Party speakers aside, the last revision of @andy19191 's OP is starting to make more sense. It would be a good thing if this project becomes something like an update of the Paul Carmody Tarkus mentioned by @diyiggy -- but I remain dubious that a modest cost speaker can play the role of monitor (presumably a studio monitor for mixing?) can also be a good all-arounder for home hifi. Not without DSP to switch between different profiles.
We can make a design that fits both roles, depending on the crossover. A good box with 2 options, the passive crossover for more vintage style monkey coffin and the active dsp version (or the version that allows it) without crossover. The cabinet and the drivers can be the same for both.You can even make it that the dsp sits in the footer/stand, so it's easy to remove or add later. A passive crosover does not have to be inside the box of the woofer (like many do it) to work, it just need to be between the amp and the drivers.
I really do love the Tarkus shape, but it is surely too much wood work for the people interrested, imo.
If I had to do a Tarkus today and balancing the cost of drivers, I would try a NE-149W-04 as we are going to see them poping up again. A Seas 22 TAF/G with an electrical cut off around 2K5 hz.
Choice of the woof is a little more problematic. Could be again the SLS 12" of the Tarkus or the Faital FE300 or Beyma talked before. Problem is the Peereless as a 5" must croosed over 150/200 stiff or a little higher if 12 dB (something 250/300 electrical) , acoustic filter 300/350 hz.
Problem is we are at the limit of what can do cleanly a hifi 12" (above 200 hz acoustical, most are less clean as subwoofer or if they are lack bass as woofers.; PA being all mid-woofer but the 10FE430 & 12FE430 (too much expensive maybe).
One of the problem of bass can be solved by choosing a woofer and make it vented. Or going to woofer in H or U frame with high efficienty PA one but loosing 10 dB (so needing at minima something 95 dB or more efficienty which is most the attribute of PA 15"... and anyway this is not MK coffin anymore).
Maybe a 10" vented could feel the gap till the Peereless 5.5", something SS Discovery or the SB Wop24 (the 5.5" and the Seas tweeter are not expensive but sounds high end). Of course it will be limited in term of SPLand needs a good amp as the impedancewilll be on the low side (2.8" to 3" for the least).
As one of the goal is the mid/tweet combo on a plate to be moved in the front baffle cause of horizontal placement choice, Simplier way could be to do the classic two box: bass + mid/tweeter, but on the same width in order it still shape as a rectangle if horizontaly .
If I had to do a Tarkus today and balancing the cost of drivers, I would try a NE-149W-04 as we are going to see them poping up again. A Seas 22 TAF/G with an electrical cut off around 2K5 hz.
Choice of the woof is a little more problematic. Could be again the SLS 12" of the Tarkus or the Faital FE300 or Beyma talked before. Problem is the Peereless as a 5" must croosed over 150/200 stiff or a little higher if 12 dB (something 250/300 electrical) , acoustic filter 300/350 hz.
Problem is we are at the limit of what can do cleanly a hifi 12" (above 200 hz acoustical, most are less clean as subwoofer or if they are lack bass as woofers.; PA being all mid-woofer but the 10FE430 & 12FE430 (too much expensive maybe).
One of the problem of bass can be solved by choosing a woofer and make it vented. Or going to woofer in H or U frame with high efficienty PA one but loosing 10 dB (so needing at minima something 95 dB or more efficienty which is most the attribute of PA 15"... and anyway this is not MK coffin anymore).
Maybe a 10" vented could feel the gap till the Peereless 5.5", something SS Discovery or the SB Wop24 (the 5.5" and the Seas tweeter are not expensive but sounds high end). Of course it will be limited in term of SPLand needs a good amp as the impedancewilll be on the low side (2.8" to 3" for the least).
As one of the goal is the mid/tweet combo on a plate to be moved in the front baffle cause of horizontal placement choice, Simplier way could be to do the classic two box: bass + mid/tweeter, but on the same width in order it still shape as a rectangle if horizontaly .
Should have clarified: I didn't necessarily mean the 2-box complex shape of the Tarkus, but the performance & component selection. There have been single box versions. Also the Tarkus used a SLS10" -- not 12". Attractive to me because I have 4 SLS10s awaiting...
Last edited:
Look at the faital 10FE330 and 12FE330 also, or the SB34NRX78-6 (for sealed then). They would also fit for this. Those were the drivers i had in mind first when this topic came to life. And there are probally more good drivers to find.
Yup, I was thinking to those Faital. The SB 12.3 is using this Sb34rnx but it needs more than 100l or a linkwitz transform as did @hifijim at around 70 L.
I am not sure the seas 22 taf /g can be horned cause the grill. But the sb26 adc/cdc has a WG made by @augerpro .
If taking the SLS-10" in the Tarkus then 1/3 of the design solved (low pass is 200 hz if I am correct.)
So resting tasks become easier ( measuring the mid plus tweet, then sim with datas given by Paul Carmody for the bass cabinet of the Tarkus...)
https://techtalk.parts-express.com/...-in-the-shop-a-loud-paper-cone-3-way?t=219023
I am not sure the seas 22 taf /g can be horned cause the grill. But the sb26 adc/cdc has a WG made by @augerpro .
If taking the SLS-10" in the Tarkus then 1/3 of the design solved (low pass is 200 hz if I am correct.)
So resting tasks become easier ( measuring the mid plus tweet, then sim with datas given by Paul Carmody for the bass cabinet of the Tarkus...)
https://techtalk.parts-express.com/...-in-the-shop-a-loud-paper-cone-3-way?t=219023
Last edited:
It's the 10" in the Tarkus.
The 12" needs a lot more enclosure volume.
Crossover point in the Tarkus from SLS10 to mid is ~400 Hz. See his FR graph in the crossover section.
I wasn't suggesting to the group abandon the classic 3 way box shape that @andy19191 and others would like.
I was certainly not intended to suggest that a "partybox" is the goal, or to derail the thread.
My sentiment was to suggest that these these "3-way boxes with 12" woofers" of the 1970s have not really gone away; they've just been just changed (for better or worse) in the current era. The JBLpartybox 100 really IS a "12" 3-way" concept, albeit in a much larger cabinet (and packaged with some modern conveniences eg. wireless communication, built in amplifiers).
It uses a few different technologies that weren't available at the time (1970), like waveguides for the tweeter, round-overs for the cabinet by use of a formed/shaped baffle.
May I take a little detour-
Apparently this was the advertising for the JBL L100 back in the 1970s:
This fellow is just BLOWN away by his speakers.
We know that speakers don't sound best placed like this.
In stereo, almost all multi-way speakers benefits from the design axis pointing at the main listening position(s).
It's hard to imagine that speaker designers intended for it to be placed this far below the listener.
But decades, people put their speakers on the floor, for better or worse.
Now here is the re-issued Classic L100 (now in MkII guise), with stands, and the speakers tilted upwards- presumable to get the reference/design axis closer ear level.
50 years ago, No stand was provided, Apart from putting them on the floor, the other thing people did was put them on a console or shelf:
As you can see, there was also no L/R pairing:
The recently updated L100 Classic/Mk2 has a improvement- vertical symmetry of the midrange and tweeter (notice the WG on the tweeter)
This gives this kind of sound radiation into space, measured on the horizontal axis:
Note the darker areas and light areas, in a kind of asymmetric pattern, and somewhat larger "blob" around 1Khz to 5KHz
Should/could this of polar map could be improved upon?
Some years ago I made a 3-way speaker. It's radiation is also, asymmetric:
Here we see the darker colors "hot spots" at various places., The main "blob" around 2-3KHz reduction and discontinuities between ~3KHz and ~6Khz between -30 to -60 degrees, but more output on the +30 to +60 degrees. And note the asymmetry above about 3 KHz:
Here is a clearer view, with demarcations at 3dB, which makes this easier to observe.
Something like this would be may be better- symmetrical and smoother, although the beam width (yellow line ~-8dB) smoother, it is constant from about 3KHz to 22KHz, it's not quite as wide below the Mid-tweeter transition.
So I ask, does the OP we want to replicate the polar patterns of the classic 3-way speakers?
Because @5th element, @DcibeL, @fluid, @HiFiCompass @hifijim, @john k... @kimmosto @PKAudio @vineethkumar01 - knowing what we know today, do we think the performance could be improved if we aimed for a smoother and more symmetrical directivity?
(How steep, and how wide is better (or worse) may be a different chapter)
So I think- Yes! Let's keep with the original proportions of the classic 3 way box. And no, we are NOT building some kind of loud party box.
But I would advocate for some improvements from the 1970's design methodology, such the midrange tweeter to be aligned vertically with the midrange, and a waveguide for the tweeter.. Perhaps a waveguide for the tweeter, or sculpted baffle shape for the midrange and tweeter (3D printing/CNC machined)
It's also hard to see how the positioning of these drivers on the baffle is optimal:
I was certainly not intended to suggest that a "partybox" is the goal, or to derail the thread.
My sentiment was to suggest that these these "3-way boxes with 12" woofers" of the 1970s have not really gone away; they've just been just changed (for better or worse) in the current era. The JBL
It uses a few different technologies that weren't available at the time (1970), like waveguides for the tweeter, round-overs for the cabinet by use of a formed/shaped baffle.
May I take a little detour-
Apparently this was the advertising for the JBL L100 back in the 1970s:
This fellow is just BLOWN away by his speakers.
We know that speakers don't sound best placed like this.
In stereo, almost all multi-way speakers benefits from the design axis pointing at the main listening position(s).
It's hard to imagine that speaker designers intended for it to be placed this far below the listener.
But decades, people put their speakers on the floor, for better or worse.
Now here is the re-issued Classic L100 (now in MkII guise), with stands, and the speakers tilted upwards- presumable to get the reference/design axis closer ear level.
50 years ago, No stand was provided, Apart from putting them on the floor, the other thing people did was put them on a console or shelf:
As you can see, there was also no L/R pairing:
The recently updated L100 Classic/Mk2 has a improvement- vertical symmetry of the midrange and tweeter (notice the WG on the tweeter)
This gives this kind of sound radiation into space, measured on the horizontal axis:
Note the darker areas and light areas, in a kind of asymmetric pattern, and somewhat larger "blob" around 1Khz to 5KHz
Should/could this of polar map could be improved upon?
Some years ago I made a 3-way speaker. It's radiation is also, asymmetric:
Here we see the darker colors "hot spots" at various places., The main "blob" around 2-3KHz reduction and discontinuities between ~3KHz and ~6Khz between -30 to -60 degrees, but more output on the +30 to +60 degrees. And note the asymmetry above about 3 KHz:
Here is a clearer view, with demarcations at 3dB, which makes this easier to observe.
Something like this would be may be better- symmetrical and smoother, although the beam width (yellow line ~-8dB) smoother, it is constant from about 3KHz to 22KHz, it's not quite as wide below the Mid-tweeter transition.
So I ask, does the OP we want to replicate the polar patterns of the classic 3-way speakers?
Because @5th element, @DcibeL, @fluid, @HiFiCompass @hifijim, @john k... @kimmosto @PKAudio @vineethkumar01 - knowing what we know today, do we think the performance could be improved if we aimed for a smoother and more symmetrical directivity?
(How steep, and how wide is better (or worse) may be a different chapter)
So I think- Yes! Let's keep with the original proportions of the classic 3 way box. And no, we are NOT building some kind of loud party box.
But I would advocate for some improvements from the 1970's design methodology, such the midrange tweeter to be aligned vertically with the midrange, and a waveguide for the tweeter.. Perhaps a waveguide for the tweeter, or sculpted baffle shape for the midrange and tweeter (3D printing/CNC machined)
It's also hard to see how the positioning of these drivers on the baffle is optimal:
Last edited:
Tried the ripole from jazzman but with the 4x 10" instead.? Hoever needs a low cut off...around 100 hz ...elepticalish.
Kimmo has recently designed a 3 way coffin for an extremely large monkey the Aalto 9 https://aaltospeakers.fi/en/product/aalto-9/
I think these are the measurements
https://www.audiosciencereview.com/...onstant-linear-directivity.58560/post-2148950
For me the tweeter and mid need to be in a vertical line to avoid a situation where the off-axis lobes become twisted. They can be offset on the baffle to smooth out baffle diffraction, but I don't think baffle diffraction should be prioritised over overall polar response.
I also like the classic 3 way look, and I wrestle with the idea of whether I should give in to the looks and simplicity or design something much more complicated that might make the other side of my brain feel better. I haven't decided which doesn't really matter as I have no time to build it anyway!
I wonder how much an absorptive grill like some of the Wharfedales use could avoid the need for large roundovers.
I already have the drivers but they are possibly even more expensive than the OSMC project.
I remember that advertisement. As I said in another thread... Speakers in the 70's were set on the floor, no one used speaker stands. When listening to the speakers, we sat on the floor so that the tweeter was at ear level... or even laid on the floor between the speakers. Remember, it was very rare for anyone over the age of 35 to have any interest in Hi-Fi. Most audio enthusiasts were pretty young, so sitting cross-legged on the floor seemed very normal. It did not matter if our musical tastes were aligned with Joni Mitchel, Black Sabbath, John Denver, P-Funkadelic, or Led Zeppelin... the listening process was the same.
Yes I agree. And thanks for the post, I enjoyed it !
@fluid,
Yes, and Kimmo's measurement is: (italics- my emphasis)
"Original measurement data is quasi anechoic in free space from speaker with all drivers at vertical center line i.e. center speaker."
This is an outstanding mandrill coffin for sure- huge dynamic range with smooth DI.
more measurements here: https://www.spinorama.org/speakers/Aalto Speakers Aalto 9/Aalto Speakers/index_vendor.html
Thank you shifu
Yes, and Kimmo's measurement is: (italics- my emphasis)
"Original measurement data is quasi anechoic in free space from speaker with all drivers at vertical center line i.e. center speaker."
This is an outstanding mandrill coffin for sure- huge dynamic range with smooth DI.
more measurements here: https://www.spinorama.org/speakers/Aalto Speakers Aalto 9/Aalto Speakers/index_vendor.html
Thank you shifu
Last edited:
The Alto 9 uses a compression driver in horn, that reacts very different than a dome tweeter. It's a different kind of speaker with that. A dome tweeter has a wider dsipertion, except when you limit it with a waveguide, and is less loud in capacity than a compression driver tweeter in horn.
https://aaltospeakers.fi/fi/product/aalto-9/
21" bass, 10" mid, compression horn, dsp, Four Audio amps. Height approx. 120cm, price 29900€/pair
I think that was a Maxwell ad, not JBL. For magnetic recording tape. IIRC.
Wow! I don't recall ever seeing a 21" driver in a single box 3-way hifi spkr. That's ambitious!
Actually Aalto9 was designed by request as a studio main monitor, hifi is secondary and speakers are made on order only. Number of speakers sold is minimal like Taipuu 3-way was (with only 18" woofer). If I understood correctly the comments at a Finnish forum...
Graphics at spinorama are of the proto which had all drivers on midline. Quasi-anechoic and simulation with VCAD2
Graphics at spinorama are of the proto which had all drivers on midline. Quasi-anechoic and simulation with VCAD2
Last edited:
Seems to be a trade off of sensitivity, smoothness and directivity
I was surprised to see that a dome can hit 110dB/m, yet still smooth in response +/- 65° out to 10K Hz.
Distortion vs a compression driver with thanks to @stoneeh :
The goldilocks?