Nigel,
I see now you are at a typical crossroad deciding which driver and enclosure may fit your needs 🙂
On the driver side, we still would need complete info from the manufacturers on the T/S parameters. Only Selenium shows all the needed parameters for a complete simulation. I would suggest asking next week to the manufacturers the following missing parameters: Force Factor (Bl), Moving Mass (Mms) and the Mechanical Resistance of the Driver's Suspension (Rms). Looking at any Selenium driver specs, you can get the Portuguese translation for these parameters. As I wrote before, I would also ask for frequency plots of SPL, THD, Impedance and Phase.
On enclosures:
1. Sealed enclosures. I like sealed enclosures because they are simple and they have high WAF when you design the speaker to face toward the floor. When you combine a speaker with low Qts and Vas, you generally end with a feasible design for a room as big as yours (I'm supposing 100 m3). I made a simulation with the Selenium speakers which have complete T/S parameters. The Selenium 18SW1P would need a huge amount of power to provide enough SPL and a big enclosure. Then I tried the same with 15SW3P and ended with a 40 cm-sided cube with a 150W amplifier. In this design, an equalizer would be needed.
2. H-Frame OB. I like the H-frame enclosure because they are also very simple to build and do not have inherent problems with the ressonances of long transmission lines. You have to choose speakers with higher Qts with this kind of enclosure. The H-frames I have seen with one pair of speakers are generally dipoles but monopoles are getting more common nowadays. I think they do not have a good WAF because the subwoofers generally don't have a good appearance (that's why I look forward on the Bravox PSW, it's the only good looking one!) and they face towards the room. I don't have the Martin J.King's worksheet yet in able to work in some simulations though. He has worked with several cheap speaker on his OB bass designs and I suppose he measured some of the T/S parameters (i.e. Goldwood Sound speakers don't have T/S parameters on their site at all).
3. Folded Horns. I like to take advantage of this design on full range drivers where you have to get the most of SPL which I don't think it would be the case here. The cons of this design are the resonnances that are introduced every time you fold the horn. The amount of trial on stuffing gives a clue on how this really intereferes. Highef wrote that he took the calculations from Martin J. King. I am curious to know how much power the amp delivers to fill the room and if he has any measurements on distortion. On the WAF side, I can tell you that tall pipes maybe share the highest WAF of all.
You may have noticed that I included the WAF parameter in all of them. That is what can make the difference between your wife calling your design as "another box" or a "nice loudspeaker that blends with the furniture".
I see now you are at a typical crossroad deciding which driver and enclosure may fit your needs 🙂
On the driver side, we still would need complete info from the manufacturers on the T/S parameters. Only Selenium shows all the needed parameters for a complete simulation. I would suggest asking next week to the manufacturers the following missing parameters: Force Factor (Bl), Moving Mass (Mms) and the Mechanical Resistance of the Driver's Suspension (Rms). Looking at any Selenium driver specs, you can get the Portuguese translation for these parameters. As I wrote before, I would also ask for frequency plots of SPL, THD, Impedance and Phase.
On enclosures:
1. Sealed enclosures. I like sealed enclosures because they are simple and they have high WAF when you design the speaker to face toward the floor. When you combine a speaker with low Qts and Vas, you generally end with a feasible design for a room as big as yours (I'm supposing 100 m3). I made a simulation with the Selenium speakers which have complete T/S parameters. The Selenium 18SW1P would need a huge amount of power to provide enough SPL and a big enclosure. Then I tried the same with 15SW3P and ended with a 40 cm-sided cube with a 150W amplifier. In this design, an equalizer would be needed.
2. H-Frame OB. I like the H-frame enclosure because they are also very simple to build and do not have inherent problems with the ressonances of long transmission lines. You have to choose speakers with higher Qts with this kind of enclosure. The H-frames I have seen with one pair of speakers are generally dipoles but monopoles are getting more common nowadays. I think they do not have a good WAF because the subwoofers generally don't have a good appearance (that's why I look forward on the Bravox PSW, it's the only good looking one!) and they face towards the room. I don't have the Martin J.King's worksheet yet in able to work in some simulations though. He has worked with several cheap speaker on his OB bass designs and I suppose he measured some of the T/S parameters (i.e. Goldwood Sound speakers don't have T/S parameters on their site at all).
3. Folded Horns. I like to take advantage of this design on full range drivers where you have to get the most of SPL which I don't think it would be the case here. The cons of this design are the resonnances that are introduced every time you fold the horn. The amount of trial on stuffing gives a clue on how this really intereferes. Highef wrote that he took the calculations from Martin J. King. I am curious to know how much power the amp delivers to fill the room and if he has any measurements on distortion. On the WAF side, I can tell you that tall pipes maybe share the highest WAF of all.
You may have noticed that I included the WAF parameter in all of them. That is what can make the difference between your wife calling your design as "another box" or a "nice loudspeaker that blends with the furniture".
Last edited:
Hi pmolina,
Thanks for all the info. I've a couple of queries about what you wrote:
Definitely, the good news is that since I am building the preamp and xo first I have plenty of time to think the design of the speakers through carefully. (And to absorb the help I'm getting from you guys.)
We already have some of this info - NAR gives "BL -TM" on the chart, which I guess is Bl, and Keybass give Rms. Keybass also give reponse curves, but I am a little sceptical about these, since two drivers apparently have exactly the same curve! I have asked NAR for curves already, but I imagine tomorrow is their first day back after New Year's , so we will see...
First, what sort of ranges would you say were "low Qts and Vas" for a 10" or 12" driver? (Room volume is pretty close to 100m3, I'd say.) I've been playing a little with gspeakers and WinISD doing volume calculations for sealed and ported boxes, to get an idea of how things would be, and I have AkAbak running under wine (I use linux), so when I understand it a little better I can try to simulate some slightly more complicated things.
Higher Qts would mean what sort of range? I could stretch speaker cloth over the front of an H-frame to make it a little more attractive, if the driver is ugly... One attractive feature of H-frames is that if I don't like the result I can knock them aprart and build something else with little effort lost. (Same applies to sealed boxes, for the most part.)
Highef: If you see this, could you tell us how powerful your amps on the NAR10" are? Are you planning on doing two boxes for stereo, or just a central sub?
I've decided a few things: First, I'm going to do do a pair of boxes, one for each channel, rather than just one sub. Second, there are at least three driver options from Brazilian manufacturers that will do the job, one way or another, in either 10" or 12" versions. (NAR 10SW2 or 12SW2, Keybass Panther 10 or 12, and Bravox PSW10 or 12. While I like the build quality of the Selenium speakers I've looked at they all have much higher Fs, as far as I can tell, and moving up to 15" drivers seems uneccessary when there are good options available in smaller sizes.) Third, I am going to lower the crossover point (following advice above) from the 300Hz I originally planned; if I use sealed boxes or H-frames 150-200Hz seems reasonable, if I do the TLs then lower. In this regard I should mention that I have dug out information on the B&Ws that will do mid and highs; they are 3db down at 70Hz, so crossover point should be high enough that this fall-off isn't in the crossover region.
OK, thanks again for all the help.
Cheers
Nigel
Thanks for all the info. I've a couple of queries about what you wrote:
Definitely, the good news is that since I am building the preamp and xo first I have plenty of time to think the design of the speakers through carefully. (And to absorb the help I'm getting from you guys.)
We already have some of this info - NAR gives "BL -TM" on the chart, which I guess is Bl, and Keybass give Rms. Keybass also give reponse curves, but I am a little sceptical about these, since two drivers apparently have exactly the same curve! I have asked NAR for curves already, but I imagine tomorrow is their first day back after New Year's , so we will see...
First, what sort of ranges would you say were "low Qts and Vas" for a 10" or 12" driver? (Room volume is pretty close to 100m3, I'd say.) I've been playing a little with gspeakers and WinISD doing volume calculations for sealed and ported boxes, to get an idea of how things would be, and I have AkAbak running under wine (I use linux), so when I understand it a little better I can try to simulate some slightly more complicated things.
Higher Qts would mean what sort of range? I could stretch speaker cloth over the front of an H-frame to make it a little more attractive, if the driver is ugly... One attractive feature of H-frames is that if I don't like the result I can knock them aprart and build something else with little effort lost. (Same applies to sealed boxes, for the most part.)
Highef: If you see this, could you tell us how powerful your amps on the NAR10" are? Are you planning on doing two boxes for stereo, or just a central sub?
I've decided a few things: First, I'm going to do do a pair of boxes, one for each channel, rather than just one sub. Second, there are at least three driver options from Brazilian manufacturers that will do the job, one way or another, in either 10" or 12" versions. (NAR 10SW2 or 12SW2, Keybass Panther 10 or 12, and Bravox PSW10 or 12. While I like the build quality of the Selenium speakers I've looked at they all have much higher Fs, as far as I can tell, and moving up to 15" drivers seems uneccessary when there are good options available in smaller sizes.) Third, I am going to lower the crossover point (following advice above) from the 300Hz I originally planned; if I use sealed boxes or H-frames 150-200Hz seems reasonable, if I do the TLs then lower. In this regard I should mention that I have dug out information on the B&Ws that will do mid and highs; they are 3db down at 70Hz, so crossover point should be high enough that this fall-off isn't in the crossover region.
OK, thanks again for all the help.
Cheers
Nigel
Has it been said yet? A 'Brazilian woofers'- that's a heck of a lot of woofers...
Still lurking; working on vrious 'projects'- hope to contribute something 'discussion worthy' eventually.
Still lurking; working on vrious 'projects'- hope to contribute something 'discussion worthy' eventually.
I suggest in making a bit of housekeeping so we narrow our conclusions. I've attached a spreadsheet as a starting point for compilation of the T/S parameters of the drivers we are considering. I just included the Selenium driver because it has the complete parameters but I agree it may not be a good contender due to its large size. If you could update the spreadsheet as soon as you get the missing parameters from the other manufacturers, I think we can end up with a very useful reference document.
I don't expect that every manufacturer has plotted the curves for their speakers (unfortunately) so maybe you'll miss some answers. But I do expect that they complete their T/S specs.
On sealed enclosures, low Qts is something less than 0.5. Qts between 0.5 and 0.8 plays well on ported and sealed enclosures. This is a general guideline and I wouldn't solely rely on a single T/S parameter. Vas is something you'll see that for sealed enclosure you'll get values less than or around 100l.
On H-frame OB, high Qts would mean something greater than 0.8 but it's just a guideline. If you read Martin J King's abstracts, you'll see that he also considers lower Qes drivers (although all his project end up using high Qts drivers). I'm missing a deeper study on his calculations and I don't have his spreadsheet to give you a more considerate conclusion on this.
I can't avoid asking 😀 but why are you designing the xover now if you might end up needing equalization on either transmission lines after the tests/simulations with each one of the cabinets (woofer and monitors)?
I don't expect that every manufacturer has plotted the curves for their speakers (unfortunately) so maybe you'll miss some answers. But I do expect that they complete their T/S specs.
On sealed enclosures, low Qts is something less than 0.5. Qts between 0.5 and 0.8 plays well on ported and sealed enclosures. This is a general guideline and I wouldn't solely rely on a single T/S parameter. Vas is something you'll see that for sealed enclosure you'll get values less than or around 100l.
On H-frame OB, high Qts would mean something greater than 0.8 but it's just a guideline. If you read Martin J King's abstracts, you'll see that he also considers lower Qes drivers (although all his project end up using high Qts drivers). I'm missing a deeper study on his calculations and I don't have his spreadsheet to give you a more considerate conclusion on this.
I can't avoid asking 😀 but why are you designing the xover now if you might end up needing equalization on either transmission lines after the tests/simulations with each one of the cabinets (woofer and monitors)?
I think it would be better with just one of 12", you better make your own design loudspeaker, not is difficult to calculate:
DiyAudio.com.br: Calculando um gabinete em Linha de Transmissão - Parte 1
I did not look around the topic, but they Horns they are loudspeaker of high yield not are resonant and use the compliance of air to increase the efficiency who gets to have 30% more the other box.
Are used in professional audio here are known as "cornetas" used in the range of 200 Hz. Also are used in tube amp due to efficiency, I do not on this box.
Polypropylene cone
cone is too heavy! Hi
Did not understand what you say, but the speaker has not yet chosen after that comes the project and building. 🙂
Hi Everyone,
I woke up to find loads of posts on this thread, so I'll try to answer while I have coffee....
Well, it *is* a lot of woofers, if you include PA and car audio stuff, but there are (apparently) *none* made for normal home use, hence the need to discuss what can be done.
The spreadsheet is an excellent idea; I will include all the numbers I've got, and send emails for the ones missing. I think it is a very good idea to include all the drivers we have discussed we have, even if not appropriate for my project, since it may be a useful resource for other builders.
A good question - maybe I should explain the electronic part of the plan in more detail. (And why I ended up doing it like this...) My main system in the living room sounds better when I use the B1 preamp I made for my office system, so for some time I have wanted to do a new preamp for the living room. (And make it pretty also, for WAF - the B1 is in an old computer PSU case...) On the PassLabs forum, juma posted a thread
http://www.diyaudio.com/forums/pass-labs/146310-bf862-preamp.html
on a similar preamp using the BF862 jfet, which is designed to mate well with the F5 power amp, and just looked perfect for what I want. Separately, Jacques Merde started a thread
http://www.diyaudio.com/forums/pass-labs/156094-b1-active-crossover.html
which discusses a very elegant xo using several copies of a B1-like buffer. You can read there about the version of this I am planning, although the circuit won't be in final form until I decide what I'm doing about speakers. So, chronologically, the order is: Build a working preamp first, using juma's circuit, with the ouput buffer usable either full range, or for high-pass. (This option will not be switchable from the outside, but will be changeable inside the case, with some kind of jumpers.) A second output buffer will be necessary for LP also, but doesn't have to be done until later. The xo circuit (which will be some version of Jacques Merde's idea) is to be a separate board in the same case as the preamp, and will be made last, so the design can be altered if necessary once the speakers are done (or at least designed). The resistors and caps that set the crossover frequency are to be changeable (dip sockets and headers, as AndrewT suggested on the other thread) so the same preamp/xo combo can be used with other speaker combinations later with relatively small changes. If any equalisation is necessary it would be good to make it similarly "adjustable"... So, the preamp is under construction (and will take a while - all the parts haven't arrived), but the xo isn't yet, and (as you say) will not have the design finalised until this discussion here is finished, or at least farther along...
When you say "just one" you mean on each channel, right? Earlier on I considered the possibility of just one subwoofer, but I have decided to do a stereo pair. This doubles the cost, of course, but there are several reasons why I'm going this way.
I've been reading the same link you posted...
Would you only use paper cones?
Um abraco
Nigel
I woke up to find loads of posts on this thread, so I'll try to answer while I have coffee....
Well, it *is* a lot of woofers, if you include PA and car audio stuff, but there are (apparently) *none* made for normal home use, hence the need to discuss what can be done.
The spreadsheet is an excellent idea; I will include all the numbers I've got, and send emails for the ones missing. I think it is a very good idea to include all the drivers we have discussed we have, even if not appropriate for my project, since it may be a useful resource for other builders.
A good question - maybe I should explain the electronic part of the plan in more detail. (And why I ended up doing it like this...) My main system in the living room sounds better when I use the B1 preamp I made for my office system, so for some time I have wanted to do a new preamp for the living room. (And make it pretty also, for WAF - the B1 is in an old computer PSU case...) On the PassLabs forum, juma posted a thread
http://www.diyaudio.com/forums/pass-labs/146310-bf862-preamp.html
on a similar preamp using the BF862 jfet, which is designed to mate well with the F5 power amp, and just looked perfect for what I want. Separately, Jacques Merde started a thread
http://www.diyaudio.com/forums/pass-labs/156094-b1-active-crossover.html
which discusses a very elegant xo using several copies of a B1-like buffer. You can read there about the version of this I am planning, although the circuit won't be in final form until I decide what I'm doing about speakers. So, chronologically, the order is: Build a working preamp first, using juma's circuit, with the ouput buffer usable either full range, or for high-pass. (This option will not be switchable from the outside, but will be changeable inside the case, with some kind of jumpers.) A second output buffer will be necessary for LP also, but doesn't have to be done until later. The xo circuit (which will be some version of Jacques Merde's idea) is to be a separate board in the same case as the preamp, and will be made last, so the design can be altered if necessary once the speakers are done (or at least designed). The resistors and caps that set the crossover frequency are to be changeable (dip sockets and headers, as AndrewT suggested on the other thread) so the same preamp/xo combo can be used with other speaker combinations later with relatively small changes. If any equalisation is necessary it would be good to make it similarly "adjustable"... So, the preamp is under construction (and will take a while - all the parts haven't arrived), but the xo isn't yet, and (as you say) will not have the design finalised until this discussion here is finished, or at least farther along...
When you say "just one" you mean on each channel, right? Earlier on I considered the possibility of just one subwoofer, but I have decided to do a stereo pair. This doubles the cost, of course, but there are several reasons why I'm going this way.
I've been reading the same link you posted...
Polypropylene cone
Would you only use paper cones?
Um abraco
Nigel
I was talking about a one sub each box, I also like the stereo configuration more should be to 3-way each box. I have some drawings in boxes 3 way for LT.
Maybe two sub 12" an exaggeration, see :
HT Forum - Ver Mensagem Isolada - SUBLIME uma pergunta indiscreta!!!
SUBLIME uma pergunta indiscreta!!! - HT Forum
Yes preferably, cone light and a low frequency resonance indicate speed of response, quite different from a cone Polypropylene where they can use a heavy cone to lower the resonance frequency is hard to dampen the cone. Only with closed-Box you will get.
I was checking on the sites. It seems all of the drivers we are looking at here are polypropylene cones, except maybe some of the heavier ones which may be aluminium. Do you know of any paper ones?
I'll check the links shortly.
Um abraco
Nigel
The information not is correct at in site Keybass, the models Panther are cellulose, the models NAR are also cellulose.
- Status
- Not open for further replies.