I've only got as far as designing the enclosure for now, but the drivers I plan to use are the ND91-8 paired with the ND20FA-6, the box is a 0.1cubic feet after port, driver and crossover have been taken into account, I'm aiming for a tuning frequency of 60hz and these little speakers will be paired with one or more subs. I usually build from the standard MDF but for these I'm using 100% aluminium, 3mm thick sheet fully welded with a high gloss automotive finish, it may seem like an expensive finish for such cheap drivers but I'm using this design as an experiment, mainly to see how the enclosure affects or improves the sound quality. It may not work well which I'm aware of, in that case I'll switch back to using MDF.
So once the box is made I'll take the measurements of the drivers using the omnimic and then start on designing the crossover, which hopefully you guys can take a look at and tell me how good or bad I did, I have designed a few speakers over the years now but I still like to get constructive criticism from people who know better. But anyway, a few pics of the planned design,
So once the box is made I'll take the measurements of the drivers using the omnimic and then start on designing the crossover, which hopefully you guys can take a look at and tell me how good or bad I did, I have designed a few speakers over the years now but I still like to get constructive criticism from people who know better. But anyway, a few pics of the planned design,
@motokok I did, but I decided on a ported design in case I want to make pairs of these as gifts or to listen to them in stereo, would you have chosen to make them sealed instead? Theres also the size of the sealed box, I'm not sure I could fit crossover parts inside very well. As for the subs, I was planning to make those using a small tang band sub driver to help with integration but I'm open to suggestions on that.
The reason for such a small speaker is to fit into a small living room, preferably wall mounted, which probably would suit a sealed enclosure better but I think becuase I'm hoping to use them as a stand alone design too I wanted a little bit more output down low.
The reason for such a small speaker is to fit into a small living room, preferably wall mounted, which probably would suit a sealed enclosure better but I think becuase I'm hoping to use them as a stand alone design too I wanted a little bit more output down low.
Sealed is easier to model and build and would seem optimal for sub integration and near wall placement; you could just craft a faux rear hatch for the crossover, on the back of the needed volume.
Though if you want flexibility in case you want to use them on their own, probably go for the flattest tuning, see what that Vb would give for Qtc if closed; you can always stuff the port as desired.
Though if you want flexibility in case you want to use them on their own, probably go for the flattest tuning, see what that Vb would give for Qtc if closed; you can always stuff the port as desired.
Yeah I may look into how that would turn out, I had just decided on ported as you say for flexibility, I'd just be concerned the sealed one would sound thin without the sub, having never owned a sealed speaker I'm not sure how different it would sound to be honest, I've always been on the side of ported designs.
I don't think the ND91 will go low enough sealed to crossover to a sub. Parts Express page says sealed F3 of 125hz in 0.25 cu. ft. I wouldn't cross to a sub above 100hz.
Yeah I hadn't considered that, that's a good point, I'll have to stick to the original design
What are the thoughts on the curved sides of the enclosure? I'm hoping it will reduce any standing waves and also improve the look of the speaker, I had considered tilting the baffle too to align the drivers better but with it being aluminium it would take a lot more work to fabricate, don't think I'll do that unless it's worth the effort
What are the thoughts on the curved sides of the enclosure? I'm hoping it will reduce any standing waves and also improve the look of the speaker, I had considered tilting the baffle too to align the drivers better but with it being aluminium it would take a lot more work to fabricate, don't think I'll do that unless it's worth the effort
I love curved sides because I think it looks nice. I am skeptical that it makes any audible difference with regards to standing waves. (I know the theory but every DIY experiment I have seen suggests it doesn't really matter. I'm sure there are exceptions, so I don't mean to start a 200-post hijack about standing waves.)
I wouldn't bother with tilted baffle either. Maybe if it was easy, but you can address time alignment when designing the crossover.
I wouldn't bother with tilted baffle either. Maybe if it was easy, but you can address time alignment when designing the crossover.
Small for small, sure...but 3.5" woofer? That sounds like it would be challenged even with a sub. It really gets too big with a larger woofer?
- Aluminum welded, cool! Damping material coating inside?
- I think curved sides don't very the dimensions enough to make some tremendous difference but it looks WAY nicer!
Well I mainly designed it curved just for the aesthetics, and just thought it could only help with any standing waves, it should make the enclosure stronger regardless by having that shape, and yeah omnimic has a pretty easy method for finding the z offset of the drivers so I'll just use that method and do my best to correct it in xsim, I always feel it would be simpler to do if the drivers were more closely aligned though, but for the extra work I agree that it's just not worth it
What about the design makes you think it may be challenged to work as intended? Is it just the fact its a single woofer? I did originally model it as an MTM but the extra cost of possibly making 5 of these put me off a bit, the box wasn't much bigger either which had it's own issues of fitting two drivers on the baffle and fitting in a port, I considered since I'd be crossing between 80 and 100hz that they would probably be ok, and for stereo listening they'd be adequate for the volume levels they would be used at. I could have misjudged it though, that's why I'm building just one for now and seeing how it performs. Winisd shows an f3 of around 53hz I think, and due to the room having thin walls the speakers are unlikely to be turned up that loud.
I'm just trying to keep the box as small as possible which is why I went for these drivers. I also plan to use one of these as the centre channel, so in that situation maybe it won't work too well, and I may have to make an actual matching centre but I'll figure that out later I guess
Hi,
few quick comments 🙂
yeah size is directly related how loud/low you can go before sound quality goes poo. Low frequencies need volume displacement and with small cone it means high excursion. All non-linear phenomena (distortion) get worse the more there is excursion. Porting reduces excursion around tuning frequency but below it there is no damping on the driver from the box and excursion goes through the roof, use steep high pass filter.
Porting means steep acoustic high pass filter (+ electornic filter to keep excursion at bay) and thus high group delay (for the whole system). Usually porting is on the lowest "way" of a system only, so on a sub. Its easy to block though so perhaps not issue as such, perhaps good thing if no subs. Prepare to block that port nevertheless as it probably has some mid range resonances as well, its very long and inlet is at boundary which has all modes pressure nodes ready to leak out and excite pipe resonances.
Curved sides have very little effect on what happens inside the box (modes). It has some effect on what happens outside, diffraction. It would be good to round over front edges, as much as possible, immediately beside the drivers so that no flat baffle remains.
But, don't get depressed, prototyping is fun and best way to find out if it works just fine, or if some of the problems laid out are noticeable. Any box that makes sound is nice, can't live without music 🙂 Have fun!
few quick comments 🙂
yeah size is directly related how loud/low you can go before sound quality goes poo. Low frequencies need volume displacement and with small cone it means high excursion. All non-linear phenomena (distortion) get worse the more there is excursion. Porting reduces excursion around tuning frequency but below it there is no damping on the driver from the box and excursion goes through the roof, use steep high pass filter.
Porting means steep acoustic high pass filter (+ electornic filter to keep excursion at bay) and thus high group delay (for the whole system). Usually porting is on the lowest "way" of a system only, so on a sub. Its easy to block though so perhaps not issue as such, perhaps good thing if no subs. Prepare to block that port nevertheless as it probably has some mid range resonances as well, its very long and inlet is at boundary which has all modes pressure nodes ready to leak out and excite pipe resonances.
Curved sides have very little effect on what happens inside the box (modes). It has some effect on what happens outside, diffraction. It would be good to round over front edges, as much as possible, immediately beside the drivers so that no flat baffle remains.
But, don't get depressed, prototyping is fun and best way to find out if it works just fine, or if some of the problems laid out are noticeable. Any box that makes sound is nice, can't live without music 🙂 Have fun!
Sound pressure is created by air moving back and forth. The lower the frequency, the slower this motion, and the more air you have to move for the same pressure. In fact for a sealed box, if you go down an octave (half the frequency) the excursion has to increase 4X to maintain the sound pressure. And due to human hearing, as you go into bass frequencies you really need more sound pressure since the ear is insensitive (search "Fletcher Munson"). I gotta get to work and can't find the whole excursion formula, however here is a calculator
http://www.baudline.com/erik/bass/xmaxer.html
and if you're using WinISD I believe that has functions to show cone excursion? You can compare versus the Xmax specs for the driver. It depends very much on how loud you play-the peaks, not the average.
- Sound pressure drops off quickly the farther and farther you get from the speakers, so maybe you need more sound pressure than you think.
- Some folks think "oh I have a subwoofer, it does all the bass" but that's not true. The satellites still need to output midbass, which still requires total air motion. The smaller the cone, the more excursion needed to move the same amount of air. The more the cone is moving, the more distortion from magnetic nonlinearity, suspension nonlinearity, doppler distortion etc. Compared to a 3.5" off the rough top of my head a 4" might need 1/4 less excursion, a 5" maybe half.
- Ported boxes stick in a resonator, which at the tuning frequency produces sound while the woofer hardly moves. Nice! Except that depending on the tuning it may sound boomy, and below the port tuning is like a huge leak so the poor little woofer can flap around uncontrolled. Also the port takes up some space, and in small enclosures can be tricky to make both long enough and large enough to flow air without choking and making nasty sounds.
- A sealed box by contrast may have a higher F3 but rolls off more slowly. Also, there is no need to take up space with the port, and be stuffed to gain and apparent maybe 25% greater box volume, clawing back some of the advantage of the port without the disadvantage of the cone unloading below the port frequency.
- F3 is not so important. Rooms reinforce low frequencies so F6, f10 are really more relevant. Who says so? ME! And common sense. And Dr. Richard Small who when I met him on a business trip to Harman Indiana agreed with me. (He used F3 in his seminal thesis and papers because that was the convention for electric filters he was modeling speakers after, along with Thiele and also J.E. Benson. But electrical filters are not speakers in rooms). Now satellites are not getting reinforced by a room much; nevertheless I think F6 still more relevant.
I've never seen large PA speakers mounted on the wall of a small living room. And I've never seen 3 1/2" woofers used for an arena concert. I've never heard an 18" woofer replicate cymbals well, nor a 1" silk dome tweeter shake the house at 20hz. It almost seems like maybe there are different speakers suitable for different purposes. 🤔
@head_unit
Thank you for all that info, I appreciate it. My box and port were modelled with 24watts signal power (80% of the drivers rms) which keeps the cone excursion under the drivers 5mm xmax, obviously not below the tuning frequency though, and it has a port air velocity of 29, I'm aware that should be under 17 but I do plan to flare the ends of the port if I find any issues. I had heard before that it's best to add a filter to stop the driver trying to play too low like you mentioned above, I just don't know how to do that as I've never seen it added to a crossover on any speaker I've opened up, which made me assume that it has to be done another way with some form of external filter? If you could advice me on how to do it that would help me out a lot.
As for the bass on this speaker, I should have mentioned in the first post that I was never trying to get window shaking bass from this little driver, more of the opposite really, the house we are moving too has a really small living room, and also has thin walls, the bass that my current speakers have would drive the neighbours insane. Hence why I was ok with the performance of these drivers, but if what your saying is my design in general is bad then I have no problem redesigning as for now it's just on paper.
Thank you for all that info, I appreciate it. My box and port were modelled with 24watts signal power (80% of the drivers rms) which keeps the cone excursion under the drivers 5mm xmax, obviously not below the tuning frequency though, and it has a port air velocity of 29, I'm aware that should be under 17 but I do plan to flare the ends of the port if I find any issues. I had heard before that it's best to add a filter to stop the driver trying to play too low like you mentioned above, I just don't know how to do that as I've never seen it added to a crossover on any speaker I've opened up, which made me assume that it has to be done another way with some form of external filter? If you could advice me on how to do it that would help me out a lot.
As for the bass on this speaker, I should have mentioned in the first post that I was never trying to get window shaking bass from this little driver, more of the opposite really, the house we are moving too has a really small living room, and also has thin walls, the bass that my current speakers have would drive the neighbours insane. Hence why I was ok with the performance of these drivers, but if what your saying is my design in general is bad then I have no problem redesigning as for now it's just on paper.
What are your thoughts on this bluetooth speaker I made from copper a while ago? I've been trying to decide between using aluminium or copper as the copper sheet is more expensive but once aged and lacquered it has a more unique looking finish, imo. I'm not sure how different copper would sound, I might actually make one of each metal and see if the measurements show up any differences, just throwing ideas around at the minute since I just want to try something different from the norm, and try to create something that people can enjoy looking at as well as listening too. Obviously if I made it out of copper it would have the same finish on all sides rather than having an MDF baffle like this one does.
Attachments
I presume you're using an AVR? Those all have built-in bass management, you set the speaker to SMALL. Or something like Audyssey room correction will do it automatically. As for the design, if you've simulated it in WinISD and the excursion seems OK for the SPLs you want to generate, steam ahead! I'd want like a 4" woofer minimum but maybe my brain is warped by too many years in automotive and too many concerts ha ha. Are you going to stick in any cross-bracing (if there's even room...and while the aluminum is strong, damping? Don't know what would stick to that however).
After you build these, the next challenge will be to make a matching subwoofer...😵
You're replying to the guy who was cranking 117 dBC in my dorm room so I think my brain cannot correctly process your idea...🤣🤣🤣
What if we put an 18" woofer together with a 1" silk dome? Wouldn't that cover everything? 😉
Yes I have the sony 1080 reciever, that's how I use it now, speakers set to small crossed to the sub at 80hz, seems to work well, my question was more about when I use a stereo pair using a different type of amplifier with no built in filters. I've never really understood the best way to filter the lows from the speakers in that scenario.
I think that the enclosure with it's curved sides and welded all around, won't need any internal bracing, regardless of which metal I decide to use, as for the damping material, there's a product that sticks to the metal and stops it from vibrating and sounding like a biscuit tin, its a kind of sticky back rubber but I can't remember the name of it, I'll try to find it later.
In general, there is no point to high pass lows as there is rarely enough 🙂 except on a tuned system. The thing is, a bass reflex box port output is in phase with driver output around tuning frequency, hense we get bass boost around tuning frequency. Purpose of the port is to provide resonance to the system which then makes some extra boost to SPL output that the driver would output on its own, in a closed box for example. But, soon below the tuning frequency output of the port gets out of phase compared to driver output and they cancel each other out. Also, the box and port do not provide any acoustic loading to the driver below tuning frequency, the driver is able to move freely as there is no pressure constricting in the box/system. This means that the driver would happily flap around when you feed 20Hz to it, but there would be no audible output, only huge excursion. This excursion will cause distortion on all the driver bandwidth though! Basically, all bass reflex box would need high pass filter to keep excessive distortion out. On a two way box this affects performance of the whole mid range!
Compare to closed box, pressure inside the box is not leaked anywhere so it acts on the cone all the way down to few Hz and prevents high excursion. Also, there is no out of phase output to cancel sound. This means closed box can be run without high pass filter just fine, without distortion penalty, and even the 20Hz would have some "audible" output. Well, most likely not audible still if small woofer.
Best way to high pass is to no high pass! just make system that requires no high pass filter and doesn't sacrifice mid range performance and is still capable of good enough bass output 😉 This could mean closed box, or just big high sensitivity system with reflex box so that excursion is minimal even below system tuning. If you need to have small reflex speaker you could use any kind of high pass filter, like output transformer on a tube amplifier. Or line level filter somewhere inside the amp or before it. Amplifiers have some (high) input impedance so (small) capacitor in series with input would make simple line level high pass filter.
An old trick: use closed box and current output amplifier, driver resonance is not damped electrically because high series impedance in the circuit so you'd get some boost from that, no need for bass reflex box to do resonance and boost. With voltage output amplifier, that have very low output impedance by definition, one can increase impedance in the circuit by adding (big) series capacitor. This would also let the woofer resonance run free and make some bass boost.
Compare to closed box, pressure inside the box is not leaked anywhere so it acts on the cone all the way down to few Hz and prevents high excursion. Also, there is no out of phase output to cancel sound. This means closed box can be run without high pass filter just fine, without distortion penalty, and even the 20Hz would have some "audible" output. Well, most likely not audible still if small woofer.
Best way to high pass is to no high pass! just make system that requires no high pass filter and doesn't sacrifice mid range performance and is still capable of good enough bass output 😉 This could mean closed box, or just big high sensitivity system with reflex box so that excursion is minimal even below system tuning. If you need to have small reflex speaker you could use any kind of high pass filter, like output transformer on a tube amplifier. Or line level filter somewhere inside the amp or before it. Amplifiers have some (high) input impedance so (small) capacitor in series with input would make simple line level high pass filter.
An old trick: use closed box and current output amplifier, driver resonance is not damped electrically because high series impedance in the circuit so you'd get some boost from that, no need for bass reflex box to do resonance and boost. With voltage output amplifier, that have very low output impedance by definition, one can increase impedance in the circuit by adding (big) series capacitor. This would also let the woofer resonance run free and make some bass boost.
Last edited: