The $1000 3 way design challenge

Counting everything - my labor and the tools excluded of course - my 375L Vintage Monitor cost me circa 600-700€.

  • wood, glue and screws
  • front simili covering and wood oiling
  • custom nameplates
  • inside filling material
  • one set of speakers (2nd hand)
  • a second set of speakers (2nd hand too, for spare)
  • crossover parts
  • wiring and miscellanous (terminals...)

1702849167772.png


1702849414754.png


1702849451263.png


1702849221642.png


1702849386082.png


SPECS :

  • glued-laminated Beech wood cabinet (Hêtre lamellé-collé) 18mm thick.
  • oiled natural satin finish.
  • dark grey tumbled simili leather on front and rear faces.
  • cotton wool full filling/damping.
  • "612SP" 6-12dB serial crossover, LPF = 400Hz, HPF = 5000Hz
  • BEYMA 12B100R Woofer (12", 4" voice coil)
  • BEYMA 8M60N Midrange (8", 1"1/2 voice coil).
  • BEYMA CP21/F "horn-blade" Tweeter (1"1/2 voice coil).
  • V=75L net (88L acoustic with stuffing), tuned at 35Hz.
  • LV=140mm, SV=78.6cm2.
  • n=5.4 (slightly overdamped operation)
  • EFF=93dB/w/m.
  • PWR=40-50WRMS.
  • F3=36Hz, F6=29Hz, F10=25Hz.
  • dimensions : 750x420x360mm.

T
 
I interpret your comment to be be higher sensitivity and higher dynamic range, but slightly less smooth frequency response, for a general MEH?
No less smooth than a direct radiating driver in a flat baffle. An expensive compression driver for an MEH is a BMS 4592 ND. Here is what I get after EQ (psychoacoustic smoothing, first order crossover filters at ~500 Hz and ~6100 Hz):

K-402-MEH (BMS 4592ND) at 1 m On-Axis, Transfer Function.jpg


K-402-MEH (BMS 4592ND) at 1 m On-Axis, Group Delay.jpg


Chris
 
Consider the loudspeakers of post #5 a bit more carefully, or these (a three way design via dual-diaphragm compression driver). Business-as-usual will not likely emerge unscathed after hearing a dialed-in pair of these.

I seem to recall confident statements like this from dipole enthusiasts and wideband enthusiasts but when it came to careful listening tests well designed boring conventional speakers seemed to come out on top. I have not messed about with MEHs which look fun but have some familiarity with very budget pro drivers and what gets sacrificed for high output and low cost.

To be fair, I think the FUD factor is still there vis-à-vis USPTO 8284976 (vs. the original 6411718 patent, now defunct). It shouldn't be, however, since one can easily make a loudspeaker of the older patent type without infringing on the second (still in effect) patent...and still get full performance.

What relevance does patents and patent infringement have to hobby DIY designs?
 
  • Like
Reactions: hifijim
What relevance does patents and patent infringement have to hobby DIY designs?
DIYers are still bound by the same patent laws as commercial entities (US law, as accepted by the EU and elsewhere). If you build even one patented invention for personal use, you're infringing. There is more discussion on this subject on the forums. I'll not clutter up this thread with those discussions. If you wish to infringe, that's up to you.

Chris
 
  • Like
Reactions: ScottG and GM
I seem to recall confident statements like this from dipole enthusiasts and wideband enthusiasts but when it came to careful listening tests well designed boring conventional speakers seemed to come out on top. I have not messed about with MEHs which look fun but have some familiarity with very budget pro drivers and what gets sacrificed for high output and low cost.
If you want to imply that I'm incompetent or a liar, that's also your prerogative. I will refrain from such accusations, however. I've heard both loudspeaker types (really good ones in the case of direct radiators and planars). I don't say what I have said lightly or flippantly. You should listen to some good DIY MEHs.

Chris
 
  • Like
Reactions: GM
@andy19191 @Cask05
I think you guys are confused or overcomplicating the OPs simple request. See above for thread topic. We'd love to see your driver and alignment choices. @andy19191 you even said above that you looked up some drivers that might work well together for $800 but didn't list them.

My point was about prices. I was not presenting a design only variations around the conventional configuration. Not sure about overcomplicated but I agree there seems to be some confusion related to what design challenge might mean. I was expecting some discussion around a fuller spec or objectives but others seem to be posting any 3 way speaker with a budget of around $1000.
 
DIYers are still bound by the same patent laws as commercial entities (US law, as accepted by the EU and elsewhere). If you build even one patented invention for personal use, you're infringing. There is more discussion on this subject on the forums. I'll not clutter up this thread with those discussions. If you wish to infringe, that's up to you.

Patents are issued to prevent commercial exploitation by others. If there is no commercial exploitation a patent is irrelevant.
 
If you want to imply that I'm incompetent or a liar, that's also your prerogative. I will refrain from such accusations, however. I've heard both loudspeaker types (really good ones in the case of direct radiators and planars). I don't say what I have said lightly or flippantly. You should listen to some good DIY MEHs.

Not sure there is good way to reply to a post like this that would keep a civil chat on MEHs going.
 
No less smooth than a direct radiating driver in a flat baffle. An expensive compression driver for an MEH is a BMS 4592 ND. Here is what I get after EQ (psychoacoustic smoothing, first order crossover filters at ~500 Hz and ~6100 Hz):

View attachment 1248022

View attachment 1248024

Chris

That's not bad. I'll have to try a MEH at some point.

Once I did try measuring my direct radiating system; with psychoacoustical smoothing.
But I agree- they do need careful selection of very high quality, well design (but not boutique) drivers.
My target to overshoot the excellent JBL 4367, but in more modest proportions that suited my room.
I aimed for 93dB/2.83V sensitivity.
But to achieve a smooth response, I had to go three way and DSP/tri-amped.

1702851504489.png



Here's an unsmoothed measurement

1702855795087.png


One of rare opportunity to do distortion in the dead of the night; noise floor is ~35dB

1702855652164.png



CTA2034A:
1702854196501.png


1702853399261.png





With thanks to @bikinpunk I was able to compare JBL's excellent 4367:
1702856429815.png


Now Erin's NFS shows all the lumps and bumps, warts an all, because it has a measurement resolution of 2Hz. My measurement resolution is much more, about 200Hz. But I've correlated them with near field measurements to ensure that the drivers don't have any hidden resonances that are not seen in the farfield.

Reference:
https://www.erinsaudiocorner.com/loudspeakers/jbl_4367/

Horizontal 0-90, in 10 degree steps.

1702856500268.png


IMG_2483.jpegIMG_2484.jpeg


It's unfortunate that they drivers are NLA or difficult to source. (MOQ 300 according to Peerless/Tymphany for the XXLS 12" 8 ohm woofers and B&W spare parts only shipped only to USA) (LF + MF cost ~USD $200 ea, HF driver US$50) otherwise I would document the design for others who may want a conventional direct radiating, high senstivity, high SPL system of moderate proportions (2 box cabinet 47 1/4" high x 13 3/4" wide x 15 3/4" deep)

Even further improvement could be done with facets or large round-overs close to the tweeter edges, which was my original intent, but my cabinet maker was unable to organize.

All responses above are heavily DSP optimized and a on-off build, but I took care to ensure that the design used drivers with smooth responses to that it is convertible to passive, which I have intentions to do another time, particularly in the MF/HF part.

But it's not far to compare a heavily tuned one-off speaker to a speaker put into In mass production.
Any passive system would deviate by at least another +/- 1.5 dB is my rough estimate.

I'm curious about the MEH, but the only good measurement I've seen is from Erin. I wasn't surprised at the response- some compromises are needed to get that level high level of sensitivity (95-100dB/2.83V), so of course the anechoic response is not as smooth as a direct radiating system. But I am curious about how it sounds in a room, which as you mentioned, may not sound as bad as it looks anechoically.

1702854343224.png


Reference: https://www.erinsaudiocorner.com/loudspeakers/danley_sh50/
 
Last edited:
Let me violate one prescription - 3 way.
1 each 2 way 54-20000 hz less wood & fixtures $492.37 two for $984.67
1 sub 20-54 hz $464
BOM
2 way 54-20000 hz BOM
2 ea Eminence woofer Deltapro-15a $155 ea $310
2 ea Eminence CD N314-T8 $188 $376
2 ea Eminence horn H14ea $ 73.43 $146.87

1200 hz 12 db butterworth
4 ea janzen 1.4 mf 18 ga $ 25.14 $ 50.28
4 ea solen 12 uf 400 v poly $ 17.80 $ 35.60
10ea resistor 1 ohm 10 watt $ 5.00 $10.00
4 ea dual binder jacks $ 8.00 $ 16.00
2 ea speakon jacks $ 9.00 $ 18.00
2 ea tungsten limiter $11.00 $ 22.00
Total $492.37 $984.67

20-54 hz eminence lab 15 1 ea $335.00 $335.00
Speakon jack $ 9.00 $ 9.00
crossover sub NADY used $ 50.00 $ 50.00
Amp 150 w Peavey MMA-8150T $ 70.00 $ 70.00
sub 20-50 hz $464.00 $464.00
 
Last edited:
Interesting.

When you say psychoacoustic, can you expand on this?

I interpret your comment to be be higher sensitivity and higher dynamic range, but slightly less smooth frequency response, for a general MEH?

eg. +10dB sensitivity, frequency response not as smooth but equivalent when smoothed to an to accuracy of 1/3 octave below 100Hz, 1/3 octave to 1/6 octave between 100 Hz and 1 kHz, and 1/6 octave above 1 kHz, with more weighting to peaks.


@mayhem13

I like this game. But I also like the- what can be done for

Bar of soap size (Google Nest mini)
0.125 cu ft (Wonderboom 3)
0.25 cu ft (~LS3/5A size)
0.5 cu ft
1 cu ft.
2 cu ft
4 cu ft.
Etc.

I think that is the real challenge.

As they get smaller and smaller I think plastic (or 3D printed) cabinets, and DSP and class D are a must. The transducer is important but is of secondary importance.
For instane the Wonderboom uses dual opposed 2” twoofers for vibration cancelling, and dual opposed passive radiators orthogonal to the twoofers. Insane performance for a 1 litre external enclosure.

The “ high end “ want boutique materials and artisan craftsmanship and an extra zero to the price. But that doesn’t mean much in terms of sonics.
YES!…..you understand perfectly. Small compact systems that don’t intrude on valuable floor and storage space or don’t call attention to themselves are on the commercial forefront. That being said, you could build the most resolute and value driven ‘MEH’ in the history of MEH’s And regardless, these won’t find a place in many homes…..single middle aged and later vintage men without constraint.
 
Let me violate one prescription - 3 way.
1 each 2 way 54-20000 hz less wood & fixtures $492.37 two for $984.67
1 sub 20-54 hz $464
BOM
2 way 54-20000 hz BOM
2 ea Eminence woofer Deltapro-15a $155 ea $310
2 ea Eminence CD N314-T8 $188 $376
2 ea Eminence horn H14ea $ 73.43 $146.87

1200 hz 12 db butterworth
4 ea janzen 1.4 mf 18 ga $ 25.14 $ 50.28
4 ea solen 12 uf 400 v poly $ 17.80 $ 35.60
10ea resistor 1 ohm 10 watt $ 5.00 $10.00
4 ea dual binder jacks $ 8.00 $ 16.00
2 ea speakon jacks $ 9.00 $ 18.00
2 ea tungsten limiter $11.00 $ 22.00
Total $492.37 $984.67

20-54 hz eminence lab 15 1 ea $335.00 $335.00
Speakon jack $ 9.00 $ 9.00
crossover sub NADY used $ 50.00 $ 50.00
Amp 150 w Peavey MMA-8150T $ 70.00 $ 70.00
sub 20-50 hz $464.00 $464.00
I didn’t require hardware or cabinet materials in the budget constraint..only drivers and XO components
 
The cost for the drivers of the Open Source Monkey Coffin is just below USD 1000 (per speaker, not the pair). At least it was last time I looked. Exchange rates and everything else in the world have gone wild since then, so depending on where and how you look at it the drivers may be cheaper or more expensive now.

That said, I don't see how it makes sense to put a cost limit on the drivers and then ask for suggestions on parts. You'd need a lot more detail on the design goals to come up with a design that will make sense. With the Monkey Coffin we had a long list of details on the design goals to start with, and this was still way too crude to avoid many long discussions on our target down the road.
 
DIYers are still bound by the same patent laws as commercial entities (US law, as accepted by the EU and elsewhere). If you build even one patented invention for personal use, you're infringing. There is more discussion on this subject on the forums. I'll not clutter up this thread with those discussions. If you wish to infringe, that's up to you.

Chris

Wrong. The law in France allow you to do this for your personal use without any infringement. And not only in France : Italy, Germany, Spain, among others.

T