Thank you, Chris!
I'd not seen that UNF system before, but it certainly seems designed for the same application as my desktop project here. They look to have followed at least some similar paths to where my own thinking is taking me, too. Very interesting.
Yes, I'm certainly intending to go sealed for the bass. I was in two minds about reducing the cabinet size, since it is already fairly modest (as these things go) and having a natively sensible Qtc could make the unit more flexible in future. However, more compactness would definitely be preferable in what I'm aiming for here and now. So in reality, avoiding the requirement for EQ would likely be a poor decision; time to focus, I think..
I probably also agree on omitting the tweeter. I have tested this quite a lot by now, and whilst I do prefer the presence of a tweeter it is a very small difference that I'd possibly not notice unless listening back-to-back. Whilst having one is not without compromise plus extra challenges to achieve good integration; there is at least some chance of it being worse. So on balance I'll very likely just go with woofer assistance for the 7ms.
Thanks again for your thoughts,
Kev
I spent several hours yesterday reading threads about baffle-step, edge diffraction and box shape. Consequently I'm starting to establish some direction for the Alpair driver:
It is clear that at a <200hz crossover and modest sized baffles, the baffle step will mostly occur within the alpair 7MS's range, not below it or at the crossover point, so I wanted to consider that and cause the change to be a smooth transition. I'll counter it in software, but the less ragged the easier, especially as, in the near-field, there will be less support from room effects to off-set the step. From Olson's work (pictured previously) and subsequent studies, it seems that heavily chamfered shapes, or rounded ones like spheres or tear-drops, would be the best for this.
I also want good spatial imaging and in particular to avoid being able to localise the speakers themselves, even though they're pretty close up. Which seems to be at least partly related to edge diffraction, since this eminates from the speaker edges (unlike reflections, which are perceived as coming from elsewhere). Good radius or chamfers at the corners make the effect significantly more gentle and diffuse; It is unclear (to me) precisely what the mathematical relationship is between wavelength and radius/chamfer size, since smaller ones seem to perform slightly better than they should. However, people's results seem to suggest that whilst small ones do something worthwhile, 2" radius is becoming quite a useful size, and 4" radius is even better.
Both of which suggest that I'd like decent rounding or chamfering. Placing the driver asymmetrically in the baffle is (by itself) apparently less of an answer; it usefully reduces the peak effect of edge diffraction on axis because the distance to edges no longer sum at the same frequency, however the edges are still sharp so there is similar total amount of sudden diffraction going on, it is just spread around and can still be ragged off-axis.
So, as I'm now settled on using only one driver from <200hz upwards, it seems to me that for the external shape a sphere would be an easy and fairly comprehensive answer (as much as anything is). It is also a very 'friendly' shape, without corners, for something which is to be positioned quite close. I had wondered about incorporating a quarter wave tapering chamber for internal reflections, which might result in the (even better) tear-drop shape, but actually that makes it quite long. I like that a sphere can be rotated within its own volume for positioning etc, so I'll probably just use it 'as is', but break up the internal shape with baffles and put in as much fluffy stuff as required to reduce reflections. It doesn't preclude an attempt at aperiodic damping, though.
Were it instead a flat baffle, then two 4" round-overs either side of the baffle would make it about 13" wide. For a round version, it happens that Ikea bowls are about 11" wide and would fit on the desk, plus they'd make a sphere of about the right sealed volume for the 7MS. So that might be optimum, especially if I ever want to run these without woofer-assistance. Though.. for this WAW project I could go smaller for the sake of compactness. In a WAW, the 7MS won't be run down to its natural roll-off anyway so a low Qtc isn't important (within reason). Ikea also make an 8" bowl, though being smaller makes the transition between driver face and bowl a noticeably sharper corner (I'd guess about 45 degrees). Which is less good but probably still acceptable (edge diffraction isn't the hugest factor, after all). So either size bowl would probably do quite well, I'll just have to see what spheres of this size feel like when on the desk.
So, that has been helpful, and I'll be able to start thinking more about implementation next.
It is clear that at a <200hz crossover and modest sized baffles, the baffle step will mostly occur within the alpair 7MS's range, not below it or at the crossover point, so I wanted to consider that and cause the change to be a smooth transition. I'll counter it in software, but the less ragged the easier, especially as, in the near-field, there will be less support from room effects to off-set the step. From Olson's work (pictured previously) and subsequent studies, it seems that heavily chamfered shapes, or rounded ones like spheres or tear-drops, would be the best for this.
I also want good spatial imaging and in particular to avoid being able to localise the speakers themselves, even though they're pretty close up. Which seems to be at least partly related to edge diffraction, since this eminates from the speaker edges (unlike reflections, which are perceived as coming from elsewhere). Good radius or chamfers at the corners make the effect significantly more gentle and diffuse; It is unclear (to me) precisely what the mathematical relationship is between wavelength and radius/chamfer size, since smaller ones seem to perform slightly better than they should. However, people's results seem to suggest that whilst small ones do something worthwhile, 2" radius is becoming quite a useful size, and 4" radius is even better.
Both of which suggest that I'd like decent rounding or chamfering. Placing the driver asymmetrically in the baffle is (by itself) apparently less of an answer; it usefully reduces the peak effect of edge diffraction on axis because the distance to edges no longer sum at the same frequency, however the edges are still sharp so there is similar total amount of sudden diffraction going on, it is just spread around and can still be ragged off-axis.
So, as I'm now settled on using only one driver from <200hz upwards, it seems to me that for the external shape a sphere would be an easy and fairly comprehensive answer (as much as anything is). It is also a very 'friendly' shape, without corners, for something which is to be positioned quite close. I had wondered about incorporating a quarter wave tapering chamber for internal reflections, which might result in the (even better) tear-drop shape, but actually that makes it quite long. I like that a sphere can be rotated within its own volume for positioning etc, so I'll probably just use it 'as is', but break up the internal shape with baffles and put in as much fluffy stuff as required to reduce reflections. It doesn't preclude an attempt at aperiodic damping, though.
Were it instead a flat baffle, then two 4" round-overs either side of the baffle would make it about 13" wide. For a round version, it happens that Ikea bowls are about 11" wide and would fit on the desk, plus they'd make a sphere of about the right sealed volume for the 7MS. So that might be optimum, especially if I ever want to run these without woofer-assistance. Though.. for this WAW project I could go smaller for the sake of compactness. In a WAW, the 7MS won't be run down to its natural roll-off anyway so a low Qtc isn't important (within reason). Ikea also make an 8" bowl, though being smaller makes the transition between driver face and bowl a noticeably sharper corner (I'd guess about 45 degrees). Which is less good but probably still acceptable (edge diffraction isn't the hugest factor, after all). So either size bowl would probably do quite well, I'll just have to see what spheres of this size feel like when on the desk.
So, that has been helpful, and I'll be able to start thinking more about implementation next.
Not unlike the truncated pyramid, your sphere would need to be significantly large to smooth the response as pictured.….there’s no free lunch here and as such, very few commercial designs employ this methodology successfully.
Not sure what your goals are here…..nearfield desktop listening?…….not a high spl situation and on either side of a monitor/screen?……not much Spacial separation going one here with a pair of enclosures on a congested desktop.….something us engineers know and understand in a critical sense where performance equates with long term success.
Consider a balance of use, purpose and ergonomics along with acoustic design….prioritize. Listening in close proximity at over 90db brings on fatigue pretty quickly……1/2 hour or so and your ear/brain function starts it’s own attenuation…..natural selection preserves frequencies dependant for Spacial cues important for balance and situational awareness….everything else gets cut.
Wanna develop a TRUE sense of space In the nearfield and not just mid side separation? ( hard to do with anything recording post 2000)……dipole is the only way to get it done.
Not sure what your goals are here…..nearfield desktop listening?…….not a high spl situation and on either side of a monitor/screen?……not much Spacial separation going one here with a pair of enclosures on a congested desktop.….something us engineers know and understand in a critical sense where performance equates with long term success.
Consider a balance of use, purpose and ergonomics along with acoustic design….prioritize. Listening in close proximity at over 90db brings on fatigue pretty quickly……1/2 hour or so and your ear/brain function starts it’s own attenuation…..natural selection preserves frequencies dependant for Spacial cues important for balance and situational awareness….everything else gets cut.
Wanna develop a TRUE sense of space In the nearfield and not just mid side separation? ( hard to do with anything recording post 2000)……dipole is the only way to get it done.
Yes, the sphere (or round-overs or bevels) would need to be adequately large to be especially effective. Going by the tests people have done on rounding over and beveling, a 2" radius/chamfer either side of the driver is useful but 4" would be better.
Unfortunately the 7MS has quite a wide frame, so even though the cone is only about 3" diameter, the driver itself isn't much less than 5". Which in total would mean a fairly large sphere unless I rear mounted the driver, and that is getting more complicated. I might complicate it even more by using a baffle that continues the curve inward like a small waveguide. Aside from the inconvenient driver frame, there seems no reason to use a flat area of baffle between the driver and the round-overs or sphere.
The purpose is indeed low-SPL nearfield listening. I want headroom for clean peaks, but I'd not normally want anything above say 80dB for this system; in fact that'll be at less than a meter, which is why I can stretch the driver's ranges fairly easily, without massive distortion. Though I already listen near-field to some speakers and can't say that I've noticed fatigue even after 8-12hrs at the desk; so I suspect that might arise from things other than distance (like peaky frequency response, or maybe diffraction/reflections too close to the main signal for easy distinction).
I'd like the system to have decent spatial rendering, but in the sense of locating the position of (e.g.) instruments within a sound-stage. For me, dipoles strength is more in creating a much more general feeling of spaciousness; I like them for things like live recordings of orchestras and bands, but I'm much less keen for listening to studio recordings (which is most of my material). I'm also not keen on their requirements for positioning, with all the stuff that'll be in near proximity to them on the desk. So that is why I'm going with cabinets that are e.g. sealed or aperiodic rather than which emit significant energy directly out of the back. Obviously I can't prevent 4pi radiation at low frequencies, so will instead use that to advantage in flexible positioning of the woofers.
Unfortunately the 7MS has quite a wide frame, so even though the cone is only about 3" diameter, the driver itself isn't much less than 5". Which in total would mean a fairly large sphere unless I rear mounted the driver, and that is getting more complicated. I might complicate it even more by using a baffle that continues the curve inward like a small waveguide. Aside from the inconvenient driver frame, there seems no reason to use a flat area of baffle between the driver and the round-overs or sphere.
The purpose is indeed low-SPL nearfield listening. I want headroom for clean peaks, but I'd not normally want anything above say 80dB for this system; in fact that'll be at less than a meter, which is why I can stretch the driver's ranges fairly easily, without massive distortion. Though I already listen near-field to some speakers and can't say that I've noticed fatigue even after 8-12hrs at the desk; so I suspect that might arise from things other than distance (like peaky frequency response, or maybe diffraction/reflections too close to the main signal for easy distinction).
I'd like the system to have decent spatial rendering, but in the sense of locating the position of (e.g.) instruments within a sound-stage. For me, dipoles strength is more in creating a much more general feeling of spaciousness; I like them for things like live recordings of orchestras and bands, but I'm much less keen for listening to studio recordings (which is most of my material). I'm also not keen on their requirements for positioning, with all the stuff that'll be in near proximity to them on the desk. So that is why I'm going with cabinets that are e.g. sealed or aperiodic rather than which emit significant energy directly out of the back. Obviously I can't prevent 4pi radiation at low frequencies, so will instead use that to advantage in flexible positioning of the woofers.
Thanks for clarifying……now for the inconvenient truth…..the 7ms and it’s 9db peak and trough between 8 and 11khz will be like an icepick to the brain in the nearfield as will nearly any wide range driver who‘s cone breakup is in the passband…..and that’s just about every one of ‘em.…..wide band drivers are just not good nearfield solutions and in the professional world only exist for the sole purpose of critical analysis short term. Such designs like the Aurasound Cubes and similar are not pleasing to listen to but they do help by highlighting resonance and locking in mid side cues…..something you DO NOT want to listen to for fun.
So my suggestion to you in this application would be the other way round on a two way…..wide band driver(s) that can cover 60hz all the way to 7khz with a very smooth response and a small smooth soft dome from 7khz on up. When considering drivers, look at the spectral decay very closely……those resonances in the nearfield are time based and as such in the nearfield are extremely important. Many of today’s ‘hi fi’ drivers that are ‘critically damped’ are loaded with such time related resonance but since they‘re used mid or far field can’t be heard so audiophiles clamour away at the smooth response…..but in your use case, time and phase domain issues predominate. Take the iconic Yamaha NS 10 for example……an amazing tool for critical monitoring because of its fantastic time domain and group delay performance…..quite by accident adopted for those traits by mix engineers around the world and through decades…..but hence out past 1m and into free space they sound awful with little to no endearing traits.
I don’t post here much these days but when I do see a thread focused on nearfield listening and design, I pop my head out because it’s what I do as a career and the properties of a good nearfield source are NOTHING like those of a conventional stereo speaker system. I’d look at the design criteria of commercial high end nearfield monitors for my inspiration.
So my suggestion to you in this application would be the other way round on a two way…..wide band driver(s) that can cover 60hz all the way to 7khz with a very smooth response and a small smooth soft dome from 7khz on up. When considering drivers, look at the spectral decay very closely……those resonances in the nearfield are time based and as such in the nearfield are extremely important. Many of today’s ‘hi fi’ drivers that are ‘critically damped’ are loaded with such time related resonance but since they‘re used mid or far field can’t be heard so audiophiles clamour away at the smooth response…..but in your use case, time and phase domain issues predominate. Take the iconic Yamaha NS 10 for example……an amazing tool for critical monitoring because of its fantastic time domain and group delay performance…..quite by accident adopted for those traits by mix engineers around the world and through decades…..but hence out past 1m and into free space they sound awful with little to no endearing traits.
I don’t post here much these days but when I do see a thread focused on nearfield listening and design, I pop my head out because it’s what I do as a career and the properties of a good nearfield source are NOTHING like those of a conventional stereo speaker system. I’d look at the design criteria of commercial high end nearfield monitors for my inspiration.
Thank you for the thoughts, this is clearly something that I'll need to take quite seriously and definitely investigate further.
Using a tweeter is okay with me; I've no issue with that - 'provided' it is crossed in pretty high; I'd say 7khz is about right but 9khz might be even better, if the driver below it were capable. It would take more effort to integrate seamlessly, but I'm not averse to trying. The lack of a tweeter does bring compromises that I'm only considering for simplicity (mostly permitted due to my hearing age). Even with the clever cone-break-up of the 7MS, the treble isn't quite as good.
But perhaps the key reason for a tweeter might be to permit a bigger, smoother wide/main range driver, and with consequently more dynamics lower down. Though 60hz overall isn't going to be low enough for me, so there would still be a subwoofer or woofer involved too, meaning 'some kind' of 3-way. In which case perhaps the main driver might be spared from having to produce the lowest frequencies, too - not necessarily down to 60hz. I'd be interested in opinions on good drivers (that aren't just available in the US market)
Though stepping back, I'm not necessarily wanting the same things that pro nearfield speakers are designed for. Forgiving and enjoyable are probably more the thing for my application, since critical listening results in dissatisfaction with a very large percentage of pre-recorded music. I have no idea how to phrase that in terms of driver parameters though.
Using a tweeter is okay with me; I've no issue with that - 'provided' it is crossed in pretty high; I'd say 7khz is about right but 9khz might be even better, if the driver below it were capable. It would take more effort to integrate seamlessly, but I'm not averse to trying. The lack of a tweeter does bring compromises that I'm only considering for simplicity (mostly permitted due to my hearing age). Even with the clever cone-break-up of the 7MS, the treble isn't quite as good.
But perhaps the key reason for a tweeter might be to permit a bigger, smoother wide/main range driver, and with consequently more dynamics lower down. Though 60hz overall isn't going to be low enough for me, so there would still be a subwoofer or woofer involved too, meaning 'some kind' of 3-way. In which case perhaps the main driver might be spared from having to produce the lowest frequencies, too - not necessarily down to 60hz. I'd be interested in opinions on good drivers (that aren't just available in the US market)
Though stepping back, I'm not necessarily wanting the same things that pro nearfield speakers are designed for. Forgiving and enjoyable are probably more the thing for my application, since critical listening results in dissatisfaction with a very large percentage of pre-recorded music. I have no idea how to phrase that in terms of driver parameters though.
Last edited:
Anyway, to continue thinking about the WAW possibility, a couple of drawings with the two sizes of Ikea bowl:
(the use of a pinkish colour for the spheres might have been unfortunate!)
So really there would be space for either size. The larger one would have a suitable Q for use 'as is', the smaller one would need to be crossed over to a woofer (or EQ'd) due to a somewhat higher peaky Q, BUT is usefully less physically intrusive.
However the bigger one will clearly give a wider, smoother baffle step and WRT edge diffraction its angle with the driver is also noticeably shallower. Probably a similar difference to having one 45-degree bevel or making the same thing from two bevels (each at half the angle), which people report as being similar but distinguishable:
So I still think either would work. Clearly the bigger one would be 'better' from an audio perspective, I just need to try it to see if I can live with such a thing around the desk, or if a smaller one would be noticeably preferable in a usability sense.
(the use of a pinkish colour for the spheres might have been unfortunate!)
So really there would be space for either size. The larger one would have a suitable Q for use 'as is', the smaller one would need to be crossed over to a woofer (or EQ'd) due to a somewhat higher peaky Q, BUT is usefully less physically intrusive.
However the bigger one will clearly give a wider, smoother baffle step and WRT edge diffraction its angle with the driver is also noticeably shallower. Probably a similar difference to having one 45-degree bevel or making the same thing from two bevels (each at half the angle), which people report as being similar but distinguishable:
So I still think either would work. Clearly the bigger one would be 'better' from an audio perspective, I just need to try it to see if I can live with such a thing around the desk, or if a smaller one would be noticeably preferable in a usability sense.
As an aside, both sizes of bowl are also small enough in depth for having the woofers directly behind (and somewhat below) them; i.e. facing forward rather than to the sides. Not a biggie, but might be more sensible given that a single (rather than dual opposing) woofer per side is being used.
Sorry Kev…..NOTHING in the nearfield (1 meter or less) is going to be forgivable……EVERYTHING is going to be audible. But that’s your only real disadvantage here. You don’t have to worry about high spl/power compression issues…..you can use a moderately expensive dome tweeter down to 1khz or an AMT tweeter down to 800hz.….a Scanspeak D3004 has an Fs of 440hz and it’s tiny….would reduce the size of the sphere enclosure needed substantially as well. It’s also really efficient so feed it 3 watts and you’re at 92 db.…..it can easily handle that at 1khz with a moderately steep XO. Float those balls/bowls over an upfiring 5” woofer…..off axis response to 90 degrees of any decent quality driver will be flat and smooth with any cone breakup lost to being off axis. And again, you won’t need any chest pounding woofer here…..you’re going to employ a sub behind your monitor anyways so in the nearfield, all the woofer needs to do is play to 100hz or so cleanly. Heck, there’s 4” woofers that can do that no problem…….the Dayton RS100 does that in a small sealed chamber.
As for the sub, use your situational install to your advantage……a single driver facing the wall behind the desk would be best, capturing the boundary gain of the wall to support the lowest frequencies possible. Add a piece of foam to the wall to eat up any driver mechanics noise and lingering cone breakup……at 180 degrees off axis, you won’t hear any of that much anyways.
So even though the tweeters are pricey, you’re still in the sub $500 range for a co plate sub/sat 3 way system that would sound incredible and likely last a lifetime in this application. Use a thin mount sub driver like the Dayton or Tang Band 6” low profile subs and your preserve desk space……you can even mount your monitor to the sub with a VESA mount and eliminate the bulky stand…..the weight and gravity pull of the monitor will also keep the sub box from creeping…..or you could singly screw the box to the desktop…..choices. Those little ScanSpeak tweeter mounted in bowls floating with say a heat formed piece of 1/4 thick acrylic would look pretty amazing.….polish the edges and add an LED to the base for a nice ambient glow?….or an upfiring led from the woofers onto the bottom of the bowls?………lots of design possibilities here.
Thank you once again. Now this sounds more like something I could get into. My firm intention for room speakers (a future project) comprise a compression driver down to below a khz, with woofer(s) below. It is a format I'm very keen on, but for some reason I'd not thought about how it could be reduced in scale, using some types of tweeters (rather than a huge compression waveguide). But of course it could, with such modest demands on SPL and power. Thank you for pointing it out!
More research and thinking required on my part, clearly.
I use a pair of Markaudio CHR-70 full range speaks in sealed boxes for my office. I think they sound great driven by my Pass ACA Mini amp.
Really clean and coherent with outstanding imaging. No crossover or other filtering are needed. They are good as is.
F3 is around 79 hz in a 4.5 liter box yielding a Q of .87. Just about right for a small sealed box. You might want to add a sub but I like them standalone against the wall on a shelf behind my monitor. Added sealed benefit: bass roll off is a gentle 12dB/octave rolloff so they sound fuller on most music than they have a right to sound. This is why I prefer sealed speakers. The bass always sounds fuller than expected and is always tigher and better controlled than a great many ported speakers.
Really clean and coherent with outstanding imaging. No crossover or other filtering are needed. They are good as is.
F3 is around 79 hz in a 4.5 liter box yielding a Q of .87. Just about right for a small sealed box. You might want to add a sub but I like them standalone against the wall on a shelf behind my monitor. Added sealed benefit: bass roll off is a gentle 12dB/octave rolloff so they sound fuller on most music than they have a right to sound. This is why I prefer sealed speakers. The bass always sounds fuller than expected and is always tigher and better controlled than a great many ported speakers.
Last edited by a moderator:
Just been looking around at tweeters, there are quite a few that easily go down to under 1khz at the SPLs I'm wanting (sometimes even at 1watt on their manufacturer's FR charts). So probably it is just a matter of comparing their distortion and waterfall charts where these exist. Unfortunately the better ones tend not to be cheap, though a couple from SB seem potential at reasonably good value.
Of course crossing at several hundred to a thousand hertz might take this project out of WAW category again. But it wouldn't preclude the use of tiny wide-range cone drivers instead of tweeters. Just that of those I've looked at (in the few cases I can find sufficient measurements) their off-axis response and distortion seem easily beaten by good tweeters until you get right down towards 1khz or so.
So I will pursue the wide-range tweeter idea further. In a sense I have one answer in the bag, the 7MS crossed to woofers below 200hz; whatever their technical down-sides, I have been listening to this combo for quite a while in test cabinets (with EQ to smooth the response and peaks) and it is probably the best I've so far heard for my application. Better than my bookshelves (that I already liked) which cross a 5" woofer above 6khz to a tweeter. But I won't know if using tweeters (essentially as the wide range) might be even better, without more research and testing.
Many people would instantly say no, the tweeters will sound strained. But then, around 85dB at less than 1m is not how most people would use their tweeters, either. There are examples of people successfully crossing normal/room speakers at not much over 1khz, so going lower but quieter shouldn't be impossible; certainly some have the displacement needed to achieve it - within their nominal linear range - and also a low enough Fs to be well below a crossover point.
Of course crossing at several hundred to a thousand hertz might take this project out of WAW category again. But it wouldn't preclude the use of tiny wide-range cone drivers instead of tweeters. Just that of those I've looked at (in the few cases I can find sufficient measurements) their off-axis response and distortion seem easily beaten by good tweeters until you get right down towards 1khz or so.
So I will pursue the wide-range tweeter idea further. In a sense I have one answer in the bag, the 7MS crossed to woofers below 200hz; whatever their technical down-sides, I have been listening to this combo for quite a while in test cabinets (with EQ to smooth the response and peaks) and it is probably the best I've so far heard for my application. Better than my bookshelves (that I already liked) which cross a 5" woofer above 6khz to a tweeter. But I won't know if using tweeters (essentially as the wide range) might be even better, without more research and testing.
Many people would instantly say no, the tweeters will sound strained. But then, around 85dB at less than 1m is not how most people would use their tweeters, either. There are examples of people successfully crossing normal/room speakers at not much over 1khz, so going lower but quieter shouldn't be impossible; certainly some have the displacement needed to achieve it - within their nominal linear range - and also a low enough Fs to be well below a crossover point.
Last edited:
So if you’re enjoying the presentation of the 7ms, I STRONGLY suggest you DO NOT cross it over to a tweeter anywhere near the 1khz range. Crossovers introduce phase anomalies and are only used because they are necessary for configurations where drivers do not work within their passband…….the 7ms is designed and does work VERY WELL from 200 hz all the way up to 8khz in a near field, low power application such as yours.
The ‘problem’ is at and above 8khz where the response of the 7ms is Cone breakup which manifests itself in a wild and bumpy frequency response filled with resonance and stored energy. BUT can you/do you hear it in your listening?….if not, then leave well enough alone and there you have it……a fine two way woofer/wide band two way with no crossover anomalies to smear the most detailed parts of the passband.
I suggest you continue as is with critical listening over time so you can identify what is missing, lacking or fatiguing for the 7ms first. Then I would do some nearfield measurements so your observations can translate to what you are hearing.
Drivers like the 7ms are great at what they do…..smooth, enjoyable representation in a wide passband……problem is when folks try to extend that wide range passband for 20-20 use. If your goal is a fine detailed, smooth 3 way, you’ve chosen the wrong driver for your midrange with the 7ms and it’s back to the drawing board………..but it doesn’t have to be if you’re enjoying what you’re hearing and can live with the compromise
The ‘problem’ is at and above 8khz where the response of the 7ms is Cone breakup which manifests itself in a wild and bumpy frequency response filled with resonance and stored energy. BUT can you/do you hear it in your listening?….if not, then leave well enough alone and there you have it……a fine two way woofer/wide band two way with no crossover anomalies to smear the most detailed parts of the passband.
I suggest you continue as is with critical listening over time so you can identify what is missing, lacking or fatiguing for the 7ms first. Then I would do some nearfield measurements so your observations can translate to what you are hearing.
Drivers like the 7ms are great at what they do…..smooth, enjoyable representation in a wide passband……problem is when folks try to extend that wide range passband for 20-20 use. If your goal is a fine detailed, smooth 3 way, you’ve chosen the wrong driver for your midrange with the 7ms and it’s back to the drawing board………..but it doesn’t have to be if you’re enjoying what you’re hearing and can live with the compromise
I wouldn't envisage crossing the 7MS over at anything like 1khz - the main reason for considering it over smaller drivers, with better top ends (like the 5.3MS), is that it would more easily go low enough to be 'assisted' by low/sub woofers.
Crossing near 1khz was instead related to your suggestion of using a (capable) tweeter as the wide-range. I'd be happier to cross a couple or ideally few hundred hertz less though, partly for the reasons you suggest. So I've started a few trials with some compression drivers. These are effectively still a 1" driver, but with considerably bigger diaphragms; they will be able to cross lower than even very costly dome tweeters.
If they don't turn out well, then I know the 7MS crossed below 200hz will be adequate. Although as tests go on, I'm thinking that the 5.3MS crossed (yes, quite low) to a small midbass would be a superior solution - unless I 'particularly' want a single-driver (plus sub) speaker.
Crossing near 1khz was instead related to your suggestion of using a (capable) tweeter as the wide-range. I'd be happier to cross a couple or ideally few hundred hertz less though, partly for the reasons you suggest. So I've started a few trials with some compression drivers. These are effectively still a 1" driver, but with considerably bigger diaphragms; they will be able to cross lower than even very costly dome tweeters.
If they don't turn out well, then I know the 7MS crossed below 200hz will be adequate. Although as tests go on, I'm thinking that the 5.3MS crossed (yes, quite low) to a small midbass would be a superior solution - unless I 'particularly' want a single-driver (plus sub) speaker.
While compression drivers can cross low and have little to no audible distortion, their response can be choppy and readily audible in the nearfield…..I don’t advise this approach for your application.
You mentioned a commitment to the performance of the 7ms earlier…..so my assumption is/was your plan to continue with that driver as the core of your design…….and that’s not a bad thing IF the 7ms brings a smile to your face. Even though the cone breakup above 8khz is present and real, listening just 10-15 degrees off axis can greatly mitigate those anomalies…..and in truth, everything is a compromise
As for crossing a high quality dome to a midwoofer at 1khz a compromise compared to the wide range 7ms?……not at all…..in his application which I assume remains near field?……..the cohesive and detailed output of a Scanspeak dome on a suitable baffle with 1-2 watts of quality source material and amplification would be staggering.
You mentioned a commitment to the performance of the 7ms earlier…..so my assumption is/was your plan to continue with that driver as the core of your design…….and that’s not a bad thing IF the 7ms brings a smile to your face. Even though the cone breakup above 8khz is present and real, listening just 10-15 degrees off axis can greatly mitigate those anomalies…..and in truth, everything is a compromise
As for crossing a high quality dome to a midwoofer at 1khz a compromise compared to the wide range 7ms?……not at all…..in his application which I assume remains near field?……..the cohesive and detailed output of a Scanspeak dome on a suitable baffle with 1-2 watts of quality source material and amplification would be staggering.
I do like the 7MS; it has some nice audio qualities and is a reassuringly strong contender for the final build. But I'm still only auditioning and testing options (that appeal to me) at the moment.
Unfortunately both with full/wide-range and compression drivers there are informed people on this forum that love them and others that hate them for home/hifi applications. So it seems that trying them for oneself, in the intended application, is the only way to decide really.
Even more-so for using tweeters very low; not many people do it and not many people use such close and low-SPL speakers so the general advice on here (that they get strained and distorted, even if not exceeding excursion) doesn't really apply. Unfortunately that is quite a costly thing to try though, as suitable tweeters are expensive. It is also very possible that (at my hearing age) I'd like the humble 5.3MS about as much (or maybe more, with its added benefit of being able to cross lower), as for me, the 5.3MS is better than the (already acceptable) 7MS at the top end. Many options, all with their own compromises.
Unfortunately both with full/wide-range and compression drivers there are informed people on this forum that love them and others that hate them for home/hifi applications. So it seems that trying them for oneself, in the intended application, is the only way to decide really.
Even more-so for using tweeters very low; not many people do it and not many people use such close and low-SPL speakers so the general advice on here (that they get strained and distorted, even if not exceeding excursion) doesn't really apply. Unfortunately that is quite a costly thing to try though, as suitable tweeters are expensive. It is also very possible that (at my hearing age) I'd like the humble 5.3MS about as much (or maybe more, with its added benefit of being able to cross lower), as for me, the 5.3MS is better than the (already acceptable) 7MS at the top end. Many options, all with their own compromises.
Last edited:
Just did my first speaker using the ikea bowls. Not really happy with the finish and I made a lot of mistakes. About to build the second one but this time I want to use a router to sink the driver. Im using a full range driver Dayton Audio 'PS95-8 3-1/2" Point Source Full-Range Driver'.
Attachments
Looks quite successful for a first attempt. It is the nature of what we hobbyists do, very often the things we build are also their own first prototype. So it is quite common to finish and then wish one had made a few different choices. I think it is partly why people endlessly tweak things, go on to later make an improved MKII, or even change their mind and make a quite different version.
It was a good idea to make one speaker before the other, so that you can make changes and improvements on the second one. Personally I would not change or replace the first one until I had listened to them as a pair for quite a while. To see if they sound right and if their character suits you, before making any decisions about what to do next.
It was a good idea to make one speaker before the other, so that you can make changes and improvements on the second one. Personally I would not change or replace the first one until I had listened to them as a pair for quite a while. To see if they sound right and if their character suits you, before making any decisions about what to do next.