess is that you could use this down to around 35 or 40 Hz in a horn optimized for low frequency. Probably wouldn't get much beyond 350-400 Hz on the upper end though.
On the high end, the horn folding may roll off some of the extension, if any folding is used, plus just horn loading the unit will narrow the passband and lower the top end. Still, with this driver you have a lot to give away, and it might work very well from 100 - 600 Hz.
Tim
After some concideration, I feel confident I will try to compromise my first horn setup by avoiding midbass horn, using instead direct radiating drivers.. This would normally force me to use an extra set of amps (as a non-loaded driver can't keep up with horns' SPL), but I am thinking of a *yet another compromise* solution I would like to hear some response on..
PHL produces a small 6,5 inch coax (PHL 1520 ) efficient enough so that I can couple 2 paralel pairs in series (4 drivers) and get 103dB/W/m, 8 ohm, direct radiation.. reaching down to about 100Hz.. I may be mixing religions here, but;
1) would it be possible to run this array paired with midfreq and hifreq horns, using passive XO (a few dBs attenuation of the horns)
2) Does anyone have an opinion on how these drivers might match a loaded 2-inch compression driver?
PHL produces a small 6,5 inch coax (PHL 1520 ) efficient enough so that I can couple 2 paralel pairs in series (4 drivers) and get 103dB/W/m, 8 ohm, direct radiation.. reaching down to about 100Hz.. I may be mixing religions here, but;
1) would it be possible to run this array paired with midfreq and hifreq horns, using passive XO (a few dBs attenuation of the horns)
2) Does anyone have an opinion on how these drivers might match a loaded 2-inch compression driver?
Just in case some of you haven't been here, a source of horn knowledge - Audio Asylum High Efficiency Speaker Forum -
http://www.audioasylum.com/forums/HUG/bbs.html
some of the inmates are loudspeaker engineers during the daytime...
http://www.audioasylum.com/forums/HUG/bbs.html
some of the inmates are loudspeaker engineers during the daytime...
Horn system
Rocky,
I wouldn't run coax pairs each side due to big comb filtering problems. Better to pair up on the midbass units such as the 2206, or others to raise efficiency up to 101 dB or better, then pad the horn down. Actually, most systems are run this way with direct radiator midbass.
Have a look at B&C midbass units. Many are near 100 dB efficiency, when paired, each channel would be near106 dB, and are available in 16 ohm.
SMathews -- The Edgar midbass probably needs 4 meters or more with the same D130F. I do believe some mods would tame this down somewhat, and Bruce Edgar would know just how to go about it.
I didn't try the EVM 15L, which is perhaps the preferred unit because I wanted to try to build a 40 or 50 Hz horn to span upwards to the mid.
The JBL 2123 10" is a great driver, now obsolete, and runs from ~100 Hz up to 4000 Hz dead flat, and works great in a midbass.
Many horn loaded the 2105 5" cone unit, including JBL in a tractrix mid horn, which sounded very good, I'm told, though not the resolution of the big compression drivers.
I do love the looks of a solid maple or mahogany wood horn, and the round horns are way cool. Edgar sells these as well in a choice of 350, 500, or 650 Hz for compresion driver mounting at $900/pr, $400/pr, $400/pr respectively. It is possible to build your own by laminating layer after layer of wood to stack up to the right length and contour, but you need a good router and patience.
Tim
Rocky,
I wouldn't run coax pairs each side due to big comb filtering problems. Better to pair up on the midbass units such as the 2206, or others to raise efficiency up to 101 dB or better, then pad the horn down. Actually, most systems are run this way with direct radiator midbass.
Have a look at B&C midbass units. Many are near 100 dB efficiency, when paired, each channel would be near106 dB, and are available in 16 ohm.
SMathews -- The Edgar midbass probably needs 4 meters or more with the same D130F. I do believe some mods would tame this down somewhat, and Bruce Edgar would know just how to go about it.
I didn't try the EVM 15L, which is perhaps the preferred unit because I wanted to try to build a 40 or 50 Hz horn to span upwards to the mid.
The JBL 2123 10" is a great driver, now obsolete, and runs from ~100 Hz up to 4000 Hz dead flat, and works great in a midbass.
Many horn loaded the 2105 5" cone unit, including JBL in a tractrix mid horn, which sounded very good, I'm told, though not the resolution of the big compression drivers.
I do love the looks of a solid maple or mahogany wood horn, and the round horns are way cool. Edgar sells these as well in a choice of 350, 500, or 650 Hz for compresion driver mounting at $900/pr, $400/pr, $400/pr respectively. It is possible to build your own by laminating layer after layer of wood to stack up to the right length and contour, but you need a good router and patience.
Tim
The efficiency of the 2206 is not very high but still quite good compared with most HiFi drivers. It is around 95 dB/w/m.
I use one in a two-way combination together with a 2344 horn and it sounds fine.
I deliberately took this one because I assumed that it performs better up to 1 kHz than any 15" would. It is one of the oldest drivers that I know that uses a glass fibre- (or whatever) paper- composite for it's cone (wich is quite fashionable nowadays).
Regards
Charles
I use one in a two-way combination together with a 2344 horn and it sounds fine.
I deliberately took this one because I assumed that it performs better up to 1 kHz than any 15" would. It is one of the oldest drivers that I know that uses a glass fibre- (or whatever) paper- composite for it's cone (wich is quite fashionable nowadays).
Regards
Charles
Gosh. This is almost basement blaster territory, which is what I call my 2 1/2 way using three vertically oriented 2226J's topped with a 2445J driving a B & C ME75 CD horn xovered @ 700hz. The 'half' xover to the two lower 2226J's is at about half that frequency. The 2245 is probably not the best choice for ultimate SQ, but I bought a used pair fairly cheap before I did all my homework and they're not bad up to 10K. Probably the best thing I can say about them above that is that I'm getting up to 20K pretty flat out of them but with some dips between those two frequencies. The system as a whole comes out 101-102 db/w/m and hovers around 5-6 ohms through most of the woofer's range.
One nice thing about the 2226 and I presume the 2206 also is that it can be run below its free air resonance without a huge increase in distortion due to suspension nonlinearities, and that's exactly what I'm doing here. I've tuned the BR box to 25 hz, which is wayy below the 40 hz Fs of the 2226, and then getting up to a 4 db passive boost from 25-70hz with a series C shunt L network before the woofs. One thing I like about this approach is that, in addition to adding a third octave flat extension on the bottom, its voltage gain is also above and beyond the voltage at the amp terminals, so a 300 watt amp looks like 800 watts or so where the boost is 4db, for instance. Add in room gain, and this speaker operates virtually 20hz-20,000hz as a 2.5 way. This all works without dropping the load impedance down to the one ohm region because I'm actually also flattening out the upper BR impedance peak here with this network
The only significant SPL compromise is that woofer excursion peaks at 40 hz, but you still have to have it above 110db to really see the cones move at that frequency.
It's in a 20" wide 5' tall cabinet. I've found that matching the horizontal dispersions of the upper woofer and the CD horn at xover solved many of the potential issues matching a direct radiator bass with a horn hf. However, the 2226 by itself is only 97 db at that frequency. Radiation pattern narrowing matches up with the horn lateral dispersion at xover which I think really helps make a seamless transition. Then I throw in 4 db boost just below the xover frequency with a somewhat peaked *elliptic* low pass for the upper woofer only. Why elliptic? It's non ripple, and the delay increases near xover to make a smooth group delay transition to the set back horn driver vc and compensate for the naturally increasing group delay as the frequency decreases down to below 200hz. Plus, it's easy to set the notch right where JBL's graphs show a little unruly distortion behavior from the 2226, I believe around 1.5 khz in this case. Plus, it starts out rolling off almost like a first order xover before the response drops off a cliff. The two lower woofs come in right when the peaking rolls off on the low side for the upper woofer to maintain flat response. I'm rolling them off pretty close to 2nd order Butterworth which has slightly less group delay yet but the delay peaks slightly at xover. One nice thing about this arrangement, IMO, is that above 350 hz, the speaker radiates like a 2 speaker 2 way on a 3' shelf, which give a credible height simulation of 'real' musicians in the voice band and above but the two lower woofers step in to keep cone excursion (and doppler distortion) way down below that frequency except for the most massive bass dynamics.
One nice thing about the 2226 and I presume the 2206 also is that it can be run below its free air resonance without a huge increase in distortion due to suspension nonlinearities, and that's exactly what I'm doing here. I've tuned the BR box to 25 hz, which is wayy below the 40 hz Fs of the 2226, and then getting up to a 4 db passive boost from 25-70hz with a series C shunt L network before the woofs. One thing I like about this approach is that, in addition to adding a third octave flat extension on the bottom, its voltage gain is also above and beyond the voltage at the amp terminals, so a 300 watt amp looks like 800 watts or so where the boost is 4db, for instance. Add in room gain, and this speaker operates virtually 20hz-20,000hz as a 2.5 way. This all works without dropping the load impedance down to the one ohm region because I'm actually also flattening out the upper BR impedance peak here with this network
The only significant SPL compromise is that woofer excursion peaks at 40 hz, but you still have to have it above 110db to really see the cones move at that frequency.It's in a 20" wide 5' tall cabinet. I've found that matching the horizontal dispersions of the upper woofer and the CD horn at xover solved many of the potential issues matching a direct radiator bass with a horn hf. However, the 2226 by itself is only 97 db at that frequency. Radiation pattern narrowing matches up with the horn lateral dispersion at xover which I think really helps make a seamless transition. Then I throw in 4 db boost just below the xover frequency with a somewhat peaked *elliptic* low pass for the upper woofer only. Why elliptic? It's non ripple, and the delay increases near xover to make a smooth group delay transition to the set back horn driver vc and compensate for the naturally increasing group delay as the frequency decreases down to below 200hz. Plus, it's easy to set the notch right where JBL's graphs show a little unruly distortion behavior from the 2226, I believe around 1.5 khz in this case. Plus, it starts out rolling off almost like a first order xover before the response drops off a cliff. The two lower woofs come in right when the peaking rolls off on the low side for the upper woofer to maintain flat response. I'm rolling them off pretty close to 2nd order Butterworth which has slightly less group delay yet but the delay peaks slightly at xover. One nice thing about this arrangement, IMO, is that above 350 hz, the speaker radiates like a 2 speaker 2 way on a 3' shelf, which give a credible height simulation of 'real' musicians in the voice band and above but the two lower woofers step in to keep cone excursion (and doppler distortion) way down below that frequency except for the most massive bass dynamics.
I run my 2206 in a reflex enclosure slightly larger than optimal and tuned to 45 Hz (fs = 50 Hz) and use some EQing. It gives an f3 of 40 Hz theoretically. Lower tuning would give lower cutoff frequency and more SPL capability a the lower end at the cost of les SPL capability in the 60 - 100 hz region.
At the levels I listen to nowadays it works quite well, though I don't use that one very often anymore.
Regards
Charles
Btw, the filter I referred to is not actually an equiripple stopband 'elliptic', per se although with a slight adjustment in component values it could be. This is because I'm peaking at resonance then transitioning to the stop band with an initially slow rolloff similar to a first order LP filter, although the group delay relative to DC is much greater than an actual first order LP filter at its equivalent -3db point.
I also forgot to mention that baffle step would boost the upper 2226J on axis response a db or two near 700 hz and that the baffle width was specifically chosen to take advantage of both this and the associated dispersion narrowing to better match that of the CD horn at xover. Vertically, the upper woofer is as close to the CD horn as I could make it to reduce vertical lobing, but the stopband zero filters I'm using for both the LF and HF drivers give a material assist here on top of their quasi first order xover characteristic.
I have gotten very recognizable square waves from this speaker on axis with the HF horn at just over 1khz in my basement. I haven't tried at lower frequencies.
I also forgot to mention that baffle step would boost the upper 2226J on axis response a db or two near 700 hz and that the baffle width was specifically chosen to take advantage of both this and the associated dispersion narrowing to better match that of the CD horn at xover. Vertically, the upper woofer is as close to the CD horn as I could make it to reduce vertical lobing, but the stopband zero filters I'm using for both the LF and HF drivers give a material assist here on top of their quasi first order xover characteristic.
I have gotten very recognizable square waves from this speaker on axis with the HF horn at just over 1khz in my basement. I haven't tried at lower frequencies.
Basement Blaster
Gotta be TD. Especially after the crossover discussion. Left my head spinning, but I think I get the drift.
I have a two pair of 2226s in JBL 4648-8 cabinets I got from the JBL tent sale. And yes, they can be EQ'd or bass boosted down to resonance or below with no apparent sonic penalty. I think these boxes are 8.5 cu ft, or close, so maybe 40 Hz.
I might take a pair of drivers out for the 40 Hz horns.
TD -- Why not build a midbass horn? You got the drivers, and I can give you some dimensions for a horn if your interested.
Might give you a bit of snap on the bottom end for you.
Let me know if your interested.
As far as going below resonance, I don't do it unless in a horn where the cone is somewhat controlled by the acoustic load.
But, good drivers seem to handle it just fine so long as they are not near their excursion or power handling limits for the random momentary signal. 80% is about tops with most drivers and still remain linear.
Tim
Gotta be TD. Especially after the crossover discussion. Left my head spinning, but I think I get the drift.
I have a two pair of 2226s in JBL 4648-8 cabinets I got from the JBL tent sale. And yes, they can be EQ'd or bass boosted down to resonance or below with no apparent sonic penalty. I think these boxes are 8.5 cu ft, or close, so maybe 40 Hz.
I might take a pair of drivers out for the 40 Hz horns.
TD -- Why not build a midbass horn? You got the drivers, and I can give you some dimensions for a horn if your interested.
Might give you a bit of snap on the bottom end for you.
Let me know if your interested.
As far as going below resonance, I don't do it unless in a horn where the cone is somewhat controlled by the acoustic load.
But, good drivers seem to handle it just fine so long as they are not near their excursion or power handling limits for the random momentary signal. 80% is about tops with most drivers and still remain linear.
Tim
Vitavox S2
Although being a 1,5'' throat unit, the Vitavox S2 claims response down to 200Hz.. It's spec sheets also says that crossed below 500Hz, a 12dB/oct filter should be used. Any idea how deep this thing can perform well when crossed in the lower end by a 24dB/oct active filter? 300Hz-350Hz XO with a 200Hz horn??
Any decent compression drivers that can go all the way down to 150-200Hz?
Although being a 1,5'' throat unit, the Vitavox S2 claims response down to 200Hz.. It's spec sheets also says that crossed below 500Hz, a 12dB/oct filter should be used. Any idea how deep this thing can perform well when crossed in the lower end by a 24dB/oct active filter? 300Hz-350Hz XO with a 200Hz horn??
Any decent compression drivers that can go all the way down to 150-200Hz?
>As far as going below resonance, I don't do it unless in a horn where the cone is somewhat controlled by the acoustic load.
But, good drivers seem to handle it just fine so long as they are not near their excursion or power handling limits for the random momentary signal. 80% is about tops with most drivers and still remain linear.
====
FWIW, assuming sufficient Xmax for the max PE available, 70.7% (backside of the lower impedance peak) is the limit before it's just 'swinging in the breeze' in a vented design, including backhorns.
====
>Any decent compression drivers that can go all the way down to 150-200Hz?
====
Any of the old Altec large format drivers can be XO'd at 300Hz/2nd as long as the lens is large enough to support 150Hz. For HIFI apps, it can be XO'd lower if a large enough lens is used, with lower peak power handling of course. The 299A is the Hot Ticket.
GM
But, good drivers seem to handle it just fine so long as they are not near their excursion or power handling limits for the random momentary signal. 80% is about tops with most drivers and still remain linear.
====
FWIW, assuming sufficient Xmax for the max PE available, 70.7% (backside of the lower impedance peak) is the limit before it's just 'swinging in the breeze' in a vented design, including backhorns.
====
>Any decent compression drivers that can go all the way down to 150-200Hz?
====
Any of the old Altec large format drivers can be XO'd at 300Hz/2nd as long as the lens is large enough to support 150Hz. For HIFI apps, it can be XO'd lower if a large enough lens is used, with lower peak power handling of course. The 299A is the Hot Ticket.
GM
Comp Driver
Some of the rather exotic Ale or Gotto might get close to 200 Hz, and I think one may even go below that, but you would probably need a 80 Hz horn.
RCA made drivers that came close 60 - 70 years ago, and I have read recently of renewed interest in building something like them again. Prototypes may be underway. Takes a big driver.
No exposure to the Vitavox, but you could probably figure the lower lmit based on resonance plus a bit. No gaurantee on the quality of sound that low. Ordinarily, throat distortion goes up near resonance.
Tim
Some of the rather exotic Ale or Gotto might get close to 200 Hz, and I think one may even go below that, but you would probably need a 80 Hz horn.
RCA made drivers that came close 60 - 70 years ago, and I have read recently of renewed interest in building something like them again. Prototypes may be underway. Takes a big driver.
No exposure to the Vitavox, but you could probably figure the lower lmit based on resonance plus a bit. No gaurantee on the quality of sound that low. Ordinarily, throat distortion goes up near resonance.
Tim
Depends on which series you're comparing to. It's more efficient and/or has closer tolerances and/or more power handling than some, so impulse response/HF BW/peak SPL are the variables for a given diaphragm. These drivers followed the same development curve as the small format in that as the company changed management over the decades, they went from high quality to so-so, to poor, and finally got back to ~where they started, except with more power handling.
For a variety of reasons, I haven't used large format drivers in my main rig since the early '70s, and at this point, if I do a major upgrade again, it will be to a DIY 80Hz Unity concept horn perched on top of a kit SD Contrabass.
GM
For a variety of reasons, I haven't used large format drivers in my main rig since the early '70s, and at this point, if I do a major upgrade again, it will be to a DIY 80Hz Unity concept horn perched on top of a kit SD Contrabass.
GM
used ofcourse..ping tim moorman
hi tim your system sounds cool....can you post some pics
p.s didn't understand what you were saying about evm15l's...did you consider them unsuitable for 80 hz straight horns or were you saying they wouldn't be good down to 40hz in a bass horn?
i've got a pair (building karlsons for bass up to 500hz) and they seem to model ok in mcbean for either purpose
cheers
gary mcdonnell-thomas
hi tim your system sounds cool....can you post some pics
p.s didn't understand what you were saying about evm15l's...did you consider them unsuitable for 80 hz straight horns or were you saying they wouldn't be good down to 40hz in a bass horn?
i've got a pair (building karlsons for bass up to 500hz) and they seem to model ok in mcbean for either purpose
cheers
gary mcdonnell-thomas
Rocky said:
Like this, Rocky
http://www.diyaudio.com/forums/showthread.php?postid=127766#post127766
cheers
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.