If you have an accurate way to measure the impedance and the frequency response of each driver at a couple of different angles (as mounted in your box), then I don't see why this couldn't be done remotely.
For the 3-way speaker I recently designed, I built the box and mounted the drivers and then took a series of measurements including on axis and 45 degree off axis response for each driver alone. I also measured the sum and difference of drivers that share a crossover point (to zero in on the acoustic center differences). Combined with the impedance measurements, that's pretty much all it really takes. It's enough to show where diffraction is impacting the sound, and where dispersion is suffering. For drivers with significant distortion, you would probably want that measured, too.
Maybe some here would consider it blasphemy to suggest you could complete a crossover design without hearing the result, but I honestly think you can. If you are able to hit the combined response you want on and off-axis, and if you know the result will be low distortion then I don't see an issue with that.
For the 3-way speaker I recently designed, I built the box and mounted the drivers and then took a series of measurements including on axis and 45 degree off axis response for each driver alone. I also measured the sum and difference of drivers that share a crossover point (to zero in on the acoustic center differences). Combined with the impedance measurements, that's pretty much all it really takes. It's enough to show where diffraction is impacting the sound, and where dispersion is suffering. For drivers with significant distortion, you would probably want that measured, too.
Maybe some here would consider it blasphemy to suggest you could complete a crossover design without hearing the result, but I honestly think you can. If you are able to hit the combined response you want on and off-axis, and if you know the result will be low distortion then I don't see an issue with that.
Last edited:
Thanks so much, Steve! I was referring to the 18M/8631G00 dedicated midrange. You’re are spot on! I actually have the MUN-17 kit in the box here. I’ve been trying to redesign the mid/tweeter cab with a more esthetic and desirable tapered crown look. I’m trying to keep the front baffle dimensions the same, but it’s tricky as the crossover has to be rearranged to make room... I’d rather not put the crossover under the bass cab... Ha, that’s another project.
He also uses the 18M/8631G00 in the FUSION 12” 3-way. He reveres this midrange, and holds it in the same league as the AT 6”.
I do have a 13” 3-way with a Visaton TIW300, 18U/8741T00, and D3004/66000, which was one of Tony’s simulations. At 86dB, it’s stellar for solid-state. But for my tube gear, I’m aiming for a potent 90dB+ 13” 3-way. I thought perhaps the new Scan-Speak 32W/8878T11 might be the ticket... It’s 92dB, 21Hz Fs. It seems like it doesn’t give up low end for efficiency too much?...
There’s also the 32W/8878T00 8-Ohm woofer at 89dB, 19Hz Fs, looks like more low end, but probably needs more volume. I think the system here would dip below 89dB, not much different than what I have in my current Visaton/Scan-Speak 13” 3-way.
Photo: Here’s the right channel of the man cave - Visaton/Scan-Speak 13” 3-way (Tony’s simulation). The D’Appolitos are Troels’s design. All excellent in their own way, and voiced with caps based on cone material and genre of music.
Attachments
George, I am still confused here: 18M/8631G00 doesn't exist!
This one does: 18M/8631T00
As it goes, the MUN17-3W and Fusion are very similar projects. 200Hz and 2kHz crossovers. Just the Fusion uses a passive LC bass filter, whereas the MUN17-3W uses a Hypex for the bass.
What are we trying to do here? Design a passive bass filter? Or do you have all the MUN17-3W filter parts from Troels? 😕
This one does: 18M/8631T00
As it goes, the MUN17-3W and Fusion are very similar projects. 200Hz and 2kHz crossovers. Just the Fusion uses a passive LC bass filter, whereas the MUN17-3W uses a Hypex for the bass.
What are we trying to do here? Design a passive bass filter? Or do you have all the MUN17-3W filter parts from Troels? 😕
So sorry, I meant the 18M/8631T dedicated midrange driver. Ha, too many numbers to memorize!
Yes, I have the MUN-17 kit.
What I’m trying to do is to have the flexibility of interchanging the MUN-17 mid/tweeter cab with the Hypex powered 28W as well as a passive 32W/8878T11 13” woofer cabinet down the road. That’s why I want to keep the crossover inside the mid cab.
I’m not sure if the impedance will dip too low, or some other unforeseen issue when adding a 13” passive woofer to the existing MUN-17 cab?... Or perhaps the 32W/8878T00 89dB woofer might be a better electrical match here, I don’t know?...
George, you are a man who loves his speakers:
Looks like an Ellam Dappo:
Ellam d'
And A SB MTM:
SBA-16-MTM
But really, you seem like an unguided Missile to me. 😱
I say this as someone who comes from the UK Defense sector.
Bass is an almost inaudible foundation to good music. Tweeters just add a bit of air. It's all about the 300-3Khz voice midrange IMO.
Given your incredible range of quality drivers, I could build a much better speaker out of them. TBH, I think Troels is having a laugh sometimes. I'd never build a three way with a 6" midrange. Breaks up horribly at about 4kHz. Two lightening-fast 4" mids must work better.
Looks like an Ellam Dappo:
Ellam d'
And A SB MTM:
SBA-16-MTM
But really, you seem like an unguided Missile to me. 😱
I say this as someone who comes from the UK Defense sector.
Bass is an almost inaudible foundation to good music. Tweeters just add a bit of air. It's all about the 300-3Khz voice midrange IMO.
Given your incredible range of quality drivers, I could build a much better speaker out of them. TBH, I think Troels is having a laugh sometimes. I'd never build a three way with a 6" midrange. Breaks up horribly at about 4kHz. Two lightening-fast 4" mids must work better.
Last edited:
Ha, I assure you, I won’t be flying off anywhere soon. 🙂 Great looking speaker build, Steve!George, you are a man who loves his speakers:
Looks like an Ellam Dappo:
Ellam d'
And A SB MTM:
SBA-16-MTM
But really, you seem like an unguided Missile to me. 😱
I say this as someone who comes from the UK Defense sector.
Bass is an almost inaudible foundation to good music. Tweeters just add a bit of air. It's all about the 300-3Khz voice midrange IMO.
Given your incredible range of quality drivers, I could build a much better speaker out of them. TBH, I think Troels is having a laugh sometimes. I'd never build a three way with a 6" midrange. Breaks up horribly at about 4kHz. Two lightening-fast 4" mids must work better.
I’ve always liked the 15W/8530K midwoofer, and I like the 3-D, airy imaging from D’Appolitos. So, the Ellam D’Appo had to be built. As soon as the SBA MTM came out, I built that. At 91dB, its imaging and transient response is phenomenal on tube gear. The MW16P is a transparent, dynamic, clear midwoofer. It does, however, sound a little lean below 50Hz. I built them a little shorter, with the intention of placing them on subs should they be lean... My assumption was correct - I have them sitting on a pair of 32W/4887T00 Subs powered by a Behringer iNuke DSP3000 Amp.
I totally agree that 4”, or even 5” mids are better than a 6” or 7” mid. Even the 18U/8741T on my 13” 3-ways has to be crossed below 2kHz to avoid beaming and break up modes with such stiff cones. I think Scan-Speak may have a winner with the 18M/8631T where the response seems good north of 3kHz.
Back to the MUN-17, I wonder if a 80-90L, 32W/8878T11 or 32W/8878T00 woofer simulation can be retro-designed to fit the roll off and output of the mid/tweeter section?...
Last edited:
Here’s one channel of the SBA MTM and the 32W/4887T00 Sub.
The other is the darling eVeII with All Scan-Speak Classic drivers - 21/8555 + 18W/8545 + 98000. It was my first build - 2.5-way, series-fed crossover. That 98000 aluminum dome is phenomenal. I went with Mundorf Supreme Caps, and Mundorf L200 coils - great match for the paper drivers. Next to the eVe is the 13” 3-way Visaton/Scan.
The other is the darling eVeII with All Scan-Speak Classic drivers - 21/8555 + 18W/8545 + 98000. It was my first build - 2.5-way, series-fed crossover. That 98000 aluminum dome is phenomenal. I went with Mundorf Supreme Caps, and Mundorf L200 coils - great match for the paper drivers. Next to the eVe is the 13” 3-way Visaton/Scan.
Attachments
George, forgive me. I am just trying to keep up with this....
Having built two 12" plus 6" plus 1" speakers, why are you going for a third?
MUN17-3W
Surely at some point you say ENOUGH! 😀
I know those old Scanspeak 18W and 18M midwoofers. Heard them too.
There is the scruffy looking one. The Sliced one. And the pretty one with the nice curve.
I honestly think you might as well just swap tweeters and mids here. See how it sounds.
Nothing will break. But these are tiny differences.
Having built two 12" plus 6" plus 1" speakers, why are you going for a third?
MUN17-3W
Surely at some point you say ENOUGH! 😀
I know those old Scanspeak 18W and 18M midwoofers. Heard them too.
There is the scruffy looking one. The Sliced one. And the pretty one with the nice curve.
I honestly think you might as well just swap tweeters and mids here. See how it sounds.
Nothing will break. But these are tiny differences.
I often swap the 13” stereo subs interchangeably with small monitors, Ellam D’Appo, SBA MTM, and with the eVeII. I can’t swap the mid/tweeters because the crossover is inside the mid cab in the eVes.George, forgive me. I am just trying to keep up with this....
Having built two 12" plus 6" plus 1" speakers, why are you going for a third?
MUN17-3W
Surely at some point you say ENOUGH! 😀
I know those old Scanspeak 18W and 18M midwoofers. Heard them too.
There is the scruffy looking one. The Sliced one. And the pretty one with the nice curve.
I honestly think you might as well just swap tweeters and mids here. See how it sounds.
Nothing will break. But these are tiny differences.
My intention for another 13” 3-way system was having higher efficiency in the 90-92dB range to use on tube gear. My Visaton/Scan-Speak 13” 3-way is only 86dB - great for solid state drive with a 100W/ch, but on SET tube gear I need a more efficient system. I guess I could build the MUN-17 first, and see how it turns out...
Interesting, how does a 32W4878 which by SS datasheet is a 90dB driver measured 2pi, turns out according to B1371 specs, 93dB/2.83V/1m.
I'd be looking at high efficiency, pro-audio woofers from the likes of FaitalPro, B&C and Eminence etc for tube gear.
EDIT: MUN-17 should be fine because of the active woofer.
Last edited:
@Johasz, I’ve looked at the B1371. The series 120uF, and 220uF midrange film caps would be a fortune. The crossover with a 3-order tweeter and 4-order mid and woofer is a dated design. I’d prefer LR2 filter.
@Lojzek, I agree. The crossover doesn’t look convincing.
@fatmarley, that makes total sense as these are much more efficient drivers, albeit higher Fs. I have a suspicion that I may still need a sub with a 12” 3-way with these high efficiency drivers.
@Lojzek, I agree. The crossover doesn’t look convincing.
@fatmarley, that makes total sense as these are much more efficient drivers, albeit higher Fs. I have a suspicion that I may still need a sub with a 12” 3-way with these high efficiency drivers.
Do you really think the 18M/8631 "breaks up horribly at about 4kHz"? As I see it, the response drops off quite smoothly up as far as 7k or so, and behaves pretty well off axis. The MUN-17 has a 4th order M-T crossover at 2kHz, so it shouldn't be an issue anyway.
I am almost at the end of building mine, and will report when they are back in the living room. Judging from a temporary installation over the festive break, they are going to sound very good indeed.
Alex
Last edited:
On the contrary, Steve, I think Troels knows exactly what he is doing with this design. I've already posted the lateral off-axis response family from the MUN-17 page on a different thread, but here it is again, because I think it is rather impressive. The combination of the midrange driver, the choice of crossover point and slopes, and the M-T baffle design, give one of the smoothest set of lateral dispersion curves I've ever seen.
Alex
Troels used baffle step compensation to get an flatter response. The 18M/8631G doesn’t exhibit the cone break-up and beaming that the 18U/8741T does. Even the 18W/8531G00 is better than the 18U/8741T00 in this regard for mids. However, the 18U has an underhung voice coil exhibiting higher power handling, less SPL, dynamic range, and less compression than all previous Scan-Speak generations.
if you have a functional Mid-Tweet already developed by Troels the easiest way to add a bass unit is by bi-amping and using digital Xover and EQ. Most flexibility to adapt to a room. It also allows one to put a better amp on the mid-tweet section. I do not see a real purpose in redesigning Troels's Xover which apparently measured exceptionally well even off axis (he rarely goes to trouble to measure there and/or show the results).
the passiv crossover off the medium of the Mun17 looks much more complicated than that of the fusion speakers !
on the fusion, has he taken into account the tall of the front face in these calculations, for the smoothing of the response
on the fusion, has he taken into account the tall of the front face in these calculations, for the smoothing of the response