3-way studio monitor design

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
When I was considering 8" drivers, the SB23NRXS45 was the one that seemed to model the best. However, the parameters looked inconsistent between independent tests (Zaph, Audioexcite) and factory specs. This was not just true of compliance determined parameters like Fs and Vas,but Mms and BL as well. This seemed to be the case with all the non Satori SB line. I am currently using the SB29RDC tweeter with three Discovery 15W/4524T per channel. I am seriously thinking of going to the R2604/8320 tweeter instead.
 
Alright, good to know.
For the tweeter I would suggest using a waveguide, to support crossing it lower and controlling the directivity. On this forum you'll a 3D-printable (or CNC) waveguide for the SB Acoustics SB26ADC, which is a very nice tweeter.
Open source Waveguides for CNC & 3D printing!


I'll follow this thread with interest :).


A good site for tweeters + waveguides is DIY Speaker | Speaker Kits, Acoustics
He tests them and also designs speakers, not any 3-way yet, although one is coming.
 
For the woofer I need to do a choice before building cause it really influence the cabinet design and they cost more. I need a woofer that cad dig well to the 30Hz.
OK, cabinet cost is also an important factor. Given a choice between:


a) A well-tuned high Xmax 8" driver in a compact box
b) More box volume and more cone area (e.g. 2 x 10 or 12 bipole) but only 5mm Xmax


My bet would be on the latter. The fat surround (a) gives more head room, but it also has to be less linear. Stiff heavy rubber = lots of hysteresis. A small VAS also means that the suspension is very stiff, so to minimise suspension-related distortion, the box should be relatively small, so that the air suspension becomes dominant.


IMD will be a b**** if you're trying to push +/- 5mm at 20Hz while XO at 250Hz. Use 2 x 10" or 12" with modest drivers in a bipole and you won't regret it. The only downside will be the large box, but even that will likely have some advantages.



I've had a pleasant experience with the 10" version of these...
W.32.500.8.MC - Nagłośnienie STX
As you can see, they're just plain paper (or some kind of cellulose) with a stamped steel structure, and the bass is really clean. And they've actually been halfway listenable running full-range in 40L sealed boxes while waiting for new active XOs.
 

Attachments

  • Screen Shot 2019-02-16 at 18.57.47.png
    Screen Shot 2019-02-16 at 18.57.47.png
    52.4 KB · Views: 414
Can you explain further? I can't see why...


Some of the parameters such as the inductance or the spring constant will change depending on the position of the voice coil. As LF push the cone around, they will modulate the amplitude of the higher frequencies riding on top. When that happens, the higher frequencies will get side-bands.


For example a loud 20Hz signal together with 250Hz. The interaction will produce new frequencies like 230Hz and 270Hz. And with more complicated signals, they all modulate each other.



Also, there's no guarantee on the harmonic distortion. 250Hz might be nice and clean in artificial tests when the cone is centred, but what happens when a +/-1mm signal is pushed 5mm off-centre? A reliable way to reduce at least some of these problems is by reducing the peak-to-peak amplitude.


Didn't know that driver. It's simulates really well in 20lt ported, tuned @35Hz, but in a sealed box it start sloping around 100hz...
edit: are you talking of STX W.22.250.8.FCX ?
I've got the W.27.400.8.MC
 
Ah, I see.

I thought the main cause of IMD was the doppler shift (in the case of +/- 5mm excursion, for the wave length of a 250Hz tone means a FM of +/-0.9Hz), that was ok for me.
I start realizing that designing a speaker it's all about finding the best compromise for many many variables. You can't have the butter and the money from the butter...
With 2 x 10" / 12" bipole my cabs would be muck bigger, that's a constraint for me.
Maybe I just have to renounce to some LF extension...
 
I start realizing that designing a speaker it's all about finding the best compromise for many many variables. You can't have the butter and the money from the butter...


Well, maybe take a step back and re-iterate the purpose? You said 700eu per speaker, but then mentioned multi-channel surround sound later on. Will the mastering be done with some commercial format in mind like 7.1 or Dolby Atmos?


What's the 'vision' or goal? Making movie soundtracks with 3d ear-candy? Layered orchestral scores? What sort of audience is it for? Personally, I kind-of despise the hype. Atmos seems like a big step forward, but the older formats with 'hard-wired' surround channels have never really impressed me.


Your priorities may differ, but If I were setting up a music mastering studio for myself, I would do something like this:
focus on 2 really nice active monitor speakers. 3-way, active/DSP filters, 6 amplifier channels, maybe OB for the mid, and even try my hand at DIY-ing the amplifiers, so the bass gets class-D, the mid-range/highs get class-A, and experiment with some exotic output impedance adjustments to reduce distortion.


Then, if I want to do some work with 3d sound, I'd add a different set of smaller speakers, maybe 4/5" full-range, maybe open-baffle, although that may be too far removed from the "miniature cubes" that most people have to endure, I mean enjoy. OTOH, sometimes you have to lead by example, so if people want to hear the music the way you intended it, they should upgrade to your speaker system, and not the other way around.
 
Last edited:
I mix and master my location recordings of chamber music / small orchestras from baroque to contemporary. Sometimes I've been asked multichannel SACD, sometimes 5.1 mixes for small DVD productions (S channels just for room ambience).
I'm thinking about stretching a bit my budget and get SS 18M/4631T00 or satori MR16Ps mid driver, to be able to XO a bit lower. Never heard this drivers but measurement looks so nice!

I'll stick with Dayton RSS210 as a woofer and R2604/832000 with waveguide as tweeter.
About the OB suggestion, I'm thinking about having a semi-open cab for the mid, like dutch&dutch 8C, to have some more cardioid behaviour in the lowmids, but it's difficult to find some documentation about this kind of design. Any hint?

I think I'll just build two MDF cabs for testing before doing the final ones (two for the moment).
 
Last edited:
I like the cardioid idea. I went that route too and I ended up improvising with the mid-woofer. It's an Eton 7" hexacone. The boxes are about 15L, with curved wooden ribs on the sides, fibreglass netting and felt glued onto the outside. The lossy area was initially around 50%, but I found that I had to reduce it significantly. Not only that, but the system seemed to resonate a bit, suggesting that the felt was too porous, and I had not designed a good method of tuning the resistance to air flow. "Mark 2" will probably have some kind of sandwich structure where filler can be added or removed as needed.



I'm also curious about an open baffle with a transition region, e.g. a 15cm buffer zone with padded holes or slots around the edge. Edge diffraction seems like just another flavour of box resonance to me, except that the dimensions are usually a bit larger so the interference amplitude is lower.


The Scanspeak looks really nice. The cone looks like it's designed to bend, bringing the break-up modes down to a low frequency, so I'd also experiment with adding some series resistance (or a specially designed amplifier ;) ) to make the damping factor really soft.
 
You can't go wrong with the Satori 6" driver for the midrange. The only fly in the ointment is the cone/surround resonance that causes a mild dip and 2nd order distortion spike at ~1700Hz. Nothing you can do about this, it's a soft cone driver and this is typical. Going aluminium for the cone material gets around this. But this is nit picking.
 
Thanks for your reply 5th element and congratulation for your FSTNT1.
I'll take the time to do some reading to see what specs works better with a semi-open baffle, if I finally decide to go on with that idea. I give myself till the end of the week before pulling the trigger, so still open to suggestions.
I think dutch&dutch uses a custom 8" seas metal driver for the mid. But they are all quite expensive, not very linear, and they all have strong peaks of 3rd order distortion between 1 to 2k. Why somebody would chose one of these? Then you have to chose a big tweeter to cross lower and that means other issues... am i missing something?
 
The provided links are interesting and inspiring.
Dutch & dutch 8c seems to be the state of art of passive cardioid speakers... but... I searched the web for transients and phase behavior measurement but couldn't find any. Do you know if somebody did this measures on them?
What values qts should I be looking for, for a mid driver that could work well in such a design?
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.