Here I put my skin on the market.
Project is in the development stage.
The objective of this project is to build a SPL dense compact solution for low crest factor music, I.E.: heavy bass.
All that reggae, bassy chill, full on PSY, techno, tekkno and other heavy stuff, in which long bass line happen to play for long loong minutes.
How do I do that?
Well, speakers are not built for this mostly at all, so some compromises must be taken into account. No free lunch, kids!
Basically, anything seriously compact in PA can only be ported or bandpass. Not much way around it, if same (good) bandwidth is to be achieved.
The issue of ported designs is, that they have significant cone excursion minimum in "most used" part of the bass band, and that reduces speaker coil cooling. On top of that, exactly at that frequency, most power goes into the speaker coil. So if full power goes to the speaker at that point on freqency scale, it is basically a toast within short minutes, sometimes sooner.
How come it wasn´t problem till now?
It was. Either the speaker burned, or heavily compressed, or strong limiting kicked in, kicking the output down by 6 or even 10db in some cases of even high end commercial products.
For "normal" content, it is not such big problem and this limiting kicked in rather gently and for short period of time. For heavy music, it would hinder the long term performance quite a bit.
What we gain: safety, low power compression and SPL compression, full bass, SPL density for box size, just 3dB limiter safety setting instead of 9-10dB. Low frequency mode giving fair SPL down to 25Hz.
What we lose: Maximum possible/peak SPL reached by about 3dB lower than conventional design. Also little of the Price/performance ratio.
How we do that properly?
In order to lower the thermal load on the speaker in compact design, it needs to operate in such a mode and way, that the impedance minimum and cone excursion minimum is avoided. In exact words, put the port tuning frequency outside the crossover pass band (lower). That will solve the issue. But of course, it will create another one. Otherwise this would not be novel approach at all. In this scenario, the speaker cone is forced to do most of the work most of the time. That means, it has to move a lot.
Most PA speakers "to date" are not built for that, so most speakers except the well chosen high end simply cannot do that. Those speakers will run out of linear excursion capabilities.
Objective in detail:
Box volume: Max ~210l / 7.4cu.ft brutto, ~92l / 3,25cu.ft net depending on the speaker.
Weight: ~34kg / 75lbs. Easy single person operation.
Frequency range: 40-80Hz flat, 35-110Hz -6dB.
SPL: 121dB/W/m+ according to Hornresp in halfspace.
This design could be made smaller, but A) for serious compromise B) If planned upgradability is dropped. More on that later.
Low power compression, low port compression.
In this case, tuning frequency will be 29Hz.
Basic sketch:
It is not my first rodeo with such design, and with RCF LF21N551 and barely with B&C 21DS115 it worked. Yet I sadly sold these, so it´s time to move little forward.
The wood used for outer walls is 15mm/0,59in baltich birch, speaker mount wall 18mm/0,71in the port, braces and back wall possibly 15mm/0,6in poplar birch.
The port is flow and "horn" optimized to squeeze little bit more performance out of the design than regular. Certain corners will be ground and smoothed some more. Thinking even about printing the port front-side into something more resembling of L-port. 🙂
Driver needs:
As mentioned, high displacement volume. We are talking here 20-21mm of one way stroke to fully exploit the design. Either IPAL class drivers or something like that in future. Actually, I am kind of counting on that the technology will catch up with the design. We are almost there. Also 5,3" to 6" coil for more "oomph" in 35-40Hz region would help. But we are doing it now, so we use what we have...
In order to fairly utilize the box, we need over 1,8l of displacement (one way) and optimum would be 2.5l. Meaning, 3dB more to reach all box limits as port flow, distortion figures for speaker stroke in such volume, doppler distortion limit almost.... That´s where this enclosure options end. And that is 21mm of excursion for 18".
The new RCF LN19S450 would qualify, but not in the current box dimensions. This is just a prototype, some changes to allow for 19" driver could be made.
I have already ordered that, but the story is unfortunate - Released in March 2024, odrered in December 2024 delivery NOT confirmed by may 2025. I just can´t. You know what I would have to say about RCF...
Drivers that qualify:
Generally these with displacement volume of 1,8l and more, and Bl²Re over 200:
That is, for example 18Sound 18NTLW5000 4Ohm version
LaVoce SAN184.50
All the IPALish drivers.
Drivers that do not qualify for feasibility reasons, but would work:
All similar drivers like B&C 18DS100/18DS115.
18SW115/4, RCF LF18X451.
Given the displacement limits, it would be just too expensive and not SPL dense anymore, or efficiency would be lower (more heating).
Why I gave the LaVoce pass: Remains to be tested, but Lavoce bass drivers allow for very large linear stroke. The SAN184.50 is marked as 15,5mm Xmax. If it is anything like 21" series, then it WILL work admirably till 16mm on regular basis. That gives it 1,95l of displacement volume. B&C on the other hand, has Xvar of 14mm, which is a bit of a stretch in my book, because suspension compliance dropping below 50% is at about 12,5mm That gives us 1,51l of displacement volume at 30% price hike. And so as much as I love the LOVELY 18DS115 it is rather out of the list. Not feasible. Would work, but at 1,5-2dB loss or more audible distortion. I am certain it WOULD stretch to these displacements, but possibly at too high distortion figure. Sad story so to speak.
Well, what if you pulled it all out of your a$$?
That´s where the reality comes in. Sims incoming, Prototype build will be done possibly by the end of this week, measurements next.
I also plan to make YT video about me rambling on audio topics and SAN184.50 review.
Suggestions, questions, opposition, all is welcome!
Project is in the development stage.
The objective of this project is to build a SPL dense compact solution for low crest factor music, I.E.: heavy bass.
All that reggae, bassy chill, full on PSY, techno, tekkno and other heavy stuff, in which long bass line happen to play for long loong minutes.
How do I do that?
Well, speakers are not built for this mostly at all, so some compromises must be taken into account. No free lunch, kids!
Basically, anything seriously compact in PA can only be ported or bandpass. Not much way around it, if same (good) bandwidth is to be achieved.
The issue of ported designs is, that they have significant cone excursion minimum in "most used" part of the bass band, and that reduces speaker coil cooling. On top of that, exactly at that frequency, most power goes into the speaker coil. So if full power goes to the speaker at that point on freqency scale, it is basically a toast within short minutes, sometimes sooner.
How come it wasn´t problem till now?
It was. Either the speaker burned, or heavily compressed, or strong limiting kicked in, kicking the output down by 6 or even 10db in some cases of even high end commercial products.
For "normal" content, it is not such big problem and this limiting kicked in rather gently and for short period of time. For heavy music, it would hinder the long term performance quite a bit.
What we gain: safety, low power compression and SPL compression, full bass, SPL density for box size, just 3dB limiter safety setting instead of 9-10dB. Low frequency mode giving fair SPL down to 25Hz.
What we lose: Maximum possible/peak SPL reached by about 3dB lower than conventional design. Also little of the Price/performance ratio.
How we do that properly?
In order to lower the thermal load on the speaker in compact design, it needs to operate in such a mode and way, that the impedance minimum and cone excursion minimum is avoided. In exact words, put the port tuning frequency outside the crossover pass band (lower). That will solve the issue. But of course, it will create another one. Otherwise this would not be novel approach at all. In this scenario, the speaker cone is forced to do most of the work most of the time. That means, it has to move a lot.
Most PA speakers "to date" are not built for that, so most speakers except the well chosen high end simply cannot do that. Those speakers will run out of linear excursion capabilities.
Objective in detail:
Box volume: Max ~210l / 7.4cu.ft brutto, ~92l / 3,25cu.ft net depending on the speaker.
Weight: ~34kg / 75lbs. Easy single person operation.
Frequency range: 40-80Hz flat, 35-110Hz -6dB.
SPL: 121dB/W/m+ according to Hornresp in halfspace.
This design could be made smaller, but A) for serious compromise B) If planned upgradability is dropped. More on that later.
Low power compression, low port compression.
In this case, tuning frequency will be 29Hz.
Basic sketch:
It is not my first rodeo with such design, and with RCF LF21N551 and barely with B&C 21DS115 it worked. Yet I sadly sold these, so it´s time to move little forward.
The wood used for outer walls is 15mm/0,59in baltich birch, speaker mount wall 18mm/0,71in the port, braces and back wall possibly 15mm/0,6in poplar birch.
The port is flow and "horn" optimized to squeeze little bit more performance out of the design than regular. Certain corners will be ground and smoothed some more. Thinking even about printing the port front-side into something more resembling of L-port. 🙂
Driver needs:
As mentioned, high displacement volume. We are talking here 20-21mm of one way stroke to fully exploit the design. Either IPAL class drivers or something like that in future. Actually, I am kind of counting on that the technology will catch up with the design. We are almost there. Also 5,3" to 6" coil for more "oomph" in 35-40Hz region would help. But we are doing it now, so we use what we have...
In order to fairly utilize the box, we need over 1,8l of displacement (one way) and optimum would be 2.5l. Meaning, 3dB more to reach all box limits as port flow, distortion figures for speaker stroke in such volume, doppler distortion limit almost.... That´s where this enclosure options end. And that is 21mm of excursion for 18".
The new RCF LN19S450 would qualify, but not in the current box dimensions. This is just a prototype, some changes to allow for 19" driver could be made.
I have already ordered that, but the story is unfortunate - Released in March 2024, odrered in December 2024 delivery NOT confirmed by may 2025. I just can´t. You know what I would have to say about RCF...
Drivers that qualify:
Generally these with displacement volume of 1,8l and more, and Bl²Re over 200:
That is, for example 18Sound 18NTLW5000 4Ohm version
LaVoce SAN184.50
All the IPALish drivers.
Drivers that do not qualify for feasibility reasons, but would work:
All similar drivers like B&C 18DS100/18DS115.
18SW115/4, RCF LF18X451.
Given the displacement limits, it would be just too expensive and not SPL dense anymore, or efficiency would be lower (more heating).
Why I gave the LaVoce pass: Remains to be tested, but Lavoce bass drivers allow for very large linear stroke. The SAN184.50 is marked as 15,5mm Xmax. If it is anything like 21" series, then it WILL work admirably till 16mm on regular basis. That gives it 1,95l of displacement volume. B&C on the other hand, has Xvar of 14mm, which is a bit of a stretch in my book, because suspension compliance dropping below 50% is at about 12,5mm That gives us 1,51l of displacement volume at 30% price hike. And so as much as I love the LOVELY 18DS115 it is rather out of the list. Not feasible. Would work, but at 1,5-2dB loss or more audible distortion. I am certain it WOULD stretch to these displacements, but possibly at too high distortion figure. Sad story so to speak.
Well, what if you pulled it all out of your a$$?
That´s where the reality comes in. Sims incoming, Prototype build will be done possibly by the end of this week, measurements next.
I also plan to make YT video about me rambling on audio topics and SAN184.50 review.
Suggestions, questions, opposition, all is welcome!
Last edited:
With today's high Xmax drivers, it seems like it might could match/best my long gone dual kit Danley Contrabass if they were fan/whatever cooled.
https://ficaraudio.com/neodymium/4-series/411/
The Fi Car Audio NEO series looks like it could kill for output density.
The 60 pound weight is over what you want, but 34mm Xmax is way out of the "pro audio" limits.
Yes, there are existing speakers already, but no free lunch. 2x-3x higher price, weight, hard to feed in PA. You certainly could do that at a cost.
If I knew how to effectively feed 2Ohm speakers, I'd get the IPALish versions of LaVoce and be done with it. Feeding four 8Ohm or 4Ohm drivers can be as cheap as ~$1000. Two ohms, tough luck, much more than double. So overall for the system, you'd pay more than double for about 2.5dB gain in loudness. Doesn't make sense to me. But good find and good point!
The LaVoces should arrive today. Will let them sit at room temp for a moment, and then test the excursion behavior in hope they are like the 21"s.
If I knew how to effectively feed 2Ohm speakers, I'd get the IPALish versions of LaVoce and be done with it. Feeding four 8Ohm or 4Ohm drivers can be as cheap as ~$1000. Two ohms, tough luck, much more than double. So overall for the system, you'd pay more than double for about 2.5dB gain in loudness. Doesn't make sense to me. But good find and good point!
The LaVoces should arrive today. Will let them sit at room temp for a moment, and then test the excursion behavior in hope they are like the 21"s.
How much power @ 2R? Erica.C offers some 2R capable modules
I grabbed one of this smaller 2K2R unit from Taramps for my mobile system
https://www.ebay.com.au/itm/354493911283
This is advanced problematics. First, I need to feed em from 230V mains. Then, the driver has such high impedance rise that in my case it is basically and virtually 4 Ohm driver. BUT. It has Re of 1.3Ohms, so it still needs at least 2Ohm stable amplifier. I need just 1500W, BUT in order to match the voltages that would feed the driver this power, such amps are classified as ~2500W.
One technical detail: to take advantage of the "chimney effect" (heat rises) it would be better to orient the port vertically to get rid of heat.
The only difference in construction would be the feet and socket location.
What you're essentially saying is the enclosure and driver needs to mostly be oprrating in some form of resonance over most its bandwidth. Running it this way keeps the amp from seeing the nominally lower driver impedance and therfore avoids heating the VC by keeping VC curent low at high drive voltages. That pretty much means sloppy, boomy, resonant bass from lack of cone control.
The right way to do it (which promotes better SQ) would be to raise the sensitivity of driver and enclosure via more efficient enclosure design. Hoffmans iron law works against this, so a small enclosure won't net much sensitivity if bandwidth and sensitivity is to be higher.
Why hasn't this supposedly been an issue before? It's the music itself placing the demand on the speakers. It just requires alot of power to generate high SPLs. There's no free lunch. Using resonance to achieve high SPL reliably for long periods of time is the only option. It will end up sounding awful and it has to do so, mainly because you need current to tightly control the cone motion. That can't be done with the enclosure running in resonance mode, as the amp can't command the driver properly without some corrective force. That needs current at a lower impedance, which means power and VC heating.
In order to have decent reliability at higher SPL and a small enclosure, more controlled excursion can help, but thats tricky to keep an eye on. It calls for specific enclosure tuning, which can work ok, but only if the driver is carefully dialed in and VCtemp is constantly being monitored. That has to be done through complicated DSP modeling. Feed forward correction doesn't really solve the issue, as it just pushes the driver harder if its EQed to go lower than the physical enclosure cutoff. This is just like the driver impedance being reduced, causing further VC heating.
Back in the day we used LF FLHs to generate high SPLs, but those require huge, heavy enclosures. You had to keep a close eye on the excursion and heating the VC from lack of airflow inside a relatively small rear chamber. Tapped horns came along and improved this, but made SQ deteriorate from the multiple resonances of the enclosure design itself. Some tapped horns sound pretty good, but not great. IMO its easier to use many smaller subs and DSP correction. The multi band limiting / compression, dynamic EQ, power derating and arraying of multiple smaller ported cabs will be the most effective way to generate high SPLs for a long time. It depends on the venue and engineer as to how well these tools can be utilized in order to keep SQ relatively decent.
The right way to do it (which promotes better SQ) would be to raise the sensitivity of driver and enclosure via more efficient enclosure design. Hoffmans iron law works against this, so a small enclosure won't net much sensitivity if bandwidth and sensitivity is to be higher.
Why hasn't this supposedly been an issue before? It's the music itself placing the demand on the speakers. It just requires alot of power to generate high SPLs. There's no free lunch. Using resonance to achieve high SPL reliably for long periods of time is the only option. It will end up sounding awful and it has to do so, mainly because you need current to tightly control the cone motion. That can't be done with the enclosure running in resonance mode, as the amp can't command the driver properly without some corrective force. That needs current at a lower impedance, which means power and VC heating.
In order to have decent reliability at higher SPL and a small enclosure, more controlled excursion can help, but thats tricky to keep an eye on. It calls for specific enclosure tuning, which can work ok, but only if the driver is carefully dialed in and VCtemp is constantly being monitored. That has to be done through complicated DSP modeling. Feed forward correction doesn't really solve the issue, as it just pushes the driver harder if its EQed to go lower than the physical enclosure cutoff. This is just like the driver impedance being reduced, causing further VC heating.
Back in the day we used LF FLHs to generate high SPLs, but those require huge, heavy enclosures. You had to keep a close eye on the excursion and heating the VC from lack of airflow inside a relatively small rear chamber. Tapped horns came along and improved this, but made SQ deteriorate from the multiple resonances of the enclosure design itself. Some tapped horns sound pretty good, but not great. IMO its easier to use many smaller subs and DSP correction. The multi band limiting / compression, dynamic EQ, power derating and arraying of multiple smaller ported cabs will be the most effective way to generate high SPLs for a long time. It depends on the venue and engineer as to how well these tools can be utilized in order to keep SQ relatively decent.
Last edited:
Indeed. I was thinking about that too!
You mean horizontally? Anyways that´s what will be done in a way. Pure horizontal orientation would mean that the handle would protrude into the small port, and that is rather no-go. Although with some help it would work too. What I think about as a sensible option, is to just put feet on both sides. Therefore it can stand and it can lay, in order to be lower AND get the heat out somewhat better. I will contemplate on details of that, but it is given, than better than basic design will be taking this into account. Many thanks for suggestion. I might miss something here and there for sure.
@profiguy
.
But I would love to get to the bottom of this.
The amp is holding the position of the speaker too, not letting it swing freely. Small box volume is not letting it swing freely, PA design with managed excursion and DC offset management in mind is not letting it swing freely, and the San184.50 behaves somewhat against it too - during burn-in test, The Fs moved up, Impedance lowered. That thing stiffened when moved a lot.
Before, I had combined net box volume of about 250l, two speaker motors to excite it and 4800Wpeak to feed it. Now, there is four motors, 360l of net volume, much more port area, 8000Wpeak to feed it. basically, I am doing everything possible at once, to not have to drive it that hard.
The ports are working, otherwise They would not deserve to be that big. They are adding to the SPL between 30-40Hz still, and even with the management, there will be 15m/s flow in these. It is just that these will be used fairly lightly.
But, I would love to know more arguments or more exact suggestions about how and where to look at it, what to measure and compare, etc.
This angle of view is interesting and somewhat welcome for further investigation. Bevause it has merits. Well, any design has this impedance peak, This usually is not an objection though if someone tunes the port around lower frequency, but uses it there too. I mean this claim probably has no way to be taken on in any usable way yet. Needs more refining. It would mean other systems would be sloppy as well. By the way, the 21DS115 was used this way, and the sound was lovely and hard. The (MS-sound) guy who heard it with both 21DS115s and RCF LF21N551s, develops and sells gear, took the enclosures from me and purchased new 21DS115s for himself in order to have this sound again. It is not all just crapolla. Knowing how the RCF double 21"s sound (very sloppy), I can just argue it is unfounded claim in this young form
.
But I would love to get to the bottom of this.
The amp is holding the position of the speaker too, not letting it swing freely. Small box volume is not letting it swing freely, PA design with managed excursion and DC offset management in mind is not letting it swing freely, and the San184.50 behaves somewhat against it too - during burn-in test, The Fs moved up, Impedance lowered. That thing stiffened when moved a lot.
Sure. Absolutely. Not an option though. We are sacrificing sensitivity, half way efficiency and almost everything for SPL density and low crest factor signal handling. I did not say it is universally better solution. Well aware of that. The distortion rises because volume displacement in smaller enclosure causes more pressure difference in the box, thus greater nonlinearity, and the avalanche of bad parameters could keep going forever. The point is, that no system in such size will play this loud and hard. One could get 15" for such volume, that would feel good in that volume, but it would simply not play better.
In a way, I am doing that too. This rig will basically be just for myself, and some gigs where things don´t matter. In home circumstances, I will welcome the usability down to 25Hz with this low tuning.
Before, I had combined net box volume of about 250l, two speaker motors to excite it and 4800Wpeak to feed it. Now, there is four motors, 360l of net volume, much more port area, 8000Wpeak to feed it. basically, I am doing everything possible at once, to not have to drive it that hard.
The ports are working, otherwise They would not deserve to be that big. They are adding to the SPL between 30-40Hz still, and even with the management, there will be 15m/s flow in these. It is just that these will be used fairly lightly.
But, I would love to know more arguments or more exact suggestions about how and where to look at it, what to measure and compare, etc.
Test box is up.
Wet though... One day to become mountable and measurable, another day to cure/harden.
Whew, the woofer fit in within last milimeter. don´t know where I lost 5mm in my box measurements.
Not sure where to gain possible 25mm in height in order to allow fitting the 19" RCFs. It´ll probably go the other way. Stealing dear milimeters everywhere to stay most compact with 18"s.
Wet though... One day to become mountable and measurable, another day to cure/harden.
Whew, the woofer fit in within last milimeter. don´t know where I lost 5mm in my box measurements.
Not sure where to gain possible 25mm in height in order to allow fitting the 19" RCFs. It´ll probably go the other way. Stealing dear milimeters everywhere to stay most compact with 18"s.
Couldn’t the heat problem instead be solved through either
- a pipe from the magnet to the outside
- attaching heat pipes to the magnet and a heatsink embedded in the enclosure wall
- a fan?
Yes, some designs are like that. But as there is fair venting in the box installed, it does not bother me to put the box on the side when playing hard, at all.
No need for additional tools, hardware and potential issues. Imagine somethig falls off or stops working. Then the speaker is cooked. No need for such risks when the solution is easy, working, not problematic etc....
Well, I could not wait to at least mount the beast. With average to low padding, the box tuning is 29.5Hz. I guess I missed the design by ~1Hz, because I thought it will be under 29Hz with padding. Hornresp is way off... That´s okay, some milimeters will be shaved off of the port area in final design anyways.
The minimum impedance is 12.5R, and with speaker heated up I guess it will be above 13Ohm. If the amp protection circuity will not have problem with speakers Re, I guess I could run two of these on one 8Ohm channel of the amp that is not supposed to run at 4Ohm. Better so if the amp is 4Ohm stable of course. I hesitate to try that on my slightly upgraded T.Amp TA2400 MK-X. But wont build second development box just for that...
No need for additional tools, hardware and potential issues. Imagine somethig falls off or stops working. Then the speaker is cooked. No need for such risks when the solution is easy, working, not problematic etc....
Well, I could not wait to at least mount the beast. With average to low padding, the box tuning is 29.5Hz. I guess I missed the design by ~1Hz, because I thought it will be under 29Hz with padding. Hornresp is way off... That´s okay, some milimeters will be shaved off of the port area in final design anyways.
The minimum impedance is 12.5R, and with speaker heated up I guess it will be above 13Ohm. If the amp protection circuity will not have problem with speakers Re, I guess I could run two of these on one 8Ohm channel of the amp that is not supposed to run at 4Ohm. Better so if the amp is 4Ohm stable of course. I hesitate to try that on my slightly upgraded T.Amp TA2400 MK-X. But wont build second development box just for that...
Hi, this remark doesn't make sense, because voltage amplifier has no control on the cone ever, becaise it's not currebt amp but voltage amp where driver (circuit impedance) determinens current through voice coil. The cone itself dictates current, the cone controls itself through electrical damping, which results the high impedance peaks,at the resonance where force turns directly into cone velocity as spring defeats mass. Basically the cone itself has great control, electrical brake on, when impedance is high. Impedance dips at port tuning frequency because there is no velocity on the cone, which makes the electrical damping, but it's not needed either because the cone doesn't move.
Your observations about sloppy sound are likely correct, but the explanation just doesn't make sense, to me at least.
Perhaps it's about junp resonance or solething, where non-linear surround and reducing Bl with high excursion makes operating point shift and the response is all over the place. Obe would get it back in line so to speak, by reducing excursion, eg. choosib
ng box tuning. Good modern driver should work with higher excursion before these problems start. Nothing to do with amplifier controlling cone though, but more plausibly about non-linearities of the driver I think.
Last edited:
@tmuikku You're likely looking at the premise of voltage and current from a different POV than what I'm trying to explain from an electromechanical perspective.
Yes, at exact ported box tuning, you're running at the lowest impedance point the tuned system is capable of achieving, depending on box loss, port loss, driver TSPs, etc. That is the only point in a ported system where the most current is being demanded from the amp, which also means excessive VC heating at this specific frequency. Its a small portion of bandwidth compared to the remaining portion of the enclosure's operating bandwidth, in which average impedance is much higher than nominal driver impedance. In a ported system such as this, the resulting output actually sounds best at its tuning frequency due to low cone excursion and high port velocity, but only if the port isn't restrictive and the minimum impedance point is closer to VC DCR. Because the cone has excellent control over the box resonant mode at this point (ie. tightly coupled with enclosure and port air mass), the start and stop points are also well controlled and the output will tightly follow the input. This is mainly a mechanically induced resonance mode, not an electrical resonance. It depends on the air load (resistance) the cone sees and translates into a lower electrical impedance.
In a box assisted with filters and other resonant circuits (LCR, Zobel, HP, LP, bandpass), the behavior around resonance can be changed or affected electrically. The components in the circuit with the enclosure dictate the final impedance curve (aside from the driver / box tuning itself, which is mechanical).
You can only raise, not lower Qts (and therefore Qtc) with electrical components. The driver's dampening characteristics around Fs / Fc can't be changed electrically. You can change the impedance curve (and resulting amp load) with filter components, but it isn't going to change the driver TSPs. Even if you mimic an impedance hump with electrical components, it still leaves the actual driver TSPs being unchanged. That means the driver / enclosure won't sound any better just because you can change the impedance curve externally using electrical filter components.
The amp will experience a higher current demand the lower the impedance dips at a given frequency and vice versa. This isn't about current or voltage amplification, neither is it a matter of comparing amplifier topologies or construction. Its simply a matter of impedance vs load and the resulting VC heating the driver will experience at higher drive levels.
A box with multiple resonances, such as tapped horns, will be even more efficient and thermally easy on drivers. Its due to the bulk of the enclosure's impedance curve being significantly higher than nominal driver impedance. The multiple resonance modes of the enclosure type will lower the average amplifier current demand and as a result VC temps. The amp can't control the cone motion as well when the enclosure is running in a resonance mode at specific frequencies. Current is required to control the VC at these points and the amplifier can only deliver a set amount of voltage, translating into (a lack of) current. It does mean you can run more enclosures of like design on one amp and also enjoy the array gain of multiple, tightly placed enclosures.
The simple reason why sealed enclosures sound tighter and cleaner (within limits of cone excursion) is due to most of the enclosure's impedance curve being lower, therfore allowing more current to be delivered into the VC, using brute force to control the cone motion. The (mechanical) air load on the cone is high across a wide frequency range. It doesn't allow for VC cooling unless the driver is inversely mounted. If you want more efficiency at lower frequencies with smaller enclosures, it requires resonance to achieve. That is what generates the most output per watt and is the cheapest way. The bass quality will suffer this way. There's no way around it. As long as you have excursion capability, the output will be high at these resonant modes, but not so clean or well controlled compared to sealed, FLH or strategically tuned ported.
Please feel free to elaborate if I'm missing something.
Yes, at exact ported box tuning, you're running at the lowest impedance point the tuned system is capable of achieving, depending on box loss, port loss, driver TSPs, etc. That is the only point in a ported system where the most current is being demanded from the amp, which also means excessive VC heating at this specific frequency. Its a small portion of bandwidth compared to the remaining portion of the enclosure's operating bandwidth, in which average impedance is much higher than nominal driver impedance. In a ported system such as this, the resulting output actually sounds best at its tuning frequency due to low cone excursion and high port velocity, but only if the port isn't restrictive and the minimum impedance point is closer to VC DCR. Because the cone has excellent control over the box resonant mode at this point (ie. tightly coupled with enclosure and port air mass), the start and stop points are also well controlled and the output will tightly follow the input. This is mainly a mechanically induced resonance mode, not an electrical resonance. It depends on the air load (resistance) the cone sees and translates into a lower electrical impedance.
In a box assisted with filters and other resonant circuits (LCR, Zobel, HP, LP, bandpass), the behavior around resonance can be changed or affected electrically. The components in the circuit with the enclosure dictate the final impedance curve (aside from the driver / box tuning itself, which is mechanical).
You can only raise, not lower Qts (and therefore Qtc) with electrical components. The driver's dampening characteristics around Fs / Fc can't be changed electrically. You can change the impedance curve (and resulting amp load) with filter components, but it isn't going to change the driver TSPs. Even if you mimic an impedance hump with electrical components, it still leaves the actual driver TSPs being unchanged. That means the driver / enclosure won't sound any better just because you can change the impedance curve externally using electrical filter components.
The amp will experience a higher current demand the lower the impedance dips at a given frequency and vice versa. This isn't about current or voltage amplification, neither is it a matter of comparing amplifier topologies or construction. Its simply a matter of impedance vs load and the resulting VC heating the driver will experience at higher drive levels.
A box with multiple resonances, such as tapped horns, will be even more efficient and thermally easy on drivers. Its due to the bulk of the enclosure's impedance curve being significantly higher than nominal driver impedance. The multiple resonance modes of the enclosure type will lower the average amplifier current demand and as a result VC temps. The amp can't control the cone motion as well when the enclosure is running in a resonance mode at specific frequencies. Current is required to control the VC at these points and the amplifier can only deliver a set amount of voltage, translating into (a lack of) current. It does mean you can run more enclosures of like design on one amp and also enjoy the array gain of multiple, tightly placed enclosures.
The simple reason why sealed enclosures sound tighter and cleaner (within limits of cone excursion) is due to most of the enclosure's impedance curve being lower, therfore allowing more current to be delivered into the VC, using brute force to control the cone motion. The (mechanical) air load on the cone is high across a wide frequency range. It doesn't allow for VC cooling unless the driver is inversely mounted. If you want more efficiency at lower frequencies with smaller enclosures, it requires resonance to achieve. That is what generates the most output per watt and is the cheapest way. The bass quality will suffer this way. There's no way around it. As long as you have excursion capability, the output will be high at these resonant modes, but not so clean or well controlled compared to sealed, FLH or strategically tuned ported.
Please feel free to elaborate if I'm missing something.
Good points. I believe it is not that straight forward. If you excite a port, it has one strong and about two decaying periods before it stops. And the cone can do nothing about it, because it is basically stopped anyways. Only saving grace is that noone knows how the bass is supposed to sound, so it sounds fairly natural.
For the cone resonance, on top of what I mentioned, I believe the amp has some hold of the driver even at resonance frequency. I wonder how do I get to the explanation of that, but I think so. It is DCR bound.
When I knock on the speaker right in the currently build design, the sound is "DOH, DOH, DOH". When it wired to the powered amplifier, the sound is TOK TOK TOK. Clearly braking whatever resonance it has. Does it debunk my first paragraph? I don´t know, I think it is speaker issue here, not port issue.
For the cone resonance, on top of what I mentioned, I believe the amp has some hold of the driver even at resonance frequency. I wonder how do I get to the explanation of that, but I think so. It is DCR bound.
When I knock on the speaker right in the currently build design, the sound is "DOH, DOH, DOH". When it wired to the powered amplifier, the sound is TOK TOK TOK. Clearly braking whatever resonance it has. Does it debunk my first paragraph? I don´t know, I think it is speaker issue here, not port issue.
^ Yeah electrical damping, the driver backEMF makes current which opposes the movement at the drivers resonance. Above resonance it doesn't oppose due to mass being accelerated delaying it 90deg. You hear the resonance when you knock and the amp is not connected, because electrical damping is missing. You can make electrical damping just by shorting the driver cables, so without an amp, and repeat the knock test with same results as with amp connected.
@profiguy thanks for the follow up! I see your perspective, it's fine although some of the explanation is backwards 🙂
@profiguy thanks for the follow up! I see your perspective, it's fine although some of the explanation is backwards 🙂
Hmm, I'm not at simulator right now, but i think you could exchange port output SPL by reducing efficiency of the port<>driver resonance by adding damping material inside the box. This would increase cone movement and reduce current at the tuning frequency helping with the thermal management some. Also helps dampen some driver resonance above the port tuning to protect from over excursion a bit. This seems silly at first but perhaps is useful in your application. Anyone have vituixcad open?🙂
Hypothesis is the system would transition toward kinda closed box response with more damping added inside the box, but with some overall cooling ability as it's bass reflex still. The expense is reduced efficiency at least lowest lows, but these you highpass anyway which is why I think this might be useful if you need reduce heating further. Looking at impedance plot the two peaks would reduce and low point between get bit higher. Or perhaps it's more heat for same output? 🤔I' ll try to simulate tomorrow, I must get some sleep now. Good night everyone!
Hypothesis is the system would transition toward kinda closed box response with more damping added inside the box, but with some overall cooling ability as it's bass reflex still. The expense is reduced efficiency at least lowest lows, but these you highpass anyway which is why I think this might be useful if you need reduce heating further. Looking at impedance plot the two peaks would reduce and low point between get bit higher. Or perhaps it's more heat for same output? 🤔I' ll try to simulate tomorrow, I must get some sleep now. Good night everyone!
Last edited:
Cool it's exactly what happens with impedance plot, although it's not useful as power is increased at the drivers resonance. High pass filter would already cut the power from the tuning frequency. Anyway, fun stuff 🙂
A woofer I had open in the simulator tuned to very low in relatively small box, power minimum around 50hz and max at ~22Hz
2nd order Highpass filter at ~30Hz cuts the power in half with only ~1db output reduction at 40Hz, assuming pass band is something like 40Hz up.
This happens if empty box is adjusted to "heavy stuffing" dropping Qa from 120 to 7. Efficiency drops so output drops. Power draw at the tuning frequency drops eyeballing about 25% while power draw increases about 50% on the driver resonance, so perhaps it's a silly trade-off even for this application 😀
Here with more conventional tuning, eyeballing 3db more output at 40Hz, with not much more power draw. Had to adjust HP bit to keep excursion within xmax below tuning.
But making this experiment I think I get the twist, Crashpc box is tuned for very specific genre in a way, to be able to exchange the ~3db drop in max output that is lost due to very low tuning, to prevent the >6db power compression and destroyed drivers, making it worthwhile. Otherwise more conventional tuning would give mucho more output.
A woofer I had open in the simulator tuned to very low in relatively small box, power minimum around 50hz and max at ~22Hz
2nd order Highpass filter at ~30Hz cuts the power in half with only ~1db output reduction at 40Hz, assuming pass band is something like 40Hz up.
This happens if empty box is adjusted to "heavy stuffing" dropping Qa from 120 to 7. Efficiency drops so output drops. Power draw at the tuning frequency drops eyeballing about 25% while power draw increases about 50% on the driver resonance, so perhaps it's a silly trade-off even for this application 😀
Here with more conventional tuning, eyeballing 3db more output at 40Hz, with not much more power draw. Had to adjust HP bit to keep excursion within xmax below tuning.
But making this experiment I think I get the twist, Crashpc box is tuned for very specific genre in a way, to be able to exchange the ~3db drop in max output that is lost due to very low tuning, to prevent the >6db power compression and destroyed drivers, making it worthwhile. Otherwise more conventional tuning would give mucho more output.
There is a reason why scoops are still popular for those genres. the backloaded horn with a compression chambre helps a lot to have the strong bass without smoking the amp and driver when done right. The only boxes i know that can do similar are tapped horns and tranflex cabinets. Both use the driver with the magnet out to cool the magnet more.