@tmuikku
Yes, I am somewhat aware. Who(what) do you think I am? :-D
It is also a compromise. It is somewhat desired to have the impedance, power and efficiency linearized, but while it is at the cost of something, one just has to decide where on the scale and which compromise to pick. Reduced cone movement helps, but you know cone movement here makes sound. So there is less sound. More power needs to be put in, to make more sound again, and you are back with the cone movement almost exactly where you were, just for some additional power. That´s not a good strategy. I use ~minimum padding such as in the picture to shave the most uglies (although bitey, sharp sound to the box has some magic too!), but no more, otherwise bleeding performance, efficiency, cooling efficiency.
The top part of the useful range power input is concerning thing. Those long coil drivers not moving around 100Hz really do not like it. Like, AT ALL.
The power profile, where most power goes into driver where it moves a lot is very desirable. Switching that means automatic failure, Heavy compression and risks of damage. That´s one part where these high motor force excell. They have high sensitivity and efficiency too in that range also, so you can afford to EQ them down around 100Hz, and cut sooner, so less compression happens in that range too. Very desirable.
Yes, I am somewhat aware. Who(what) do you think I am? :-D
It is also a compromise. It is somewhat desired to have the impedance, power and efficiency linearized, but while it is at the cost of something, one just has to decide where on the scale and which compromise to pick. Reduced cone movement helps, but you know cone movement here makes sound. So there is less sound. More power needs to be put in, to make more sound again, and you are back with the cone movement almost exactly where you were, just for some additional power. That´s not a good strategy. I use ~minimum padding such as in the picture to shave the most uglies (although bitey, sharp sound to the box has some magic too!), but no more, otherwise bleeding performance, efficiency, cooling efficiency.
The top part of the useful range power input is concerning thing. Those long coil drivers not moving around 100Hz really do not like it. Like, AT ALL.
The power profile, where most power goes into driver where it moves a lot is very desirable. Switching that means automatic failure, Heavy compression and risks of damage. That´s one part where these high motor force excell. They have high sensitivity and efficiency too in that range also, so you can afford to EQ them down around 100Hz, and cut sooner, so less compression happens in that range too. Very desirable.
Yes, that is exactly what I do. I handle compromises. And the conventional design is not working for me favourably. If I played live stuff, or light stuff, I would certainly not recommend this route. When I play "power hammers" it needs to be handled accordingly for best outcomes for the driver survival. 🙂
While I have good record of not burning speakers down (zero total damages worthy of recone), it is always because of me having some sense and taking care of the proper signal management for the speaker. Othewise I believe the ratio of coil carnage would turn exactly opposite. No design is safe if it has impedance valleys in the passband. It always takes just "one DJ that has an idea to push certain steady bass note into "listeners" intestines for few minutes" that makes it impossible for any system to survive his rampage.
That is very true, that was also why a limiter on the dsp (password protected) is needed to make sure the amps don't get overloaded on the input by the dj/band. When i was working in pro audio rentals, that was mandatory to protect the system. Especially on cases where bass heavy music was played. And i had many arguments with dj's (being a dj playing bass heavy music also) about that limiter, but when you heared the limiter kicking in, you were playing to loud...
In your case, i would look at tapped horn designs or paraflex/transflex designs, not reflex cabinets. Something like the (free plan) Paraflex Cram Type O (2x12" drivers) could do it in a relative small enclosure.
Paraflex Cram Type O plans
and there are many other designs out there that do similar things. These will do the low subbass a lot better at high volume than any reflex. The theory is also availeble (on facebook groups i'm affraid), and these speakers are more and more used to replace the scoops in custom bass heavy dj soundsystems.
Paraflex Cram Type O plans
and there are many other designs out there that do similar things. These will do the low subbass a lot better at high volume than any reflex. The theory is also availeble (on facebook groups i'm affraid), and these speakers are more and more used to replace the scoops in custom bass heavy dj soundsystems.
In the same compact box volume, they wont. I had this discussion many times, and if the passband and versatility is to be the same, a ported design will offer most SPL output density from all.
To date, there was not a case of documented paraflex design around 200l-250 of outer box volume that could compete. The paraflex scales favourably well past 250l, rather 300l..
To date, there was not a case of documented paraflex design around 200l-250 of outer box volume that could compete. The paraflex scales favourably well past 250l, rather 300l..
In theory yes, but you will very easy overheat and burn your driver in a very compact reflex at high (max) volume. There is not enough cool air flow or space to push drivers to the max in a small cabinet. That has been proven.
And power compression will kick in very soon in a small cabinet because of that so the output will be reduced compared to the theory. With a tapped or tranflex/paraflex you don't have that, or far less at least.
And power compression will kick in very soon in a small cabinet because of that so the output will be reduced compared to the theory. With a tapped or tranflex/paraflex you don't have that, or far less at least.
Sigh...
We can probably only agree to disagree on this. If that was the case, I wouldn't be here making this topic, to not be laughed at/out.
My design exactly avoids this, and now you come with the argument of overheating and burning down.
There are cases, even here on this forum that we can look into. I think the long Keystone thread for example. The issue is, that they use the same speakers in fairly big boxes that run out of their design specs soon. Once you run drivers that don't do that, the ported design wins in small designs. Could easily run two 21"s with 6" coils in the volume of Keystone. That would kill anything the sub was thinking it was doing. Is it efficient? No. Is it effective? No. Does it produce most SPL per volume? Damn it is....
And that's all that is to the goal.
We can probably only agree to disagree on this. If that was the case, I wouldn't be here making this topic, to not be laughed at/out.
My design exactly avoids this, and now you come with the argument of overheating and burning down.
There are cases, even here on this forum that we can look into. I think the long Keystone thread for example. The issue is, that they use the same speakers in fairly big boxes that run out of their design specs soon. Once you run drivers that don't do that, the ported design wins in small designs. Could easily run two 21"s with 6" coils in the volume of Keystone. That would kill anything the sub was thinking it was doing. Is it efficient? No. Is it effective? No. Does it produce most SPL per volume? Damn it is....
And that's all that is to the goal.
If you insist on building reflexes, do it. But think about what i said about overheating and compression. The lavoce actually sims better (in WinISD at least) than that expensive 18sound i think, and would be a good candidate to try. That is true. But you will need a huge port for that to keep the air velocity low enough, so your cabinet will end up arround 250L in total anyway.
This is a quick sim of both in a 200L ported at 4kW with filters like you probally should use (8th order hpf at 20Hz and 4th order LPF at 100Hz). The thick line is the Lavoce.
This is a quick sim of both in a 200L ported at 4kW with filters like you probally should use (8th order hpf at 20Hz and 4th order LPF at 100Hz). The thick line is the Lavoce.
See.... It is not that I insist on that or on anything. It is that after numerous measurements, tests, discussions and $$$$ spent ofon speakers, I got to this point. Not full conclusion, I am open to things, but it seems to me like it.
I thought of overheating and compression more than quite a bit. That´s why this design. It fairly avoids port compression and driver thermal compression due to the use around its very high impedance peak (borderline able to feed any kind of power into it) and cutoff above port tuning. At this point (no need, just to clarify) it seems you did not get it or read it. There will be no issue with port air velocity because it is only used half way. The very peak being around 15m/s, nor power compression, because the driver will be well cooled and less power will be going in due to the impedance curve being relly high. Yes, bigger box can do more, but that was established, and it is outside the goal.
The WinISD is rather clueless with its native system input power, because it does not take into account quite few things. Same with efficiency.
Maybe I am due with the simulations to show something of that off..... Will do that as soon as possible.
I thought of overheating and compression more than quite a bit. That´s why this design. It fairly avoids port compression and driver thermal compression due to the use around its very high impedance peak (borderline able to feed any kind of power into it) and cutoff above port tuning. At this point (no need, just to clarify) it seems you did not get it or read it. There will be no issue with port air velocity because it is only used half way. The very peak being around 15m/s, nor power compression, because the driver will be well cooled and less power will be going in due to the impedance curve being relly high. Yes, bigger box can do more, but that was established, and it is outside the goal.
The WinISD is rather clueless with its native system input power, because it does not take into account quite few things. Same with efficiency.
Maybe I am due with the simulations to show something of that off..... Will do that as soon as possible.
@waxx
Here:
Look what wattage the WinISD proposes it to feed. Even without EQ, it is very unreliable piece of data. One cannot take this into account while developing any kind of serous box.
The classic design is 92l@39Hz ported box. The proposed unconventional is 29Hz tuning with sooner cutoff. The goal is compact, you simply cannot run from this and say 200l box is better. It is, DOH, but then two 92l boxes will walk over the single 200l box and outplay anything it was thinking it was doing....
Now, cone excursion:
There are ways and angle of views to look at it indeed.
The matter of fact is, that my design with such tuning is, that at high power frequencies, it moves much more, and cools down much better. You don´t want to feed 1200,1500,2000W into speaker that is barely moving. That will not work out.
Now you can counter with higher distortion. And you would be right. But it is still better than driver burned down, and I was specifically testing and buying speaker with long stroke and liner suspension, that will work admirably there. If we took for example the original B&C Speakers 18DS115 with Xvar of 14mm and suspension compliance halving at about 12.5mm, it would end up bad. Not with this one! But you cannot simulate that or read it from datasheet.
Now you might argue that the classic design could be fed more, because it has excursion to spare. Well, good luck with burning that thing down when the DJ hits that 40Hz note for 5 minutes straight.
Now, power! This is funny one!
↑ Not so for the classic design with 0°phase angle at 39Hz....
The WinISD shows apparent power. That is Volts versus Amps. Volt-amperes, not Watts. But that does not make all the heat. Volts * amps * cosinus phase angle - Watts do. Once you account for that, you will realize that the unconventional design eats LESS power in the most demanding areas where most cooling happens. Yes, my design eats more between 60-100Hz, but Make it half for ANY music with 3dB dynamic range or more, and we are talking 150-200VA load on the driver. Peanuts compared to 1000+ in the bass range. We can afford to basically discart that.
And so that´s that. Different view on things, and I believe I am not wrong here.
Now, You are right that bigger systems are better. They are. Try to make better one in THIS volume.
Here:
Look what wattage the WinISD proposes it to feed. Even without EQ, it is very unreliable piece of data. One cannot take this into account while developing any kind of serous box.
The classic design is 92l@39Hz ported box. The proposed unconventional is 29Hz tuning with sooner cutoff. The goal is compact, you simply cannot run from this and say 200l box is better. It is, DOH, but then two 92l boxes will walk over the single 200l box and outplay anything it was thinking it was doing....
Now, cone excursion:
There are ways and angle of views to look at it indeed.
The matter of fact is, that my design with such tuning is, that at high power frequencies, it moves much more, and cools down much better. You don´t want to feed 1200,1500,2000W into speaker that is barely moving. That will not work out.
Now you can counter with higher distortion. And you would be right. But it is still better than driver burned down, and I was specifically testing and buying speaker with long stroke and liner suspension, that will work admirably there. If we took for example the original B&C Speakers 18DS115 with Xvar of 14mm and suspension compliance halving at about 12.5mm, it would end up bad. Not with this one! But you cannot simulate that or read it from datasheet.
Now you might argue that the classic design could be fed more, because it has excursion to spare. Well, good luck with burning that thing down when the DJ hits that 40Hz note for 5 minutes straight.
Now, power! This is funny one!
↑ Not so for the classic design with 0°phase angle at 39Hz....
The WinISD shows apparent power. That is Volts versus Amps. Volt-amperes, not Watts. But that does not make all the heat. Volts * amps * cosinus phase angle - Watts do. Once you account for that, you will realize that the unconventional design eats LESS power in the most demanding areas where most cooling happens. Yes, my design eats more between 60-100Hz, but Make it half for ANY music with 3dB dynamic range or more, and we are talking 150-200VA load on the driver. Peanuts compared to 1000+ in the bass range. We can afford to basically discart that.
And so that´s that. Different view on things, and I believe I am not wrong here.
Now, You are right that bigger systems are better. They are. Try to make better one in THIS volume.
....And here is port velocities in roughly same port box volume:
With long rectangular port, all simulators are greatly overestimating. The box and the port is standing, and it is about 100cm long, properly tuned within 1Hz.
Now Let´s keep it the way the WinISD simulated. 14.3m/s versus 18m/s. Win for the unconventional design yet again.
Group delay, just in case, basically same within 35-500Hz.
The conventional design indeed has more maximum SPL. But in face of burning down, I have no use for it, or be bothered additionally by limiter pumping or fadeouts as it tries to save the speaker from inevitable. I´m sick of the strongly audible dynamic processing.
With long rectangular port, all simulators are greatly overestimating. The box and the port is standing, and it is about 100cm long, properly tuned within 1Hz.
Now Let´s keep it the way the WinISD simulated. 14.3m/s versus 18m/s. Win for the unconventional design yet again.
Group delay, just in case, basically same within 35-500Hz.
The conventional design indeed has more maximum SPL. But in face of burning down, I have no use for it, or be bothered additionally by limiter pumping or fadeouts as it tries to save the speaker from inevitable. I´m sick of the strongly audible dynamic processing.
When the "new generation" of higher power/excursion/Bl neo drivers appeared around 2010, EAW used a similar approach, they lowered the Fb to 30Hz, loosing some low end sensitivity in the SB1001, but increasing power density while still using a "flat to 40Hz" processed target:
I prefer your design's shallower depth to the SB1001, though it's offset cone arrangement also being part of the port exit increases the internal volume/port volume slightly by comparison.
In case there was any doubt of what I meant by orienting the port vertically to increase heat transfer from the bottom to top, I meant like this:
That's interesting (and odd), will be looking forward to your impressions with high power.
Art
I wonder if they had the same reasoning and what led them to basically discontinue this approach. Though these 130+ and 140dB figures are sweet. I'm going to market these figures on mine too hehe...
It appears that the test box is compact enough to fit two into the volume of the SB1001.
I understood what you mean regarding cooling, just didn't make the connection with the port dimensions-orientation. Yes, that's what I will do in order to secure better heat exchange with air.
I was thinking of design where the cone makes part of the port. I have even executed such thing before, and it does not work out favourably box volume wise.
Regarding the driver, I unfortunately was not able to replicate that behavior. Either it was some very weird Arta/Limp glitch, or it was part of the burn-in process, where the spider stiffened before it loosened again. I don't know. Now the drivers behave fairly as expected, although Fs is not going down much even after serious abuse.
It appears that the test box is compact enough to fit two into the volume of the SB1001.
I understood what you mean regarding cooling, just didn't make the connection with the port dimensions-orientation. Yes, that's what I will do in order to secure better heat exchange with air.
I was thinking of design where the cone makes part of the port. I have even executed such thing before, and it does not work out favourably box volume wise.
Regarding the driver, I unfortunately was not able to replicate that behavior. Either it was some very weird Arta/Limp glitch, or it was part of the burn-in process, where the spider stiffened before it loosened again. I don't know. Now the drivers behave fairly as expected, although Fs is not going down much even after serious abuse.
Funny, this morning while laying in bed, I realized the SB1001 offset cone arrangement also being part of the port exit does not increase the internal volume/port volume slightly by comparison, other than the small volume gained by not recessing the driver on the front panel.
If the cone was reversed within the port it could roughly equate to the L-Acoustics laminar L-Vent, while slightly increasing the internal cabinet by the air volume occupied by the driver, ~10.5 liters (0.37 ft3) for a driver like the B&C18SW115.
That said, the tuning would rise slightly on the inward stroke compared to the outward stroke, not sure what the those displacement changes in the port volume would do to distortion profiles.
Art
It is not very good idea for various reasons. Well you can make it an opening port enough till it is not a problem. That was how my first prototype was. More horn-optimized port shape was used. Reminds me from Beyma SB18 for example. And it works. But is too kicky and bitey, and boomy, not much fundamental bass. It would need bigger chamber AND bigger port, which would make it something very different. I tried that all, and settled at classic ported for good reasons.
Also tested the warmed up speaker in the box, and the impedance peak leans into the bottom range. So it looks like I can keep it tuned tiny teeny higher for less excursion and keeping that peak where it is needed.
By the way, I found a Basstronic music/track, where 1 period bass signals are played in rapid succession. In slo-mo it looks like the speaker only has 1/2 period overshoot, the rest is heavily attenuated/damped. Not bad.
Also tested the warmed up speaker in the box, and the impedance peak leans into the bottom range. So it looks like I can keep it tuned tiny teeny higher for less excursion and keeping that peak where it is needed.
By the way, I found a Basstronic music/track, where 1 period bass signals are played in rapid succession. In slo-mo it looks like the speaker only has 1/2 period overshoot, the rest is heavily attenuated/damped. Not bad.
I NEVER had a driver break in a bandpass enclosure. I had drivers die in direct radiator enclosures though.
Regarding BP4's, I always design the enclosure with the driver basket in the vented chamber for voicecoil cooling.
Quick model.
8ft TH (BP6S) = 26.25" high x 19.75" wide x 24.75" deep with 19.05mm (0.75in) wood.
The 18.62 m/s horn mouth peak particle velocity is k!ll!n' me tho'.
Avsforum.com guys are cool with <25 m/s with the Devastators (BP6P).
Diyaudio.com = <17 m/s.
Certainly, after measurements and final developments. But the original design posted is VERY close to the final. If ever, it might be 51x64x64cm, and very little change in the port profile. This is basically IT. But I will post the final design, no worries.
@BP1Fanatic :
I have hard time believing the TH will fit into that volume, but I might try to draw it. Yes, given the experience I am not cool with 25m/s either. There certainly will be high port losses that way, and that´s exactly what I am avoiding - almost all causes of power compression.
Again, for normal program content, many designs are better. Try that one tone bass note for 5 minut in such design that exactly fits into the impedance walley. There is no physical way for it to survive.
Such argument is hard to take on. I also never had a driver break on me, but that was because of careful management. I have seen drivers burnt in basically everything.
I have hard time believing the TH will fit into that volume, but I might try to draw it. Yes, given the experience I am not cool with 25m/s either. There certainly will be high port losses that way, and that´s exactly what I am avoiding - almost all causes of power compression.
Again, for normal program content, many designs are better. Try that one tone bass note for 5 minut in such design that exactly fits into the impedance walley. There is no physical way for it to survive.