Hypercube Loudspeakers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
WF, how did building go?

Please keep in mind that you will have to deal with the rising response of the fullrange driver towards mid to high frequencies. That means that without a frequency correction circuit you will get very bass shy and sometimes annoying frequency response.

A simple circuit in front of the speaker consisting of a coil in series with the speaker and a resistor in parallel with the coil will most of the time take care of this.

Good values to start with are 1mH for the coil and 8R for the resistor. Then variations on the resistor will determine if you get more or less high frequency level, as the resistor is a bypass for the low-pass effect of the coil.
 
WF, how did building go?

Please keep in mind that you will have to deal with the rising response of the fullrange driver towards mid to high frequencies. That means that without a frequency correction circuit you will get very bass shy and sometimes annoying frequency response.

A simple circuit in front of the speaker consisting of a coil in series with the speaker and a resistor in parallel with the coil will most of the time take care of this.

Good values to start with are 1mH for the coil and 8R for the resistor. Then variations on the resistor will determine if you get more or less high frequency level, as the resistor is a bypass for the low-pass effect of the coil.

Hi, diyAchim,

thanks for asking. Actually I have been running after some stuff I had accumulated, I just bought the wood, I will try to make the patterns tomorrow and rough cut it, and I will finish and assemble them later in Belgium.

I decided 3 sizes, I will start with the medium one, ~45 l. for the two BG20 drivers I have. I will do one ~70 l. for the marimbula, and 4 ~31 l. for the BG17 ones that I have lying around, getting dust.

The problem is that I live between two places, and I only have the good machines here, in France, where I spend the less amount of time (arghhh!!!). Also, the two places are a 5 hour drive away, so I try to make the most of my time here (and it always seems to fly!).

Thanks for the suggestion for the passive circuit, it seems so simple that even I could understand it. I could try to use the eq. of the instruments and amps, etc., but it could be a nice feature to have some sound shaping built in.

Did you call the Swiss for the wood? My guess is that if you do not hurry, you better wait until next summer, I wouldn't like to go there when it's all covered in snow and ice, I barely have the courage to get there in summertime, and then I keep my fingers crossed I will not run into someone and then have to drive a couple of kilometers in reverse in those winding mountain roads. Although last time I was there they were building some ramps here and there so two cars can cross. But I was born in the plains, I get dizzy very quickly driving over there.

Cheers,

WF
 
OK, I just finished pre-cutting 3 hypercube kits.

The hardest part is to make a really nice and precise pattern or jig, that I have done in maple, one for the rhombi and triangles, plus another for the square.

Then I roughly cut 3 sets of the 13 pieces needed, the 8 rhombi, 4 triangles, and 1 square.

The plan is to make 2 enclosures and a marimbula out of that. For the marimbula, the square face instead of being the baffle will be the 'soundboard'.

I will be sending some pictures. I am leaving for the other place, where I don't have a table saw, but I will finish the assembly with a router, and some hand-tools. I am hoping that the rest is easier, I figure once you get the pattern right, it will be fast to trace it with a router. I have the three kind of bits, the only problem is that they are not long enough (the wood i bought is too thick), but I will try to arrange that by thinning down the boards near the edges.

If the thing sounds convincing, next time I will upgrade the driver and the wood quality.

Cheers,

WF
 
Hi WF,

looking forward to your pictures.

I did not contact the tonewood people as I am waiting for my proper hypercubes in 9mm plywood that will be here next week. I will then compare my thinner 5mm prototype to the 9mm ones, and from there hope to conclude if a more resonant tonewood type is preferable or not.

One remark: On my prototype that I had glued up with hide glue, one connection cracked open when I screwed the loudspeaker to the baffle. Being 5mm the wood was obviously a little bent into shape when taping and glueing, and the drawing flat the baffle when screwing on the speaker created too much tension for the glue seal. So I think that glue with more strength = standard carpenters glue is better in order to avoid such cracking. Titebond states that their hide glue is not recommended for structural work that has static function.
 
I may add that yesterday I listended to a Neil Cowley Trio recording on my single prototype hypercube - and, in addition to the good overall performance, the low piano notes especially have a vibrancy and presence that I have never had before in my system...

That's nice diyAchim, I am also eager to test my three hypers, particularly the one that will function as a musical instrument.

I used quite cheapo wood, not-quarter sawn, and with many knots and imperfections, but it's a start. Also, it is quite soft pine, so my guess is that it is somewhat absorbent for sound.

In any case, I like the ring of porous softer woods much more than hardwoods for musical instruments. I have a Japanese 1980 Ibanez P-Bass, with a thick maple neck, and a body of ash, it sounds really wonderful, I am in love with it.

I don't understand why the Tesserax guy stated it was too difficult to build them, with a proper setup, the right tools and patterns, you can build them at a cost very similar to a typical cubical enclosure. If you are even more serious you can even envision some kind of mounting block that ensures the pieces are all glued at the right angles. It is not that hard to make, not now and not back then.

You could produce the enclosures up to quite a big size with only two planks of 200 by 40 cm, which can be bought quite cheap (total 28 euros per enclosure). Even if you were to make them with the best tonewood money can buy, it would be at around 120-150 euros a piece, for thick ones, and half of that for really thin ones. Frankly, if they sound so good as everyone here is saying, you could mount some setups to produce them for quite a reasonable fee as kits, or already assembled.

I think that the key is the quality of the wood (highest quality, quarter sawn wood shrinks and expands quite uniformally, and doesn't twist) if you want to make then really thin. My guess is that your unit cracked precisely because you had to force it in place, if the fit is perfect, and the wood is fine, there is no reason why it should happen.

Take a violin, for instance, the gluing surface is quite minimal, and it works really good (of course the shape contributes to the structural strength, but nonetheless, those instruments are quite sturdy).

I have quite thick wood, so I can use hide glue, there is a lot of contact surface.

I am glad you liked the prototype. In any case, I think this is the right shape of enclosure for me, since I like the idea that the case gives some color or ring to the sound, and it's not a dead thing.

By the way, it's quite off-topic, but wood is a "hi-tec material that just happens to grow as trees". I have seen the phrase in a presentation about violins, check it out, it is really interesting:

YouTube

Cheers,

WF
 
Thanks for the video link, I will watch this.

The "Tesserax guy", respectfully, I think is the wrong wording ;-)

I am very greatful for Matthew and his colleague to have made their finding available to us and still investing time in making some waves after having hit the wall when they tried it commercially. Kudos!

I was thinking along the same lines as you, though, when after the prototype I pondered how a larger production of these would be viable.

Like you I think that with proper templates for the angles/sizes of the pieces, exact degree router bits for the edges and a scaffolding/mould for putting together the pieces in 3D space production should be quite smooth. It's doable with the taped net plan, but you are fighting gravity once the pieces get heavier than the 5mm ones.

As to "colouring the tone": Look forward to what you will hear and say. My perception is that there is NO colouration, but just added energy. For example, it is wrong to assume that the HC plays nice piano, but gives voices or cymbals a wooden varnish. Nothing of that sort happens. This may differ with build, and be a function of the driver and wood parameters.
 
Last edited:
Thanks for the video link, I will watch this.

The "Tesserax guy", respectfully, I think is the wrong wording ;-)

I am very greatful for Matthew and his colleague to have made their finding available to us and still investing time in making some waves after having hit the wall when they tried it commercially. Kudos!

I was thinking along the same lines as you, though, when after the prototype I pondered how a larger production of these would be viable.

Like you I think that with proper templates for the angles/sizes of the pieces, exact degree router bits for the edges and a scaffolding/mould for putting together the pieces in 3D space production should be quite smooth. It's doable with the taped net plan, but you are fighting gravity once the pieces get heavier than the 5mm ones.

As to "colouring the tone": Look forward to what you will hear and say. My perception is that there is NO colouration, but just added energy. For example, it is wrong to assume that the HC plays nice piano, but gives voices or cymbals a wooden varnish. Nothing of that sort happens. This may differ with build, and be a function of the driver and wood parameters.

Hi diyAchim, sorry about my unfortunate wording, I don't mean to be disrespectful, so no offense, please! I know how many things don't go right even when they should, in the 'invention' business. As a matter of fact I witnessed very closely how some people developed a cure for a STD for which there was no specific treatment, and the whole thing will follow the same road, the patent will expire, since they didn't have enough founding to do a systemic test, and the topical version did not reach the targeted virus.

I agree that we owe to them a lot, I am just saying that I do not fully understand why the commercial production of those enclosures was not viable, since as you rightly say, it is not particularly hard to do if you have the proper tools and patterns (by the way, the gravity issue you mention could be easily dealt with by doing partial assemblies, maybe in 3 parts, on top of your mold, and then one final go when instead of dealing with 13 pieces, you only glue together 2 or 3).

Anyway, I am also very curious to try them, but as you might have gathered from my posts, I am not from a hi-fi background, so I don't have a clue of how many-thousand-dollars speakers sound like. But if you give me a hand, and explain me how to do it, I would gladly supply some experimental data once they are assembled.

I will also try to document how it is done, since I have the feeling it might help some other builders, as it was my case when gmad explained me the right angles to construct the truncated rhombic dodecaedron.

By the way, regarding the volume formula, once you realize that this is in fact a cube that has it's volume divided in 6 pyramids, and turned around and added to its volume, it's easy to calculate the enclosure volume, you take the square of the truncated side, you calculate the volume of a cube with it, and you multiply it by 7/4, and that's all.

Cheers,

WF
 
Hi WF,

well understood about your/my thoughts on "production should be easy". Nevertheless, we weren't there... (to quote New Model Army). We will know more after having tried ourselves.

Assembly piece by piece: This is what I meant. Having a mould, and the square baffle at the bottom, would allow one piece at a time added to the baffle, building bottom to top. Gravity would actually work with the assembler... And after having added the last pieces of the back, a sandbag weight would give the pressure needed for curing the glue.

Please PM me about your questions re loudspeakers. I should be able to help you, but details would probably be of topic to this thread.
 
Very important:

As I am just finding out as I start using the new 9mm hypercubes it is very, very important to ascertain airtight mounting of the speaker as well as airtight sealing of any vias for cabling.

I was a bit sloppy with that on the prototype, so will retake the impedance measurements after having set everything straight.
 
Very important:

As I am just finding out as I start using the new 9mm hypercubes it is very, very important to ascertain airtight mounting of the speaker as well as airtight sealing of any vias for cabling.

I was a bit sloppy with that on the prototype, so will retake the impedance measurements after having set everything straight.


Just out of curiosity, where do you have feed the cable to the speaker? I was thinking on using regular jack as it is done usually with guitar and bass cabinets.
 
Founder of XSA-Labs
Joined 2012
Paid Member
Typical banana plug screw terminal binding posts use a standard 1/4in hole mount and can be airtight if you add a bit silicone RTV or other sealer to the hole before inserting and tightening. Most terminal cups or dual binding posts on a plastic base have a sponge rubber airtight gasket. It’s not a problem really. With some that use a smaller interference fit hole, you don’t even need sealer as it is very tight and requires great pressure to install the fluted shaft into the wood.

s-l300.jpg
 
Last edited:
Typical banana plug screw terminal binding posts use a standard 1/4in hole mount and can be airtight if you add a bit silicone RTV or other sealer to the hole before inserting and tightening. Most terminal cups or dual binding posts on a plastic base have a sponge rubber airtight gasket. It’s not a problem really. With some that use a smaller interference fit hole, you don’t even need sealer as it is very tight and requires great pressure to install the fluted shaft into the wood.

s-l300.jpg


Thanx, that's very helpful. But I probably just install the jack in an outer plate and then run the cables inside through a hole and then seal that with something. All or most amps for instruments have jack outputs, so I rather stick to the norm.
 
New measurements

So with the new 9mm hypercubes in the house, and everything now sealed airtight, I made new measurements.

I have to apologize for spreading false wisdom with my first measurements, which seem partly false now.

What stands is that the impedance peak is in fact more akin to a box volume larger than the actual hypercube volume.

What proved false was the thought that the impedance is actually lowered by the HC.

I measured open baffle, HC 5mm and HC 9mm. I also measured frequency response. It shows that HC5 and HC9 do not differ much, except for small resonance peak differences which are also indicated in the impedance plots.

The 5mm HC has more peaks but they are smoother.

In practice the thinner wood HC is more to my liking. It does not show more coloration, but still sounds a tad livelier. Both are without any stuffing.

Compared to open baffle (both equalised for similar response) the hypercube provides more energy and presence. It's 100 Hz peak needs some ironing or proper placement in the room. The open baffle is too laid back for my liking.

With the Visaton B200 the hypercubes need a 6dB cut at 1024 Hz with a Q of 0.5 - so roughly equalling the application of a 1mH or 1.5mH coil with appropriate bypass resistor to ones taste. I currently eq in the digital domain.

The hypercubes will need a proper stand and tilt angle needs to be determined as well as placement in the room.

The way the hypercubes energize the room is indeed special.
 

Attachments

  • impedance_ob_hc5mm_hc9mm_B200.png
    impedance_ob_hc5mm_hc9mm_B200.png
    46.9 KB · Views: 282
  • near_plus_100cm.png
    near_plus_100cm.png
    112.9 KB · Views: 148
  • splnear.png
    splnear.png
    49.6 KB · Views: 167
  • WP_20181025_003.jpg
    WP_20181025_003.jpg
    778.4 KB · Views: 173
  • WP_20181025_001.jpg
    WP_20181025_001.jpg
    751.3 KB · Views: 153
  • impedance_hc5mm_hc9mm_B200.png
    impedance_hc5mm_hc9mm_B200.png
    46.5 KB · Views: 75
  • impedance_ob_hc9mm_B200.png
    impedance_ob_hc9mm_B200.png
    46.6 KB · Views: 279
  • impedance_ob_hc5mm_B200.png
    impedance_ob_hc5mm_B200.png
    46.6 KB · Views: 292
Has anyone tried putting drivers on three of the square surfaces of the HC?
It could make a rather nice semi omni.

It would be placeable on the floor due to the upwards facing drivers, giving the cone of quality sound as much relevant space as possible.
It would always a give more emphasis on the reflected sound, which is good since reflection attenuates a lot.
 
...I made new measurements...

Holy crap this look be data of great worth and value you share thanks for that, each diy'er probably have their target and taste somehow cut out and was it me HC looks should as soon as possible be seriously single/double or trible material stuffed somehow to get rid of those impedance peaks and resonances, but again great thanks show those honest graphs and by the way those handmade builds look great and same does row of vinyl colletion :)
 
Last edited:
Made another pair of these. This pair uses Tang Band W8-1808 8" neodymium full range drivers. These came out about twice as big as my first pair. They are about 19" tall/wide. My first set needed a sub to fill in the lows. The goal was to reduce that need. They still need some sub but not as much. Theoretically the f3 should be at 72 Hz vs. 84 Hz for the smaller pair that used Mark Audio CHR-70A 4". These are 3/4 MDF. Next set if I make any will be plywood. Since starting these I've concluded that plywood is stiffer for the thickness/weight and doesn't produce as harmful dust as MDF. You can see my first set behind the new one in the last picture.



Thanks.
 

Attachments

  • IMG_9238.JPG
    IMG_9238.JPG
    94.4 KB · Views: 100
  • IMG_8136.JPG
    IMG_8136.JPG
    64.2 KB · Views: 83
  • 57024049328__8CF29E35-3C5D-4262-82EA-2645E43E3E75.JPG
    57024049328__8CF29E35-3C5D-4262-82EA-2645E43E3E75.JPG
    96.6 KB · Views: 78
  • IMG_7558.JPG
    IMG_7558.JPG
    127.7 KB · Views: 79
  • IMG_7557.JPG
    IMG_7557.JPG
    79.1 KB · Views: 70
  • IMG_7556.JPG
    IMG_7556.JPG
    85.6 KB · Views: 81
  • IMG_7555.JPG
    IMG_7555.JPG
    91 KB · Views: 85
  • IMG_7554.JPG
    IMG_7554.JPG
    87.2 KB · Views: 82
  • 55979780676__0665E7CE-4BB4-4BD8-ABAF-E1B1E512B303.JPG
    55979780676__0665E7CE-4BB4-4BD8-ABAF-E1B1E512B303.JPG
    158.8 KB · Views: 86
  • 58614560399__1821A31F-BCF7-41B4-BC97-8786BA44E078.JPG
    58614560399__1821A31F-BCF7-41B4-BC97-8786BA44E078.JPG
    72.8 KB · Views: 105
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.