Hi,
I am an old home speaker builder from the 1970s and 80s who has not built anything since 1992, before the easy availability of all the home speaker computer programs. I was thinking of building one last pair of speakers just for the heck of it. A company called Spatial Audio has a crazy loudspeaker design called the M3 Sapphire. It uses the well regarded Peerless 1.25 inch corundum dome tweeter in an almost crossoverless design using just a 33uf capacitor in series with the tweeter. This large cap allows the tweeter to run down to its resonance frequency near 600hz. The company owner told me that the M3 Sapphire could play comfortably to 90db, but he did not specify room size or give a power rating for the loudspeaker in watts. He mates the tweeter to two large woofers in an open baffle design.
Spatial Audio M3 Sapphires - MyAudioPhrenia
My crazy idea is to try to mate the same tweeter with a single 33uf cap to a 8 inch diameter Peerless 830869 8" Nomex Cone HDS Woofer, with just a 2.5 mH coil in series with the woofer to simulate a 6db per octave crossover at 600hz. I would put the woofer in a 9.5 inch wide, 11 inch high, and 12 inch deep cabinet made of 3/4 inch thick MDF and covered in Formica. The tweeter would be mounted on a separate home made MDF stand and placed on top of the woofer cabinet in an air mounted, minimum diffraction design that is cushioned from vibrations by a foam pad. So the idea is to end up with a pulsating hemisphere with a range from just below 600hz to beyond 20,000hz mounted in free space on top of a pretty mellow and fast low q (Qts .31) 8 inch woofer in a sealed cabinet with a Qtc of about .9. I would use the speakers as satellites mated with two subwoofers at 80hz or so. I know it sounds like a crazy stupid idea, but it just might work. If Spatial can get away with using the tweeter almost full range, then anyone should be able to do it. I do not listen to loud music anyway, so it could work for me and would be easy to put together. The woofer and tweeter have very close sensitivity ratings, so there would probably be no resistor compensating network needed, just one cap and one coil per speaker. I would appreciate comments, suggestions, etc. See detailed specs below.
Peerless DA32TX00-08 1-1/4" Corundum Dome Tweeter
8 ohms, Re 6.78 ohms, 89.78dB 2.83V/1m, Listed at a whopping 120 Watts RMS! Fs = 573.04Hz
https://www.parts-express.com/pedocs/specs/264-1678--peerless-da32tx00-08-spec-sheet.pdf
Peerless 830869 8" Nomex Cone HDS Woofer
8 ohms, (Re) 5.9 ohms, 90.2dB 2.83V/1m, 60 Watts RMS - 150 Watts Max.
https://www.parts-express.com/pedocs/specs/264-1098--tymphany-hds-p830869-spec-sheet.pdf
I am an old home speaker builder from the 1970s and 80s who has not built anything since 1992, before the easy availability of all the home speaker computer programs. I was thinking of building one last pair of speakers just for the heck of it. A company called Spatial Audio has a crazy loudspeaker design called the M3 Sapphire. It uses the well regarded Peerless 1.25 inch corundum dome tweeter in an almost crossoverless design using just a 33uf capacitor in series with the tweeter. This large cap allows the tweeter to run down to its resonance frequency near 600hz. The company owner told me that the M3 Sapphire could play comfortably to 90db, but he did not specify room size or give a power rating for the loudspeaker in watts. He mates the tweeter to two large woofers in an open baffle design.
Spatial Audio M3 Sapphires - MyAudioPhrenia
My crazy idea is to try to mate the same tweeter with a single 33uf cap to a 8 inch diameter Peerless 830869 8" Nomex Cone HDS Woofer, with just a 2.5 mH coil in series with the woofer to simulate a 6db per octave crossover at 600hz. I would put the woofer in a 9.5 inch wide, 11 inch high, and 12 inch deep cabinet made of 3/4 inch thick MDF and covered in Formica. The tweeter would be mounted on a separate home made MDF stand and placed on top of the woofer cabinet in an air mounted, minimum diffraction design that is cushioned from vibrations by a foam pad. So the idea is to end up with a pulsating hemisphere with a range from just below 600hz to beyond 20,000hz mounted in free space on top of a pretty mellow and fast low q (Qts .31) 8 inch woofer in a sealed cabinet with a Qtc of about .9. I would use the speakers as satellites mated with two subwoofers at 80hz or so. I know it sounds like a crazy stupid idea, but it just might work. If Spatial can get away with using the tweeter almost full range, then anyone should be able to do it. I do not listen to loud music anyway, so it could work for me and would be easy to put together. The woofer and tweeter have very close sensitivity ratings, so there would probably be no resistor compensating network needed, just one cap and one coil per speaker. I would appreciate comments, suggestions, etc. See detailed specs below.
Peerless DA32TX00-08 1-1/4" Corundum Dome Tweeter
8 ohms, Re 6.78 ohms, 89.78dB 2.83V/1m, Listed at a whopping 120 Watts RMS! Fs = 573.04Hz
https://www.parts-express.com/pedocs/specs/264-1678--peerless-da32tx00-08-spec-sheet.pdf
Peerless 830869 8" Nomex Cone HDS Woofer
8 ohms, (Re) 5.9 ohms, 90.2dB 2.83V/1m, 60 Watts RMS - 150 Watts Max.
https://www.parts-express.com/pedocs/specs/264-1098--tymphany-hds-p830869-spec-sheet.pdf
Very cool idea. I would love to see your project progress.
It does seem like Zangers might be a great place to start!
It does seem like Zangers might be a great place to start!
Do not run the tweeter with just 33uf, full-stop. That's **** poor design and is not an acceptable use of that tweeter under any circumstance.
You are probably right. A second order crossover at 1,500Hz would be a lot safer. But, what would it sound like at 600Hz? I am curious. Perhaps I will order parts for both crossover frequencies and do some cautious testing.
Why don't you consider deploying a dome midrange unit and cross it at 600-700 Hz? That is the right way and will bring you terrific sonics.
Running anything that small down to 600Hz is a bad idea. Why would you want to settle for elevated distortion levels and such a restriction in dynamics?
Why not do a filler driver design a la B&O? It would play louder, sound better and you won't blow up tweeters under typical conditions. You could still achieve that theoretical flat phase response without hearing the tweeter working in its fundamental resonance range, even with electrical resonance compensation.
Why not do a filler driver design a la B&O? It would play louder, sound better and you won't blow up tweeters under typical conditions. You could still achieve that theoretical flat phase response without hearing the tweeter working in its fundamental resonance range, even with electrical resonance compensation.
I suggest a great deal of caution!
Consider if you use a 1st order at 600Hz and turn the speakers up to produce 90dBspl the tweeter will still try to reproduce 81dBspl at 300Hz and 75dB at 150Hz.
Consider if you use a 1st order at 600Hz and turn the speakers up to produce 90dBspl the tweeter will still try to reproduce 81dBspl at 300Hz and 75dB at 150Hz.
why not a second order filter instead ?
This 8" unit is said a little difficult at producing great bass I was said.
In such a project I would consider maybe :
8″ SB23MFCL45-4 / Polypropylene – Sbacoustics for playing with the bafle step with a Rock n'Roll XO 12 db slope XO around 1300/1500 hz with the DA32TX and accepts some nulls that arms less than peaks.
Or more seriously its 6" little brother... or the sb17cac but with higher slope.
Maybe the Visaton 8" aluminium but with a LR24. (greater bass I surmise than the sb17cac).
XrK971 in the Full Range section has made a 2 ways with a 8" and a 3" with a first slope at around 300 hz but the 3 is a full range which is supporting 250 watts at those cut-offs... but it doesn't play loud.
First slope are difficult to do well...
This 8" unit is said a little difficult at producing great bass I was said.
In such a project I would consider maybe :
8″ SB23MFCL45-4 / Polypropylene – Sbacoustics for playing with the bafle step with a Rock n'Roll XO 12 db slope XO around 1300/1500 hz with the DA32TX and accepts some nulls that arms less than peaks.
Or more seriously its 6" little brother... or the sb17cac but with higher slope.
Maybe the Visaton 8" aluminium but with a LR24. (greater bass I surmise than the sb17cac).
XrK971 in the Full Range section has made a 2 ways with a 8" and a 3" with a first slope at around 300 hz but the 3 is a full range which is supporting 250 watts at those cut-offs... but it doesn't play loud.
First slope are difficult to do well...
Last edited:
Look at the displacement of a driver to determine how low(and at what spl) you can go without hurting it. Distortion may limit you further, you also should allow for ample headroom per your taste in music.
In fact, the closer you get to fs, the steeper you must filter.
The lower the slope, the further you must go.
Ironically, most successful low-order crossovers are anything but simple, including impedance correction circuits. Thus defeating 'purist' notions of simplicity.
In fact, the closer you get to fs, the steeper you must filter.
The lower the slope, the further you must go.
Ironically, most successful low-order crossovers are anything but simple, including impedance correction circuits. Thus defeating 'purist' notions of simplicity.
Now maybe it's possible in the wave guide formerly made by T Gravsen for the 1.3 Audax ?
The WG + the Corundum 1 mm Xmax and its 500 hz Fs could be ok for a 1500 But6 ?
Plus the fact than a WG helps for the physical horizontal acoustical center with the midwoofer. The WG could be below the midwoof to winn few phase recess in the same spirit if the op plans passive filter instead active delayed offset ?🙄
The WG + the Corundum 1 mm Xmax and its 500 hz Fs could be ok for a 1500 But6 ?
Plus the fact than a WG helps for the physical horizontal acoustical center with the midwoofer. The WG could be below the midwoof to winn few phase recess in the same spirit if the op plans passive filter instead active delayed offset ?🙄
Hello, I’m doing something a bit similar at the moment - after reading the Linkwitz lab pages, I bought some Aurasound NSW2 2” drive units which go from about 250Hz up to 20kHz. I’ve mounted these on top of an existing speaker and crossing over to the mid bass. I’ve chosen the xover freq to be where the midbass becomes acoustically large (near to the baffle step frequency - related to the cabinet width, not the drive unit itself). So at low frequencies, the design is omni, because the wavefront from the midbass diffracts in all directions. Then the aurasound takes over. Since this driver has no faceplate, it is acoustically small up to around 3khz. It starts beaming (due to the large cone) above 2.5kHz, at which point I crossover to a conventional tweeter.
Main design consideration is THD of Aurasound at a given spl related to its xmax and radiating surface area, then xover frequencies to trade off thd against omni-directionality.
I may post on another thread, if of interest.
Main design consideration is THD of Aurasound at a given spl related to its xmax and radiating surface area, then xover frequencies to trade off thd against omni-directionality.
I may post on another thread, if of interest.
Did Spatial Audio get a new owner?
Their original engineer generally made some nice speakers.
But using a 33uf cap as the only filter on a tweeter is just plain wrong, it's basically like running the tweeter with no filter at all.
Their original engineer generally made some nice speakers.
But using a 33uf cap as the only filter on a tweeter is just plain wrong, it's basically like running the tweeter with no filter at all.
A number of reviewers liked the finished product, so that got me interested. I have not heard the speakers myself.
I am not sure that it is the stock tweeter in the sapphires, I am recalling that in one of the many interviews Clayton Shaw said that the surround was different to allow greater Xmax. You may not have the same success or SPL levels with the standard corundum tweeter.
To give people an idea of how terrible an idea it is, I've run a sim using a Dayton RS28. No, it's not the same tweeter, but it shows the general idea of why doing this is boneheaded:
1) At 1khz the tweeter is only attenuated 4dB
2) At 300Hz the tweeter is only attenauted 12dB
So the power handling on the speaker would be something like one watt, if that.
The review says:
" I do feel this speaker is for the younger crowd that likes an upfront presentation. My previous speakers were just the opposite. The Magnepan’s and Quads are very laid-back speakers and over time I am sure I will get used to the M3’s."
I'll translate that into english:
"It sounds like the tweeter is about to explode and that's not a good sound."
Also the author's measured frequency response is terrible. (Note the scale is a ridiculous 130dB) I'm honestly kinda shocked that Clayton Shaw would release this, he's been making speakers long enough to know better. This looks like someone's first speaker design, which was why I was wondering if the company had been sold. I've seen a lot of cases where a loudspeaker company changes hands and then the new owners go in a completely different direction, while retaining the name and the reputation that preceded them.
Attachments
Last edited:
Aren’t they just using that cap on the tweeter for protection? It can play low and they just run the tweeter basically completely open besides that protection cap.
M3 Sapphires have 15in woofers.
If you use 8in you can proportionaly raise the crossover point to 1200 Hz and still get same "omnidirectional" behavior.
Also, human hearing is least sensitive to direction of sound in 800-1200 Hz range, so crossover point in that region should actually give better melt between drivers.
Not to mention better protection of tweeter.
So, my starting point would be 16uF in series with tweeter.
If you use 8in you can proportionaly raise the crossover point to 1200 Hz and still get same "omnidirectional" behavior.
Also, human hearing is least sensitive to direction of sound in 800-1200 Hz range, so crossover point in that region should actually give better melt between drivers.
Not to mention better protection of tweeter.
So, my starting point would be 16uF in series with tweeter.