Hi Jussi
I'm following your development, and am inspired by your project to get my own dream dipole system going . .
> Along sidewalls capacity increases another 6 dB.
With lower bass to be through the compact Phoenix XLS woofers that I've already built, I'm planning the rest of an open baffle system. With separate baffles about 1.2 m high * (folding out to) 1.8 m wide, to be mounted up the side walls (wide room, but little spare floor space).
The lowest range covered by the separate baffles, will be 15" midbass (about 80 - 500 Hz) with drivers centred 200 - 300 mm from the side walls. The upper mid and tweeters would/ could be a similar or closer distance to the side walls ~
I've read Linkwitz's site and John k's several times, others stuff on OBs, but don't recall the 6 dB gain. (my drivers are 97 - 100 dB: 6 dB gain would mean I could get away with a 2 watt 45 SE tube amp, which would be great!)
(this the horizontal equivalent to mirroring from 'floor bounce')
- is it a full 6 dB of gain?
- flat over 80 Hz - 20 kHz
or just at the low end?
Thanks
Rick
I'm following your development, and am inspired by your project to get my own dream dipole system going . .
> Along sidewalls capacity increases another 6 dB.
With lower bass to be through the compact Phoenix XLS woofers that I've already built, I'm planning the rest of an open baffle system. With separate baffles about 1.2 m high * (folding out to) 1.8 m wide, to be mounted up the side walls (wide room, but little spare floor space).
The lowest range covered by the separate baffles, will be 15" midbass (about 80 - 500 Hz) with drivers centred 200 - 300 mm from the side walls. The upper mid and tweeters would/ could be a similar or closer distance to the side walls ~
I've read Linkwitz's site and John k's several times, others stuff on OBs, but don't recall the 6 dB gain. (my drivers are 97 - 100 dB: 6 dB gain would mean I could get away with a 2 watt 45 SE tube amp, which would be great!)
(this the horizontal equivalent to mirroring from 'floor bounce')
- is it a full 6 dB of gain?
- flat over 80 Hz - 20 kHz
or just at the low end? Thanks
Rick
Rick,
6dB is based on Linkwitz and overally commonly known theory (and practice) of limiting radiating pattern to 1/x space. For example floor does it pretty automatically while system is close enough to it. It turns full space radiation to half space radiation and therefore increases output 6dB. According to Linkwitz this effect occurs between 100Hz and 200Hz with normal height systems. So he uses 100-200HP filterblock to adjust the overall balance. 100-200HP is a shelvingfilter that drops on axis response below 200Hz. -6dB at 100Hz.
This effect can be increased by positioning system along sidewalls which limits radiation patternr from half space (floor already limiting) to quarter-space. Again 6dB gain.
But this setup changes dipole configuration. It doesn't work as a dipole that well anymore. Basically limiting dipoles radiation pattern increases it's "D". For example sidewall forms a mirrorsource from the real dipole which increases dipole systems "D" to double it's value. This increases the output but like you propably know larger "D", lower the point where configuration isn't a true dipole anymore. Therefore I'd recommend it for the bass department only. For the midrange dipole radiation just ends causing pretty dramatical diffractions from the sidewall.
Perhaps in your case you can use your 15" lowmid close to sidewall if you place it in very modest size baffle to obtain dipole pattern up to 500Hz.
It actually isn't that hard to achieve high efficiency from dipole system. Especially if you use PA drivers to do it. Personally I consider many PA drivers (specially mids and tweeters) pretty crude for home environments and therefore I'm not going towards that path. My own configuration is about 94-94,5dB sensitive at W18Es midrange area. You can use the floor boost as well as potential wallboost for your advantage as long as you remember the limitations in radiation pattern. In my case I'd estimate the lowmids (25Ws) are 94dB efficient alone, they get 6dB boost at the lowend of their range and also woofers cross at 80-100Hz is helping them another 6dB.
Jussi
6dB is based on Linkwitz and overally commonly known theory (and practice) of limiting radiating pattern to 1/x space. For example floor does it pretty automatically while system is close enough to it. It turns full space radiation to half space radiation and therefore increases output 6dB. According to Linkwitz this effect occurs between 100Hz and 200Hz with normal height systems. So he uses 100-200HP filterblock to adjust the overall balance. 100-200HP is a shelvingfilter that drops on axis response below 200Hz. -6dB at 100Hz.
This effect can be increased by positioning system along sidewalls which limits radiation patternr from half space (floor already limiting) to quarter-space. Again 6dB gain.
But this setup changes dipole configuration. It doesn't work as a dipole that well anymore. Basically limiting dipoles radiation pattern increases it's "D". For example sidewall forms a mirrorsource from the real dipole which increases dipole systems "D" to double it's value. This increases the output but like you propably know larger "D", lower the point where configuration isn't a true dipole anymore. Therefore I'd recommend it for the bass department only. For the midrange dipole radiation just ends causing pretty dramatical diffractions from the sidewall.
Perhaps in your case you can use your 15" lowmid close to sidewall if you place it in very modest size baffle to obtain dipole pattern up to 500Hz.
It actually isn't that hard to achieve high efficiency from dipole system. Especially if you use PA drivers to do it. Personally I consider many PA drivers (specially mids and tweeters) pretty crude for home environments and therefore I'm not going towards that path. My own configuration is about 94-94,5dB sensitive at W18Es midrange area. You can use the floor boost as well as potential wallboost for your advantage as long as you remember the limitations in radiation pattern. In my case I'd estimate the lowmids (25Ws) are 94dB efficient alone, they get 6dB boost at the lowend of their range and also woofers cross at 80-100Hz is helping them another 6dB.
Jussi
Jussi
Thanks, nearly all of what you say 'rings a bell' or makes sense. To clarify a couple of things -
So the half space boosts all frequencies equally. But at higher freqencies, where output is rising 6 dB/ octave - gain isn't needed . .
. . and the half space bounce causes some frequency response aberations ?
(btw the drivers I'm likely to use if they measure well, are Lambda for midbass, PHL (PA/ studio?) mids and Aurum Cantus ribbons)
The 15" midbass could be crossed nearly an octave lower, say 300 Hz, that would lower diffractions ~ likely to be usefully lower?
(I was thinking of a 1.2* 1.8m baffle to extend the bass depth) - how would the midbass' diffractions be reduced by a smaller baffle?
Thanks,
Rick
Thanks, nearly all of what you say 'rings a bell' or makes sense. To clarify a couple of things -
So the half space boosts all frequencies equally. But at higher freqencies, where output is rising 6 dB/ octave - gain isn't needed . .
. . and the half space bounce causes some frequency response aberations ?
(btw the drivers I'm likely to use if they measure well, are Lambda for midbass, PHL (PA/ studio?) mids and Aurum Cantus ribbons)
The 15" midbass could be crossed nearly an octave lower, say 300 Hz, that would lower diffractions ~ likely to be usefully lower?
(I was thinking of a 1.2* 1.8m baffle to extend the bass depth) - how would the midbass' diffractions be reduced by a smaller baffle?
Thanks,
Rick
Why not connect two XXLS in series? The limiting factor in the bass is usually excursion, not maximum amp rail voltage.
Pro woofers: I had some reservations as well, but I wanted to try it and I was converted... I changed my two 10" HiFi woofers for a top of the line 15" pro woofer. The sound quality just jumped to a different league. I was completely taken by surprise. Even the lower midrange improved, compared to an already very good HiFi driver. So I'd recommend not to be afraid to go large and go pro sound.
Something else to consider: thinking is OK but nothing beats trying. Your system sans deep bass woofers will be almost complete for most music even with nothing below 80 Hz. So I'd finish and optimize all else first, before tackling the bass. Keeps you busy too...
And then you can try ugly test bass cabinets, with say the XXLS's first. Reason, they are well known, well respected, and if you don't like the results you can market them at little loss. If they don't work out, then I'd try one or two pro 15", again, ideally something marketable in case you don't like them... With the cost of the rest of the system, the loss of such a buy and sell approach will not be significant, and it will get you real life experience on how the components work for you. I don't think it is possible, in all humility, to build such a system without at least some prototyping, from scratch and pen and paper alone.
Pro woofers: I had some reservations as well, but I wanted to try it and I was converted... I changed my two 10" HiFi woofers for a top of the line 15" pro woofer. The sound quality just jumped to a different league. I was completely taken by surprise. Even the lower midrange improved, compared to an already very good HiFi driver. So I'd recommend not to be afraid to go large and go pro sound.
Something else to consider: thinking is OK but nothing beats trying. Your system sans deep bass woofers will be almost complete for most music even with nothing below 80 Hz. So I'd finish and optimize all else first, before tackling the bass. Keeps you busy too...
And then you can try ugly test bass cabinets, with say the XXLS's first. Reason, they are well known, well respected, and if you don't like the results you can market them at little loss. If they don't work out, then I'd try one or two pro 15", again, ideally something marketable in case you don't like them... With the cost of the rest of the system, the loss of such a buy and sell approach will not be significant, and it will get you real life experience on how the components work for you. I don't think it is possible, in all humility, to build such a system without at least some prototyping, from scratch and pen and paper alone.
MBK said:
What are the pro woofers you are using? (brand, model)
Also what other examples of bass woofers usable up to 250hz with low distortion would you list?
I use an 18Sound 15ND930. A lot of discussion and measurements found in
this post
and the following or posts referenced therein. The thread mentioned is all about large OB midranges, a project by PaulW. Bottom line, I am now tempted to skip my midrange and use the 15ND930 up to 1k.
There should be many potential candidates for your application, now that the pro market is chock full of high tech woofers, with demodulation rings and other goodies. I have only looked more closely at the 15TBX100 by B&C. It has excellent features, and Earl Geddes uses it in his Summa speakers up to 900 Hz.
this post
and the following or posts referenced therein. The thread mentioned is all about large OB midranges, a project by PaulW. Bottom line, I am now tempted to skip my midrange and use the 15ND930 up to 1k.
There should be many potential candidates for your application, now that the pro market is chock full of high tech woofers, with demodulation rings and other goodies. I have only looked more closely at the 15TBX100 by B&C. It has excellent features, and Earl Geddes uses it in his Summa speakers up to 900 Hz.
The speakers are up and running. Or at least one of them. I have only one unmodified waveguide and other speakers mid and lowmid drivers are still getting measured.
Sounds pretty promissing. It's diffucult to say anything solid listening to a mono system.
Basically it is. I haven't tried it how high it goes. Radiation pattern problems come along at some point.
At least above the point where acoustically double baffle (double "D" because of the boundary nearby) stops being a dipole.
Ok. I've considered both PHL and Aurum Cantus but they didn't seem promissing (enough). PHL mids have pretty dramatical cone breakup problems in the passband (check the impedance response) while it's very light paper cone changes shape between strong motor and stiff suspension. There seems to be some problems with AC tweeters as well. At least distorsion numbers have been average at the best. But both of them are capable playing plenty of SPL with reasonably low power.
That's more like a open baffle than dipole. So there isn't even a meaning to let the rear radiation at the front and create dipole radiation pattern. I'm not that familiar with open baffles. I've heard a few but I haven't thought that much how they really work.
That's one way to go. At least the XXLS models aren't that dramatical to low impedance as the older XLS models are. 12" XLS has 4,7 ohm minimum impedance, 12" XXLS 6,7 ohms or so.
Yep. Good point. And I have a very long way to go with them mains first. And they don't even seem to lack that much bass. Even Jurassic Lunch test track played decently. Not an earthquake but decent.
Jussi
Sounds pretty promissing. It's diffucult to say anything solid listening to a mono system.
rick57 said:
Basically it is. I haven't tried it how high it goes. Radiation pattern problems come along at some point.
At least above the point where acoustically double baffle (double "D" because of the boundary nearby) stops being a dipole.
Ok. I've considered both PHL and Aurum Cantus but they didn't seem promissing (enough). PHL mids have pretty dramatical cone breakup problems in the passband (check the impedance response) while it's very light paper cone changes shape between strong motor and stiff suspension. There seems to be some problems with AC tweeters as well. At least distorsion numbers have been average at the best. But both of them are capable playing plenty of SPL with reasonably low power.
That's more like a open baffle than dipole. So there isn't even a meaning to let the rear radiation at the front and create dipole radiation pattern. I'm not that familiar with open baffles. I've heard a few but I haven't thought that much how they really work.
MBK said:
That's one way to go. At least the XXLS models aren't that dramatical to low impedance as the older XLS models are. 12" XLS has 4,7 ohm minimum impedance, 12" XXLS 6,7 ohms or so.
Yep. Good point. And I have a very long way to go with them mains first. And they don't even seem to lack that much bass. Even Jurassic Lunch test track played decently. Not an earthquake but decent.
Jussi
Hi Jussi
Any developments on your system?
I said i was thinking of a 1.2* 1.8m baffle, and i was intrigued when you responded . ."That's more like a open baffle than dipole". I think of the 2 terms as different ways of describing the same thing: a dipole figure radiation pattern results from (whatever size) of open baffle.
"PHL mids have pretty dramatical cone breakup problems in the passband"
A friend tested the 10" 100 dB 3451 mid, and got dramatic breakup around 2.2 kHz; I have a pair as an option for this use, the large AC tweeters I have with a steep filter, have been crossed at 1.2 kHz. And after some favorable comments from a Nowegian guy here on the PHL 1040 mids, I also picked up some 6.5" 100 dB , which is another mid option. Though the Lambdas are meant to have excellent midange (very low inductance of 0.2 mH), I may find just running them up to 1.2 kHz withgous mids is fine(?)
(I also have a pair of Seas mag cone W17's, I could use, but I would rearly prefer to use them in another system)
I haven't got a measurement setup just yet - really curious, which PHLs did you test or see results on?
Cheers, Rick
Any developments on your system?
I said i was thinking of a 1.2* 1.8m baffle, and i was intrigued when you responded . ."That's more like a open baffle than dipole". I think of the 2 terms as different ways of describing the same thing: a dipole figure radiation pattern results from (whatever size) of open baffle.
"PHL mids have pretty dramatical cone breakup problems in the passband"
A friend tested the 10" 100 dB 3451 mid, and got dramatic breakup around 2.2 kHz; I have a pair as an option for this use, the large AC tweeters I have with a steep filter, have been crossed at 1.2 kHz. And after some favorable comments from a Nowegian guy here on the PHL 1040 mids, I also picked up some 6.5" 100 dB , which is another mid option. Though the Lambdas are meant to have excellent midange (very low inductance of 0.2 mH), I may find just running them up to 1.2 kHz withgous mids is fine(?)
(I also have a pair of Seas mag cone W17's, I could use, but I would rearly prefer to use them in another system)
I haven't got a measurement setup just yet - really curious, which PHLs did you test or see results on?
Cheers, Rick
Jussi
I should have said, I don't mean to do a tempoary thread hijack, it's just that you are further down the track with assessing drivers 🙂
Thanks, Rick
I should have said, I don't mean to do a tempoary thread hijack, it's just that you are further down the track with assessing drivers 🙂
Thanks, Rick
rick57 said:
Not that much. Got the tweeters and measured two different crossover setups for the Behringer. Later one sounds pretty promissing. 300Hz 2nd and 1800Hz 4th order crossover.
Have you read open baffle information from www.linkwitzlab.com? There is some well explained information about this. Basically the difference is "dipole" vs "dipolar". Certain size baffle works as a dipole up to a certain point. In my case, 34cm (13,4") wide baffle, up to around 500Hz or so. Below that dipoles figure 8 radiation pattern is pretty well achieved. But above that system is just dipolar. It radiates sound in opposite phases front and back but the baffle size doesn't allow them to form a perfect dipole radiation pattern. This causes dipole nulls at some point. In my case first one is around 1300Hz which is also very visible in the off axis measurements. Directivity drops and radiation pattern spreads open for a moment. With 1,2m wide baffle first null is much lower and before your considered crossoverpoint of 1200Hz there is propably second and maybe even third null. Check linkwitzlab for additional information and calculation patterns.
But like said, large open baffle is a different ballgame than smaller mostly dipole operating panel. Both have advantages and disadvantages. Basically whole dipole pattern is achieved and maintained best by using totally open drivers. No baffle at all. But this ruins overall efficiency. As you know the true dipole range has it's -6dB/oct roll off towards bass. For example we measured 7" Peerless HDS without baffle and got poor 75dB sensitivity at 300Hz. So it's more than just the radiation patterns.
It was PHL 1120 midrange unit. I don't know is the same breakup pattern following the drivers in the lineup but there are problems with the papercone. Period. Many people dislike metalcones since they have dramatic peak at upper midrange area requiring steep crossover and a notchfilter. Ok. But the problems is pretty much solved by it. Many 7" papercone midwoofers have a breakup around 1Khz. At the middle of their passband. Naturally it isn't even closely as severe as metalcones but it causes some stored energy and distorsion problems. Naturally this can be sensed as a positive thing depending on what different people like. So the popularity and praised "naturality" of papercone can technically be just a coloration and misbehavior of the driver.
I understanded that you have pretty small room, lack of floorspace and you'd like to have a high efficiency system that you can drive with lowish power tubeamp? So there isn't that much need for tremendously high SPL ?
Some studio manufacturers use PA drivers in their lineup. But what I've heard it's mostly because high SPL figures are needed and that's the only way to get very high sensitivity and powerhandling to do the job. But how refined do they sound, even at modest levels. Many (that I know, propably not all) compression tweeters (as example) have pretty big problems with their frequency response and even distorsion. But if +130dB levels are required it's pretty much the only way to do it. Other solutions offer a second of smoke at those levels.
Personally I'd go for the W17s you have. I use W18Es in my dipole system, 34cm wide baffle with 300Hz 2nd and 1800Hz 4th order crosses at the moment. And they sure play loud and clean. Measured sensitivity using parallel connection is around 94-95dB. Not that much power required. Single or double 15" PA woofer for the bottom, depending on your room and needs. A high quality dome for the treble. Or use the ribbons if you already have them. W17s in separated small panel with the tweeter. Something like the NaO. And then the woofers in larger construction depending on your crossover needs and speaker placement. Digital crossover and delay ease things up a bit. NaO style panel doesn't require much floorspace and is even mobile if you need to move them for vacuuming or such. Woofers along the sidewalls, plenty of output. But this idea is just me ... 😉
I've said this before but it doesn't seem to hurt saying it again.. If a system has to play loud sensitivity is important but at least as important is low distorsion. What's the use to have explosive dynamics and high SPL capacity while sound distorts changing music to noise? I think there is plenty of noise on this planet as it is. I'd concentrate on the quality.
Jussi
Jussii
Yes I have read the at www.linkwitzlab.com . .
while I recall the changing dispersion as Hz changes, I had not picked up on a distinction between "dipole" vs "dipolar" . . thanks for your explanation and I better SL read again too . .
"I understanded that you have pretty small room, lack of floorspace and you'd like to have a high efficiency system that you can drive with lowish power tubeamp? So there isn't that much need for tremendously high SPL? "
The 'floorspace' probelm is not so much space - the room is large : 6.7 * 5.3 m (and cathedral ceiling to 3.8 m) but has bulky furniture in the path of where speakers direct radiation would normally go. With the high ceiling, i will mount the baffles on the side walls, so the mirroring to extend the midwoofers' lower extension a little, will come from 'wall bounce'.
"If a system has to play loud sensitivity is important but at least as important is low distorsion. What's the use to have explosive dynamics and high SPL capacity while sound distorts changing music to noise? I think there is plenty of noise on this planet as it is. I'd concentrate on the quality."
I agree 100%. The high efficiency is not for high SPLs - 105 db is enough on extreme occasions, but will allow the use of a 2 watt tube amp that's intended for the system. The tweeters are not compression, but 'true' ribbons. I will compare the W17s to the PHLs; and try good SS amps of enough power to use them.
You suggested "W17s in separated small panel with the tweeter. Something like the NaO."
As the bass below 80 Hz will be covered by W baffles that i already have, to mimize comb filtering/ lobing, without knowing the relationship between baffle size and Hz where these effects apply, I had been thinking of a separate very small baffle for the tweeters (crossing about 1300 hz, 100 dB).
Thanks,
Rick
Yes I have read the at www.linkwitzlab.com . .
while I recall the changing dispersion as Hz changes, I had not picked up on a distinction between "dipole" vs "dipolar" . . thanks for your explanation and I better SL read again too . .
"I understanded that you have pretty small room, lack of floorspace and you'd like to have a high efficiency system that you can drive with lowish power tubeamp? So there isn't that much need for tremendously high SPL? "
The 'floorspace' probelm is not so much space - the room is large : 6.7 * 5.3 m (and cathedral ceiling to 3.8 m) but has bulky furniture in the path of where speakers direct radiation would normally go. With the high ceiling, i will mount the baffles on the side walls, so the mirroring to extend the midwoofers' lower extension a little, will come from 'wall bounce'.
"If a system has to play loud sensitivity is important but at least as important is low distorsion. What's the use to have explosive dynamics and high SPL capacity while sound distorts changing music to noise? I think there is plenty of noise on this planet as it is. I'd concentrate on the quality."
I agree 100%. The high efficiency is not for high SPLs - 105 db is enough on extreme occasions, but will allow the use of a 2 watt tube amp that's intended for the system. The tweeters are not compression, but 'true' ribbons. I will compare the W17s to the PHLs; and try good SS amps of enough power to use them.
You suggested "W17s in separated small panel with the tweeter. Something like the NaO."
As the bass below 80 Hz will be covered by W baffles that i already have, to mimize comb filtering/ lobing, without knowing the relationship between baffle size and Hz where these effects apply, I had been thinking of a separate very small baffle for the tweeters (crossing about 1300 hz, 100 dB).
Thanks,
Rick
rick57 said:
1,2m wide baffle. How about placing the woofers just along the sidewalls and then have some gap in the baffle and have mid-tweeter department at the other side of the baffle? With uniform baffle it doesn't work that well as a dipole at the midrange. Dipolar but not dipole. Or maybe you could treat the sidewall just at front of the speaker but I don't consider that so good solution.
So the speakers aren't going to be at the normal height either and therefore you need to use wall installation?
Ok. 2 watts, 105dB. Tough one. At least with normal home drivers. I don't have the measurements yet but I'd expect W18E pair play +100dB clean (<0,3-0,4% THD). Seas measurements suggest such a performance for 96dB levels with single driver. I don't know how PHL drivers or Aurum Cantus ribbons handle it. 105dB can be pretty stressed if it runs badly distorted. Ribbons tend to have nasty high order harmonics after they lose it.
So you have W-baffles below 80Hz, consider 15" or two from 80Hz up to 300-500Hz and then PHL-AC top. Ok.
Personally I wouldn't detach tweeter from the midrange construction. It just makes things less stable. They would have to be very close to the mids and still cause combfilter problems.
Jussi
Jussi,
I understand you have a relatively small room so I assume you don't play it loud normally. Have you tried to play it loud for a test? How do your Esotar tweeters perform?
Your tweeters are crossed at 1800Hz 4th order. The driver Fs is 750Hz. I don't know if there is sufficient attenuation, especially when playing loud. Your WG would boost the SPL a bit.
In the process of designing my new MTMWW I am planning on crossing the Esotar at 1900Hz-2000Hz 4th order. I am worrying about stressing the tweeters a bit but could not cross it higher because of the MTM configuration. I don't have the WG. Your experience can save me some time.
Regards,
Bill
I understand you have a relatively small room so I assume you don't play it loud normally. Have you tried to play it loud for a test? How do your Esotar tweeters perform?
Your tweeters are crossed at 1800Hz 4th order. The driver Fs is 750Hz. I don't know if there is sufficient attenuation, especially when playing loud. Your WG would boost the SPL a bit.
In the process of designing my new MTMWW I am planning on crossing the Esotar at 1900Hz-2000Hz 4th order. I am worrying about stressing the tweeters a bit but could not cross it higher because of the MTM configuration. I don't have the WG. Your experience can save me some time.
Regards,
Bill
Jussi
> So the speakers aren't going to be at normal height either and therefore you need to use wall installation?
yes . .
> How about placing the woofers just along the sidewalls, and then have some gap in the baffle and have mid-tweeter department at the other side of the baffle?
> Personally I wouldn't detach tweeter from the midrange construction. It just makes things less stable. They would have to be very close to the mids and still cause comb filter problems.
While I'd prefer to design up front more, and experiment less; especally in this unusual situation ~ it seems best to experiment more and try this . .
> With uniform baffle it doesn't work that well as a dipole at the midrange. Dipolar but not dipole.
> Certain size baffle works as a dipole up to a certain point. In my case, 34cm (13,4") wide baffle . . This causes dipole nulls at some point. In my case first one is around 1300 Hz
So to try to put it all together (and re-reading Linkwitz) nulls can be estimated as when D becomes larger than 1/3 of a wavelength . .
. . but (it's not clear), where a baffle (eg U or H) has both width and depth, D = ?
> whole dipole pattern is achieved and maintained best by using totally open drivers. No baffle at all. But this ruins overall efficiency.
is there a formula for lost efficiency?
(maybe if the 98 - 100 dB drivers were used - which can handle 100/ 160/ 300 watts, with a 100 watt SS amp, smaller baffles would get closer to an all dipole pattern?) I'm not wedded to low power tube amp ~, whatever is best overall
Thanks
Rick
> So the speakers aren't going to be at normal height either and therefore you need to use wall installation?
yes . .
> How about placing the woofers just along the sidewalls, and then have some gap in the baffle and have mid-tweeter department at the other side of the baffle?
> Personally I wouldn't detach tweeter from the midrange construction. It just makes things less stable. They would have to be very close to the mids and still cause comb filter problems.
While I'd prefer to design up front more, and experiment less; especally in this unusual situation ~ it seems best to experiment more and try this . .
> With uniform baffle it doesn't work that well as a dipole at the midrange. Dipolar but not dipole.
> Certain size baffle works as a dipole up to a certain point. In my case, 34cm (13,4") wide baffle . . This causes dipole nulls at some point. In my case first one is around 1300 Hz
So to try to put it all together (and re-reading Linkwitz) nulls can be estimated as when D becomes larger than 1/3 of a wavelength . .
. . but (it's not clear), where a baffle (eg U or H) has both width and depth, D = ?
> whole dipole pattern is achieved and maintained best by using totally open drivers. No baffle at all. But this ruins overall efficiency.
is there a formula for lost efficiency?
(maybe if the 98 - 100 dB drivers were used - which can handle 100/ 160/ 300 watts, with a 100 watt SS amp, smaller baffles would get closer to an all dipole pattern?) I'm not wedded to low power tube amp ~, whatever is best overall
Thanks
Rick
HiFiNutNut said:
This room is about 20m2. Few square meters larger than our last livingroom. I haven't tried any box speakers in this room but at least the last one was awful. Boom, messy midrange, can't even play loud without sounding tiring.
I don't have the Esotars anymore. They had too big tolerance in their response. Maybe 2dB or so. Present tweeter is H1214 model from Seas lineup.
Have only single speaker ready and playing but I've fixed the crossover pretty much as ready as I can listening it mono. I've also tried to play them loud. And they can play loud, even very loud, and it doesn't show any sign of stress. 25Ws move violently without highpass filter so 2x10" in flat baffle definitely is a lower midrange solution. With 80Hz 2nd order highpass system can play with very high levels. Jurassic lunch sounded pretty impressive. And the other speaker should add another 6dB or so...
I'm actually quite impressed by these speakers already. Nicely detailed, fast, dynamic and very clear sound. Hope this illusion remains in stereo.
H1214 has 1100Hz Fs. Waveguide assists up to 6-7dB at 1,8Khz starting from 6-7Khz. Sound is very clean. I don't have the final waveguide and driver installation done so it should get better later on...
Some Esotar based commercial speakers use 2Khz 1st order cross with it. For example Sonus Faber Extremis (check Stereophile archives). Esotars are very sturdy made units and can handle quite a hammering. Shame mine weren't in tolerance...
rick57 said:
In practice you have to measure it. Then you also should consider baffle diffraction issues both front and back. For example Phoenix system has pretty severe diffraction problems while the baffle construction is a small "U". I had similar results with my prototype some time ago. Therefore I decided to use strongly curved baffles in this project. All sides have 25x50mm elliptic curve and they seem to work fine. No diffration problems in the on axis response. +-1dB easily.
In theory you can calculate raw drivers "D" but in practice you have to measure it.
Depend on were you consider crossing. Basically there is no reason to use no-baffle solution if you're going to cross below the first null. Same with woofers as well. Design bafflesize and frameshape depending on what kind of crossover do you want to use. Low and steep, go for W. High and gentle, go with I.
Jussi
PnZrFsT said:
If you're referring to my project I don't agree. 🙂
And that's because they aren't finished at any level. Mechanical construction needs spine (10-12mm steelspine), floorplate and panelboards need some paint on them and then some cloth. But finally they should look decent.
Jussi
And here 25W/8565-01 THD @ 90dB SPL. Both measurement have too small closed cabinets so below 100Hz distorsion isn't correct.
That 3rd harmonic peak at 1800Hz with W18 response isn't that good thing. It's from the primary breakup. I guess my present 1800Hz 4th order lowpass with adequate notch filtering is pretty much the maximum for this driver. At least when it comes to distorsion.
25W didn't show that big suprices either. Few 2nd harmonic peaks, pretty similar than smaller brother 21W has. I guess some of them can be damped with magnet support. My present 300Hz 2nd order lowpass might be a bit high for this driver.
Jussi
That 3rd harmonic peak at 1800Hz with W18 response isn't that good thing. It's from the primary breakup. I guess my present 1800Hz 4th order lowpass with adequate notch filtering is pretty much the maximum for this driver. At least when it comes to distorsion.
25W didn't show that big suprices either. Few 2nd harmonic peaks, pretty similar than smaller brother 21W has. I guess some of them can be damped with magnet support. My present 300Hz 2nd order lowpass might be a bit high for this driver.
Jussi
Attachments
- Status
- Not open for further replies.