Perhaps less overall cost than a two-way, which might not be essential for the use case implied? *
While I no longer do the majority of my viewing/casual listening at a computer desktop - now it’s an iPad and Senheisser BT buds or circumaurals for that application- I was very much satisfied for several years with the FF85WK in the same style enclosure Dave posted above.
* maybe I spoke too soon? just fired up currency converter and 1856DKK = $336CAD, before freight, taxes, etc. 😳
While I no longer do the majority of my viewing/casual listening at a computer desktop - now it’s an iPad and Senheisser BT buds or circumaurals for that application- I was very much satisfied for several years with the FF85WK in the same style enclosure Dave posted above.
* maybe I spoke too soon? just fired up currency converter and 1856DKK = $336CAD, before freight, taxes, etc. 😳
Last edited:
Maybe. I have never subjected a driver to the “magic” you are intending to use. But i do know that when new people hear a good FR there are way more “wow, that is coming out of that driver” than people complaining about what they can do.
Perhaps a cheap experiment?
dave
I have limited access to the tools I use to to build enclosures, so I think I'll go with a two-way build instead of experimenting - for now. I'm sure the idea of compact one-way speakers will haunt me enough to try it out at some point 🙂
Yep.
If you use the excursion calculator to look at your system. and find that, when hitting 85dB:
- the woofer is at 100% of maximum excursion
- the wideband driver is chilling at 1% of maximum excursion (at your chosen crossover point)
...it means you the wideband is relatively overbuilt for the application, and you could consider using a smaller equivalent - which typically means a device that will give better / smoother treble. My own builds have shifted, over the years, to using the smallest HF device that a system will allow. e.g. if the choice was between:
(a) a robust but slightly rough unit that can be crossed at 300Hz
(b) a smaller, smoother unit that has to be crossed at 600Hz (for equivalent SPL)
...I would go with (b). But you may differ - this preference is very user dependent (music taste, age, degree of HF hearing loss).
How about a coaxial design instead of WAW? I came across the "TB new line of Coax FR drivers" thread and it seems that for example the W6-2313 or the SB16PFC25-4-COAX only lose to the W5-1138SMF by 3dB @ 30Hz when it comes to bass extension. But how do they compare in other aspects? WAW, if I understand correctly, makes it easier to deal with crossover related problems, but coaxials have "better" dispersion, right? What's more important for near field use, if there even is an objective answer?
Coax is better than separate mid+tweeter, but given the XO frequency, you VERY likely need some digital delay to get the tweeter within a quarter-wavelength of the mid, and you still have an XO right where the ear is most sensitive.
A well-done WAW is essentially a coax (due to the low XO), but can already be at less than quarter wavelength spearation so passive can be done that preserve time response.
Coaxials do not nesecassarily have better dispersion. Some of the better small FRs actiually have similr dispersion ro domes.
dave
A well-done WAW is essentially a coax (due to the low XO), but can already be at less than quarter wavelength spearation so passive can be done that preserve time response.
Coaxials do not nesecassarily have better dispersion. Some of the better small FRs actiually have similr dispersion ro domes.
dave
Could be a good start for a first build, but keep in mind that most coaxials in this size class generally have smaller diameter tweeter domes with higher crossover frequencies than usually employed in WAWs. Not necessarily bad thing, but the “transfer functions” of, and quality of parts incorporated in those XO’s are very often compromised for economic /size reasons. Of course there may be some co-ax driver designs that allow for external XO - but then I’ve not played much with those- in which case there could be plenty of opportunity to achieve excellent performance, but all the caveats given above would still apply.
and of course, Dave beat me to some of the same points, plus his own
and of course, Dave beat me to some of the same points, plus his own
Those Tangband and SBA coaxials don’t come with crossovers. They are raw drivers. In general, having the tweeter delayed by a coax or waveguide is a good thing for time alignment. If you can cross around 5kHz you can have the mid range act a lot like a full range driver as far as coherence of the band where we hear sound stage and imaging.
I had tweeter recessed in a waveguide and it made for perfect time alignment. I was able to make a close to transient perfect crossover with an asymmetric slope passive Harsch.
I had tweeter recessed in a waveguide and it made for perfect time alignment. I was able to make a close to transient perfect crossover with an asymmetric slope passive Harsch.
Just a quick note that I used a pair of KEF eggs (HTS3001SE) for this job. Blocked the port, EQ'd to death. They just about manage useful output down to the mid-30s, but it's easy enough to hit excursion limits with dubstep.
If I was to do it again, I'd pick 2x 4" mini-subwoofers (the Kartesian ones are worth a look) firing sideways with force cancellation. SB65 above a few hundred Hz.
Chris
If I was to do it again, I'd pick 2x 4" mini-subwoofers (the Kartesian ones are worth a look) firing sideways with force cancellation. SB65 above a few hundred Hz.
Chris
Your application (near field, 70dB max), is a nice "hack" in that it makes many normal problems / requirements go away.
In an normal WAW build, you can attempt to get:
a) wide enough dispersion to sound good across multiple listening positions.
b) fairly even dispersion, so that "room sound" (reflected sound bouncing off the ceiling and other surfaces) is pretty even.
If you listen at a large and variable distance from your speakers, in a lot of positions, all this is important, e.g, for a home gym ...but your application is the exact opposite of this.
(a) won't matter in desktop setup, because there is only one listening position. You don't need wide dispersion if you only have one skull.
(b) won't matter because, in a near field application, the reflected (room) sound will be very low, relative to the direct sound (you might have to fix reflections off your desktop and other close surfaces, but positioning + treatment are probably the best options for that).
I'm currently playing with the TB W2-2360S as a waw-widerange(300 to 600hz) and as a lower max spl tweeter that can cross where we commonly resort to a waveguide (1000 to 1500 hz). And that TB 5" can reach 1200hz before it gets wonky.
Also trying it in some classics like large Advent and ARs, since they were crossing 1500-1800(and not too efficient to match up).
Also trying it in some classics like large Advent and ARs, since they were crossing 1500-1800(and not too efficient to match up).
Does force cancellation have significant benefits for low SPL applications? And does side-mounting have any disadvantages in a WAW?
The Kartesians look nice indeed but lack excursion compared to the TB. Then again, they presumably play way cleaner at their Xmax
The application may be low-SPL, but excursion may still be high, depending on the frequency requirements. According to: http://www.baudline.com/erik/bass/xmaxer.html
A 4" driver needs to move 2.4mm one-way to produce 30Hz at 70dB. Obviously 4x 4" woofers will be able to go louder.
Force cancellation avoids the desk's mechanical resonances, cuts down on other things rattling, and stops the cabinets from walking around.
No disadvantages to side-firing woofers that I've found. It just needs accounting for in the crossover.
Having tried the Tang Band W6-1139 and the Kartesian Sub120-ferrite, the Tang Bands have considerably higher harmonic distortion, leading to a relatively muddy sound that can't be "cleaned up" with EQ. The Kartesian units surprise me every time I use them, with how clean they sound, even when you can see the cones moving from across the room.
Chris
A 4" driver needs to move 2.4mm one-way to produce 30Hz at 70dB. Obviously 4x 4" woofers will be able to go louder.
Force cancellation avoids the desk's mechanical resonances, cuts down on other things rattling, and stops the cabinets from walking around.
No disadvantages to side-firing woofers that I've found. It just needs accounting for in the crossover.
Having tried the Tang Band W6-1139 and the Kartesian Sub120-ferrite, the Tang Bands have considerably higher harmonic distortion, leading to a relatively muddy sound that can't be "cleaned up" with EQ. The Kartesian units surprise me every time I use them, with how clean they sound, even when you can see the cones moving from across the room.
Chris
Last edited by a moderator:
I don't mean to be spoonfed everything, but what kind of accounting would that mean?
Not sure if I'm sold on dual opposed with such small drivers. It would also require around 6.6" of speaker width with coaxial mounting, if I'm not mistaken. I could just put the speakers on sylomer pads, though it's surely not as efficient. You've convinced me to go with the ferrite Kartesian regardless. I wish other brands would publish some distortion figures like they seem to do. I've tried to compare some viable woofers by their SPL at 30Hz in a small box, but that doesn't really tell much when even a single 4" Kartesian can do 78.4dB 🙂
Now that I think of it, the aforementioned 4 incher from Purifi seems to do even better than the Kartesian if judged by manufacturer measurements. And 3dB more SPL, FWIW. Probably not worth the price increase for a first DIY design, though 😅
You could always offset the drivers vertically, so that the magnets are vertically aligned. Reduces cabinet width to more like 4", but should also keep vibration down.
When it comes to the crossover, you want to make sure your drivers are aligned in phase* at/near the crossover point. A physical offset between them means the more distant one will be lagging in phase a little more, so you might need to tweak the crossover accordingly, by perhaps using a lower-Q crossover with a softer knee on the low-pass.
* NB - phase is just time expressed as a fraction of a cycle. ie, 90 degrees at 100Hz = 2.5ms. 90 degrees at 1kHz = 0.25ms. There's a lot of air-of-mystery around phase online, but that's literally it.
Chris
When it comes to the crossover, you want to make sure your drivers are aligned in phase* at/near the crossover point. A physical offset between them means the more distant one will be lagging in phase a little more, so you might need to tweak the crossover accordingly, by perhaps using a lower-Q crossover with a softer knee on the low-pass.
* NB - phase is just time expressed as a fraction of a cycle. ie, 90 degrees at 100Hz = 2.5ms. 90 degrees at 1kHz = 0.25ms. There's a lot of air-of-mystery around phase online, but that's literally it.
Chris
Of course, forgot about phase alignment 🤦♂️ Definitely something I need to read more about though
Four Kartesians would cost me 400€ with shipping - not great, not terrible 🙂
Four Kartesians would cost me 400€ with shipping - not great, not terrible 🙂
I'm still trying to decide on the drivers. The SB65 got disqualified cause I find it rather ugly 🙂 I gave the NE65W another consideration and found these measurements that don't have the distortion peaks shown in Voice Coil and xrk's measurements. Maybe the driver has been refined since? The small size and rather smooth and wide dispersion are nice, although the latter might not be such a big deal in nearfield use? The resulting enclosure would be small enough to 3D print, making prototyping easy - I wouldn't lose much if the NE65W turns out to have high distortion levels.
I could use the B80 or 10F/8424 too. Them being similar in size to the Kartesian would give the speaker a rather nice Buchardt style look while being acceptably compact, but I don't know if they have much benefit over an improved NE65W.
I also noticed that the Kartesian Sub120 has an EBP of 121. The retailer said that sealed "is not recommended". If I understand correctly, these rules don't really apply for small drivers/high excursion? The simulations look fine enough with a Linkwitz transform. I could port them and have 70dB@30Hz with a port air velocity of 16m/s, but I don't know if that really makes sense. And that would be like F12, so I would still need EQ.
I could use the B80 or 10F/8424 too. Them being similar in size to the Kartesian would give the speaker a rather nice Buchardt style look while being acceptably compact, but I don't know if they have much benefit over an improved NE65W.
I also noticed that the Kartesian Sub120 has an EBP of 121. The retailer said that sealed "is not recommended". If I understand correctly, these rules don't really apply for small drivers/high excursion? The simulations look fine enough with a Linkwitz transform. I could port them and have 70dB@30Hz with a port air velocity of 16m/s, but I don't know if that really makes sense. And that would be like F12, so I would still need EQ.
Fair enough on the SB65. It's a nice driver, but not perfect: there's still some ugliness in the treble which requires taming. I've wondered if running it up to ~7kHz and crossing to a tweeter would be sensible. There'd be some vertical lobing in the HF due to the large (compared to wavelength) centre-to-centre distance, but that's unlikely to be a big problem once room acoustics are factored in.
With regards to EBP etc, don't worry about it: EBP is only useful to give a hint of where a driver might be at its most useful.
The recommendation against a sealed box might have been from the EBP, or it might simply be that a 4" driver in a sealed box won't make much bass compared to, say, an 8" in a ported box (a more typical compact subwoofer). ie, the retailer is trying to avoid potential customer disappointment.
Here, the application is fairly low SPL, and you're aiming to leverage the available excursion to get lots of LF extension.
If, for some crazy reason, you find there's not enough LF output for you, switching to ported will certainly help, at the potential expense of a larger cabinet, and a strongly-defined LF cutoff. Sealed is quite nice because you can choose, by changing a few numbers, whether you want -3dB at 10Hz and low SPL, or -3dB at 50Hz and much much louder.
Chris
With regards to EBP etc, don't worry about it: EBP is only useful to give a hint of where a driver might be at its most useful.
The recommendation against a sealed box might have been from the EBP, or it might simply be that a 4" driver in a sealed box won't make much bass compared to, say, an 8" in a ported box (a more typical compact subwoofer). ie, the retailer is trying to avoid potential customer disappointment.
Here, the application is fairly low SPL, and you're aiming to leverage the available excursion to get lots of LF extension.
If, for some crazy reason, you find there's not enough LF output for you, switching to ported will certainly help, at the potential expense of a larger cabinet, and a strongly-defined LF cutoff. Sealed is quite nice because you can choose, by changing a few numbers, whether you want -3dB at 10Hz and low SPL, or -3dB at 50Hz and much much louder.
Chris