The construction of a multicell horn

I'll try my best to inspire and to show to you that everything is possible. To do a multicellular horn takes a while due to all different moments but they are all kind of nice and it feels good to see the horn grow with every hour of work.

These horns are NOT a Altec 1505B copy. They are pure exponential with a flare frequency of 183 Hz, the same, or at least what was my and reVintages intentions bacause the throat flare in JBL 2441 and 2445.

The horns are built out of 3 mm Baltich Birch plywood, filled with polyester and the outher sides are covered with polyester and glass fibers.

Enjoy:cool:
 

Attachments

  • Multicell horn sheet.zip
    5.1 KB · Views: 2,830
Last edited by a moderator:
  • Like
Reactions: 1 user
Construction of a mold

An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



Was later copied so the mold became symmetrical

An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 

Attachments

  • Horn-5.jpg
    Horn-5.jpg
    32.6 KB · Views: 4,366
  • Horn-4.png
    Horn-4.png
    3.3 KB · Views: 4,303
  • Horn-3.jpg
    Horn-3.jpg
    39.1 KB · Views: 4,352
  • Horn-2.jpg
    Horn-2.jpg
    35.2 KB · Views: 4,321
  • Horn-1.jpg
    Horn-1.jpg
    25.4 KB · Views: 4,333
  • Horn-7.jpg
    Horn-7.jpg
    27.3 KB · Views: 3,276
  • Horn-8.jpg
    Horn-8.jpg
    51.2 KB · Views: 3,102
  • Horn-6.jpg
    Horn-6.jpg
    29.2 KB · Views: 3,384
  • Like
Reactions: 1 user
Good start:)

An externally hosted image should be here but it was not working when we last tested it.



Some help from a friend

An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.



An externally hosted image should be here but it was not working when we last tested it.
 

Attachments

  • Horn-9.jpg
    Horn-9.jpg
    32.6 KB · Views: 2,494
  • Horn-10.jpg
    Horn-10.jpg
    23.4 KB · Views: 2,349
  • Horn-14.jpg
    Horn-14.jpg
    29.6 KB · Views: 2,608
  • Horn-13.jpg
    Horn-13.jpg
    27.7 KB · Views: 2,402
  • Horn-12.jpg
    Horn-12.jpg
    60.2 KB · Views: 2,436
  • Horn-11.jpg
    Horn-11.jpg
    23.5 KB · Views: 2,429
  • Like
Reactions: 1 user
So the cells have zero edge width at their beginning.

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 

Attachments

  • Horn-17.jpg
    Horn-17.jpg
    22.5 KB · Views: 2,027
  • Horn-16.jpg
    Horn-16.jpg
    30.8 KB · Views: 1,859
  • Horn-15.jpg
    Horn-15.jpg
    35.6 KB · Views: 2,050
  • Horn-25.jpg
    Horn-25.jpg
    29.5 KB · Views: 2,332
  • Horn-24.jpg
    Horn-24.jpg
    46.3 KB · Views: 2,186
  • Horn-23.jpg
    Horn-23.jpg
    32.2 KB · Views: 1,935
  • Horn-21.jpg
    Horn-21.jpg
    25.4 KB · Views: 1,839
  • Horn-20.jpg
    Horn-20.jpg
    44.7 KB · Views: 1,929
  • Horn-19.jpg
    Horn-19.jpg
    28.9 KB · Views: 1,981
  • Horn-18.jpg
    Horn-18.jpg
    36.8 KB · Views: 2,437
  • Like
Reactions: 1 user
One side was taped

An externally hosted image should be here but it was not working when we last tested it.


And filled with polyester

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 

Attachments

  • Horn-26.jpg
    Horn-26.jpg
    38.3 KB · Views: 1,769
  • Horn-27.jpg
    Horn-27.jpg
    31.9 KB · Views: 2,013
  • Horn-28.jpg
    Horn-28.jpg
    45 KB · Views: 2,573
  • Horn-31.jpg
    Horn-31.jpg
    21.2 KB · Views: 2,051
  • Horn-30.jpg
    Horn-30.jpg
    43.1 KB · Views: 2,135
  • Horn-29.jpg
    Horn-29.jpg
    47.3 KB · Views: 2,391
  • Like
Reactions: 1 user
Construction of adapter

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 

Attachments

  • Horn-40.jpg
    Horn-40.jpg
    41.8 KB · Views: 1,746
  • Horn-39.jpg
    Horn-39.jpg
    48.2 KB · Views: 1,519
  • Horn-38.jpg
    Horn-38.jpg
    29.1 KB · Views: 1,567
  • Horn-37.jpg
    Horn-37.jpg
    26.7 KB · Views: 1,633
  • Horn-36.jpg
    Horn-36.jpg
    48.8 KB · Views: 1,487
  • Horn-34.jpg
    Horn-34.jpg
    27.5 KB · Views: 1,303
  • Horn-33.jpg
    Horn-33.jpg
    22.3 KB · Views: 1,487
  • Horn-32.jpg
    Horn-32.jpg
    71.6 KB · Views: 1,547
  • Horn-35.jpg
    Horn-35.jpg
    52.7 KB · Views: 1,475
  • Like
Reactions: 1 user
X-mas:cool:

An externally hosted image should be here but it was not working when we last tested it.





Finish!

Sorry for the filter, the layout is just preliminary and will be changed.

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.



The crossover at 500 (12 db electrical) and 3000 Hz (6 dB electrical)

Sounds like heaven:D

If you have any questions/comments please post them and I'll try to answer.



Thanks! / Petter, Sweden
 

Attachments

  • Horn-41.jpg
    Horn-41.jpg
    26 KB · Views: 2,194
  • Horn-42.jpg
    Horn-42.jpg
    31.5 KB · Views: 2,484
  • Horn-43.jpg
    Horn-43.jpg
    37 KB · Views: 2,593
  • Horn-44.jpg
    Horn-44.jpg
    20.7 KB · Views: 2,617
  • Like
Reactions: 1 user
Excellent Work!!!!!!!!

That was a very nice project. Very, very nice.
I was a bit confused on the throat adapter's construction. I may just need to look at it more closely.

Have you had a chance to make any acoustic measures of the horn, especially the polars?

Thank you very much for posting this
 
Very nice work...

If you measure them I'll be curious to see what the real F3 freq works out to be (sans xover), given the mouth size and the flare rate being 183Hz...

I have Altec 1003s, 803s and 203s here to play with... I like that they go to <300Hz. Makes the xover to midbass much better, imho.

One of the real problems with any horn is that there is significant energy there, and the sides of the horn, especially with the thin walls of the multicells (or anything made with thin wood or steel) are not sufficiently stiff... I've had to reinforce the walls of my 203s (for example) which makes a world of difference.

Among the best of the WE and Altec multicells are the ones that are "Tar Filled" - a big hunk of heavy dried asphalt/tar that was moulded around the whole thing... dead and heavy.

So I was wondering what you did about that, if anything with your wood horns. I did notice what looked like some fiberglass over the cells of the completed horn... hard to tell from the pix what it is for?

Btw, they look wonderful!
Nice work...

_-_-bear
 
WithTarragon said:
Excellent Work!!!!!!!!

That was a very nice project. Very, very nice.
I was a bit confused on the throat adapter's construction. I may just need to look at it more closely.

Have you had a chance to make any acoustic measures of the horn, especially the polars?

Thank you very much for posting this



Helmuth said:
Looks great Petter,:smash:

Thanks for sharing your solution how to construct the horn.
:worship:


Thanks :) I thought it would be a good inspierer.

The adapter has the internal throat width and height of 49 mm. I
will upload a schematic of the adapter soon, it will be easier to
understand than me trying to explain.

I have done some measurements on them to make the xover. If
some one can merge different measures to a polat plot I'll be
glad to mail them. The diffucalty to measure them is to choose
where to place the mic, the frequency response change with
both distance and angle. My room is so tight that I can't listen on
axis, so for the cross over design I measured them between the
center cell and the one at it's side.

I have no problem with vibrations what so ever in these horns. If
I knock on them they are dead silent. When all the cells were put
together three of four sides were covered vith fiberglass. Then I
poured polyester from above so all the voids between the cells
were filled with polyester. The horns are filled with about 20 kg
of polyester each. With the JBL 2445 they weigh just above 40
kg a piece. You can lift one by your selve but it is not a nice experience.


Without xover the horn have good output down to about 250-
300 Hz, but due to the variations in the acoustic reactanse and
resistanse I would not cross them lower than 400-500 Hz. this is
something that I found out after the horns were built. My
intention was to use it from 350-400 Hz. Would be intresting to
try them with JBL midrangedriver, because the 2445 is not ment
to be used lower than 500 Hz.

A friend of mine has the Altec 288 and 1505B, his horn are sand filled and that works too, the black kloggy stuff is not too nice to handle.

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.



The small one has a flare of 300 Hz and it is also a pure
exponential horn. The one who built that one is using it together
with JBL 2226 and a RCF tweeter, The small horn goes higher
and a bit more open sound if both horns (183 and 300 Hz) are
crossed high. That's why I chose a quite low xover freq, 3000 Hz
and up to that freq. it sounds really good.

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


View from a multicell, at a friends house:cool:

An externally hosted image should be here but it was not working when we last tested it.


I'll come back to continue...
 
  • Like
Reactions: 1 user
I noticed the clear fill after I posted...

Consider an oil finish on the wood, if you have not yet done anything to it....

The ratings in terms of frequency for the compression drivers are based on two things:

- power handling, or full power bandwidth at the rated power
- physical compliance of the diaphragm's surround

So, if the diaphragm can move below the rated LF cutoff in a linear way, the driver will go down to whatever freq the horn permits. But the manufacturer will only rate the driver to a particular frequency based upon its ability to handle full output, and still remain within xmax. That usually means a higher cutoff than the diaphragm can actually handle, IF the max SPL does not exceed some Xmax figure.

In typical home use that is way way below where it becomes a problem. Consider that most compression drivers (pro ones) are intended to run up to about 128dB SPL, whereas the typical average listening level is <100dB at home... quite a bit of headroom.

Certainly enough headroom for careful testing at minimum.

I use a Japanese driver with a first order filter (one cap) on my 300Hz. horns and have played direct into the system with my band (electric guitars, drums, bass, vocals) in the listening room, uncompressed, at ~100dBSPL average in the space with no ill effects. The JBL and Altecs might not do as well (different diaphragm construction) but worst case you can put a spacer under the mount (frequently they are supplied by the Mfrs to recone houses to center the VC windings in the gap) and gain enough clearance IF there is a slight excursion (Xmax) problem found...

If you say that you have variations below your present xover point, it's likely due to insufficient mouth size, and/or that the horns are foreshortened a bit to be similar to the slightly shorter 500Hz style horns... which is ok, but as I said before I have had a far better listening experience when the horn to midbass xover is lower, not higher, and the horn carries the "voice frequencies" (aka "telephone frequencies") completely - ie. <300Hz and up...

Maybe for your next horn? :D

Regards,
 
Hello Petter,

Very nice work! Congratulations!

For your eyes only , one of the very first picture of Jean Hiraga's new auditorium with a 15 cells Onken horn.


Best regards from Paris,

Jean-Michel Le Cléac'h


Petter Persson said:
X-mas:cool:

If you have any questions/comments please post them and I'll try to answer.
 

Attachments

  • hiraga_system_.jpg
    hiraga_system_.jpg
    96.7 KB · Views: 3,902
Thanks every one for the nice comments!

So, if the diaphragm can move below the rated LF cutoff in a linear way, the driver will go down to whatever freq the horn permits. But the manufacturer will only rate the driver to a particular frequency based upon its ability to handle full output, and still remain within xmax. That usually means a higher cutoff than the diaphragm can actually handle, IF the max SPL does not exceed some Xmax figure. In typical home use that is way way below where it becomes a problem. Consider that most compression drivers (pro ones) are intended to run up to about 128dB SPL, whereas the typical average listening level is <100dB at home... quite a bit of headroom.

I totally agree.

If you say that you have variations below your present xover point, it's likely due to insufficient mouth size, and/or that the horns are foreshortened a bit to be similar to the slightly shorter 500Hz style horns... which is ok, but as I said before I have had a far better listening experience when the horn to midbass xover is lower, not higher, and the horn carries the "voice frequencies" (aka "telephone frequencies") completely - ie. <300Hz and up.

The mouth size is calculated for free space full size, i.e. 2812 cm2 for an exponential flare. The patent rexomended an exponential flare back in the days. The midbass horns has impedance variations that makes them difficult to cross lower than just under 500 Hz, allso 500 Hz makes the components smaller. Got to use two notch circuts. I'll try later with lower xover.

The variations near the cut of freq. depends a lot of the mic position and also, the wave lenght make it difficult to gate the impulse response to avoid room interaction. I have to use a time window of 6-8 ms to gett accurare results at 200-300 Hz and I have to be at quite a distance from the horn mouth to avoid cell-to-cell-interaction/interferrence.

To sum up, the variations can depend on a lot of things, not only to the exponential flare. A Kugelwellen multicellular horn would be kind of cool to try build:cool: due to better coupeling to the free air at the horn mouth.



The cells have got one layer of oil, soon I'll put on an extra layer.

For your eyes only , one of the very first picture of Jean Hiraga's new auditorium with a 15 cells Onken horn.

And every one else;) this Onken, is it simular to the Altec 1505B? The roof horns are awsome! *Want a house...*

I spent two months, on evnings and weekends building, approx. 200 hours I wouls guess. But it is worth it when it's finish! I like to do carpenting too...

Näe, I keep my horns for my self;) Doubt your horns are bad but I like to listen to things I spent blood (yes, cut my self several times in the sharp edges and got a few scars), sweat (all sandpaper and stuff...) and tears (when I listen to it:D ) on building.

Fine workshop details, very informative

Thanks! That was my intension



/Petter
 
Hi Petter,

Congratulations. It's just breathtaking, I saw the pictures yesterday and is still panting now....

One question: I saw it's a curved 'surface' where the multi-cell joint together near the throat. However the adaptor between horn and driver has 2 flat flanges.

How is the conversion between the curvy side and the flat side?


PS. I saw 2 big BLH behind the couch, is that Jensen's Imperial? You guys do have something!