Yes, that might work quite well. Personally I'm more inclined towards doing equivalent things with active dsp and eq; partly because it might be easier to correct other slightly ragged areas of response too. But if the project goes another way and needs to stand alone, that could be a very good call. Thanks!
On another note, after more listening I've decided that I actually prefer the 7MS (once tamed a bit, with eq) over the 5.3MS. The latter does have a slightly better top end but at my hearing age (mid 50s) there isn't a massive difference up top, whilst the 7MS has noticeably better dynamics through the midrange. I might think differently (or more likely want a tweeter) if I had the ears of an 18-year old. Even now a tweeter is still an option for wider dispersion, as I'll be sitting quite close, though I don't feel the need when sitting bang on the axis in the small sweet spot.
Though oddly it still needs a reasonably loud SPL (at 1m) for the 7MS to get going and sounding as special as it can; played quietly it sounds fairly average to me. I suppose that is to be expected to some degree, but I've had better quiet-listening results with small 2-way bookshelves. It seems to go a bit beyond what correcting for equal loudness curves can do, too. I suspect this may be down to the small cone area when tested at full-range, so I shall be very interested to see how things might improve with woofer-assistance.
I shall also be interested to compare a pair of faitalpro drivers of comparable cone size, that I have coming (4FE35). These are very different: deeper cone, not a metal one, less linear excursion but more sensitive. They are also suited to rear-mounting, and have a less wide frame, which might be useful for some forms of baffle-edge diffraction management.
Though oddly it still needs a reasonably loud SPL (at 1m) for the 7MS to get going and sounding as special as it can; played quietly it sounds fairly average to me. I suppose that is to be expected to some degree, but I've had better quiet-listening results with small 2-way bookshelves. It seems to go a bit beyond what correcting for equal loudness curves can do, too. I suspect this may be down to the small cone area when tested at full-range, so I shall be very interested to see how things might improve with woofer-assistance.
I shall also be interested to compare a pair of faitalpro drivers of comparable cone size, that I have coming (4FE35). These are very different: deeper cone, not a metal one, less linear excursion but more sensitive. They are also suited to rear-mounting, and have a less wide frame, which might be useful for some forms of baffle-edge diffraction management.
The Faital-Pro 4FE35 drivers arrived, and I was able to test side by side with the alpair 7MS. I expected the differences might be subtle, but they are actually quite different little beasts.
The 4FE35 drivers render things like snares and symbols with quite a kick for their diminutive size, which might suit recordings of live events; quite upbeat and lively. Though they don't seem nearly as smooth as I expected from their official FR charts; they might need closer measurements and some notch treatment to get the best out of.
By comparison, the 7MS sound more laid back. I wasn't expecting that as they aren't very smooth in frequency response and have some over-bright jingly things going on for my taste, but even so they sound more rounded and produce vocals and instruments with noticeably more clarity IMO. I was hoping their flatter cone might do more for their high frequency dispersion, but the sweet spot isn't all that much bigger, that I can hear anyway. But for my application and preferences the 7MS is going to be easier to live with.
So, I'm happy with that; it performs well enough over a wide enough range at enough SPL for my modest requirements, and could arguably manage without a tweeter in my case. It has been interesting to find what cone size works best for me too, though it would be more of a compromise if I needed higher SPLs; my future room-speakers will likely use a different method.
The 4FE35 drivers render things like snares and symbols with quite a kick for their diminutive size, which might suit recordings of live events; quite upbeat and lively. Though they don't seem nearly as smooth as I expected from their official FR charts; they might need closer measurements and some notch treatment to get the best out of.
By comparison, the 7MS sound more laid back. I wasn't expecting that as they aren't very smooth in frequency response and have some over-bright jingly things going on for my taste, but even so they sound more rounded and produce vocals and instruments with noticeably more clarity IMO. I was hoping their flatter cone might do more for their high frequency dispersion, but the sweet spot isn't all that much bigger, that I can hear anyway. But for my application and preferences the 7MS is going to be easier to live with.
So, I'm happy with that; it performs well enough over a wide enough range at enough SPL for my modest requirements, and could arguably manage without a tweeter in my case. It has been interesting to find what cone size works best for me too, though it would be more of a compromise if I needed higher SPLs; my future room-speakers will likely use a different method.
Last edited:
So am I generally, but my thinking here is to minimise this sort of thing (excerpt from a Klippel doc.) :
So, at a few kHz the inductor presents several ohms of impedance in series -- the more the better -- and the variation becomes less significant because it's a smaller fraction of the total impedance.
So that would improve IMD in complex music signals where multiple frequencies are playing at the same time. But even if excessive displacement is prevented by box tuning, the bass can still pump high current, so that would be another source of impedance modulation and messy hysteresis effects.
You could take things further, and use a larger inductance together with an active filter to boost the treble back. A project that I've been psyching myself up to do (when not busy with life) is to build a current amplifier, or "mixed mode" (not 100% sure what topology yet and even just doing sims takes ages), so the impedance is emulated by the circuit.
But a 100W class D with 47 ohm in series may be quite hard to beat, especially if you already have EQ to fix the lumpy response.
Interesting stuff, thank you. I was aware that passive components offer some useful capabilities, but hadn't reasoned it through in this case.
I'll sort out the baffle and cabinet, then need to take some detailed measurements to see more closely what is required in terms of smoothness and response correction. There'll be baffle step too, since I intend to cross over low enough for it to occur within the range covered by this driver.
Thanks again
Kev
I'll sort out the baffle and cabinet, then need to take some detailed measurements to see more closely what is required in terms of smoothness and response correction. There'll be baffle step too, since I intend to cross over low enough for it to occur within the range covered by this driver.
Thanks again
Kev
Kev06, Have you used alpair 5.3 before 7ms? Im thinking of this upgrade. I could imagine more body at higher volume in my waw speakers. My highpass is around 350hz. Have you done the baffle step compensation?
Of course, you could always do something like this. 😉
Pair of 10in Scan bass units, 14 x CHN-50 in twin hexagonal clusters forming an MTM with a beryllium dome tweeter.
Granted -it's not peanuts. And these aren't mine -they're a 'final' build by a friend who asked me to help him think of something interesting that would beat his Edingdale GTs (and after he & his good lady wife decided on the general approach from a dozen or so ideas kicked around, guess which mug then had a dinosaur's share of the design work 😉 ). They're not half bad though.
Ideally I'd have gone with the original intent of over-under woofers for a WMTMW configuration, but to put the tweeter at ear level the height would have started to completely dominate Colin's room, so he (rightly IMO) shifted to the MTMWW configuration to keep size manageable for his space. No point in building a speaker that you can't live with.
Pair of 10in Scan bass units, 14 x CHN-50 in twin hexagonal clusters forming an MTM with a beryllium dome tweeter.
Granted -it's not peanuts. And these aren't mine -they're a 'final' build by a friend who asked me to help him think of something interesting that would beat his Edingdale GTs (and after he & his good lady wife decided on the general approach from a dozen or so ideas kicked around, guess which mug then had a dinosaur's share of the design work 😉 ). They're not half bad though.
Ideally I'd have gone with the original intent of over-under woofers for a WMTMW configuration, but to put the tweeter at ear level the height would have started to completely dominate Colin's room, so he (rightly IMO) shifted to the MTMWW configuration to keep size manageable for his space. No point in building a speaker that you can't live with.
Attachments
Last edited:
Yes I have, though both with woofer assistance (not full range). Imo the 5.3ms is definitely the better midtweeter. The top end is very good and the midrange is quite smooth. But of course the 7ms has more dynamics lower down, if you want or need to cross lower. It also can go louder. So each has their place.
I've done BSC in DSP, but how much you need and at what frequency is very dependent on the baffle/cabinet and room placement. So the compensation i applied won't be directly transferable to other designs and situations.
Scott, those look fairly unique! Though when I make my next big speakers, I want to try Geddes-like compression drivers in waveguides.
That crude, Tekton inpired way-to-big-spacing.
It did inspire this which could be interesting and workable:
http://wodendesign.com/downloads/CHN-40-array.pdf
dave
In my own defence, it depends on where the design priorities lie. The spacing to each cluster centre from the central tweeter is 1 wavelength, which assuming the theoretical ideal of 1/4 wavelength at the transition frequency is impractical for other performance reasons is one of the better compromises available in terms of response and polars. That's assuming the object is a pure point-source of course -in this case, the MTM with its small limit (and it's small in this design) on the vertical polars was desirable.
Thanks, this is what i thought... currently im using the alpair 5 in closed enclosure with a tang band w5 1138, and they are a good couple. Although im planning to scale it a little bigger alpair 7 with epique 7 woofer. quriosity never ends... 😀 what woofer do you use with the alpairs?
With the 5.3ms I happened to use the scanspeak 15W/8530K01, though that was only because I already had a pair and wanted something low in fs but compact. If more space had been available I'd probably have gone with bigger (and likely cheaper) woofers; even quite large ones can go well past typical WAW crossover frequencies (before cone-breakup, beaming or centre-to-center distance start causing issues).
With the 7MS I didn't get as far as settling on a specific design. But I tested it with some 10" scanspeak woofers, and that seemed like a decent combo. Though a lot of 12" types would still be fine. I also tried a couple of 8" SB woofers, and liked those results too.
However in light of recent threads, for larger speakers I'd probably not use one woofer per side, but two smaller ones. If crossed over below say 200hz or less, they can be put in the sides of the cabinet, facing away from each other (opposing), so they cancel each other's vibrations. I find it interesting that crossover frequencies at the lower-end of typical WAW-type configurations are only one or two hundred hertz above subwoofer territory. So, quite long-throw woofers might be used if they go high enough for WAW without distortion or breakup, and using these opposed seems more worthwhile.
(EDIT: That, incidentally was one of my main reasons for becoming interested in the 7MS; although the 5.3MS is IMO a better midtweeter, I wasn't convinced by it crossing over so low. An alternative would have been to stay with the 5.3MS crossed higher and do something else for the low end - like add subwoofers or use a woofer in a big transmission line or similar, but I also liked the higer SPL capabilities of the 7MS, so didn't choose the 5.3MS route. Yet.)
With the 7MS I didn't get as far as settling on a specific design. But I tested it with some 10" scanspeak woofers, and that seemed like a decent combo. Though a lot of 12" types would still be fine. I also tried a couple of 8" SB woofers, and liked those results too.
However in light of recent threads, for larger speakers I'd probably not use one woofer per side, but two smaller ones. If crossed over below say 200hz or less, they can be put in the sides of the cabinet, facing away from each other (opposing), so they cancel each other's vibrations. I find it interesting that crossover frequencies at the lower-end of typical WAW-type configurations are only one or two hundred hertz above subwoofer territory. So, quite long-throw woofers might be used if they go high enough for WAW without distortion or breakup, and using these opposed seems more worthwhile.
(EDIT: That, incidentally was one of my main reasons for becoming interested in the 7MS; although the 5.3MS is IMO a better midtweeter, I wasn't convinced by it crossing over so low. An alternative would have been to stay with the 5.3MS crossed higher and do something else for the low end - like add subwoofers or use a woofer in a big transmission line or similar, but I also liked the higer SPL capabilities of the 7MS, so didn't choose the 5.3MS route. Yet.)
Last edited:
... Results in impressively low HD in the midrange, and enough "grunt" to allow a relatively low crossover, which alleviated some of the acoustic issues in that room.
Chris
Indeed, although the considerable help of a 'certain person' who optimised the active bass section LP filter & EQ also has something to do with that. 😉 I can't imagine who it might have been, but he did a rather good job. You might even say Wilson Benesch could learn a bit from them. 
It's an interesting aspect of the multi-driver midrange though, just as with arrays. With 14 x CHN-50 per channel, you've got a radiating surface roughly equivalent to an 11in driver through their BW, but with a greater number of coils to dissipate heat (on top of the other aspects).
It's an interesting aspect of the multi-driver midrange though, just as with arrays. With 14 x CHN-50 per channel, you've got a radiating surface roughly equivalent to an 11in driver through their BW, but with a greater number of coils to dissipate heat (on top of the other aspects).
Last edited: