Another Homemade Exponential Horn Project

[Ethan]

My first paper appliqué horn has always seemed rather bass-weak, so I decided to try my hand at making something a little bigger: an EMG Xb Oversize copycat (sort of). Based on length, mouth, and throat measurements of Xa’s, Xb’s, and Xb Oversizes elsewhere on the forum, I believe this design should be about as big as an Oversize; it has a ~34” diameter mouth and is ~86.5” long from the throat to the mouth, with an 80 Hz flare. The “neck” isn’t bent quite the same as a Xb’s, mostly to reduce the number of different sections and make the construction less complicated.
I originally planned to use heavy-duty paper and Elmer’s Glue-All to make the horn, but instead switched to the same packing paper I used for the last one (more about that below); for ease of construction, it's a multi-piece design, like the last horn, except with the bell divided roughly in half in order to get the 34” mouth through my 29” door.
The bell portion of the former uses the same basic construction method as for the last horn, but it is split into two pieces, with wood discs on the ends, to provide a hard surface to cut against while trimming excess paper.
I had to dissolve the curved Styrofoam “neck” former out of the last horn, so for this one, the former has a wood skeleton filled with expansion foam and paper mâché—hopefully, it will release with some sharp taps from a hammer, instead of just crushing at the hammered end as the last one did. I also built a handle into the large of end of the former for the second (longer) bend, to help pull it out—we shall see how well it all works in practice; hopefully, the former does actually release, as it can’t be dissolved out like the last one.
I had initially planned to sand the expansion foam to the exponential flare and finish it with shellac, but within a few days after sanding, the foam had collapsed by up to half an inch, and after a few more days, it had begun to expand again, beyond the intended surface, so I sanded it all down to at least half an inch below the wood ribs, filled in with paper mâché paste, and gave it several coats of shellac to give a smooth surface and ensure that water from the horn couldn’t rehydrate the paper in the former. (I considered using general-purpose Bondo instead of paper mâché, but paper mâché is much less potentially hazardous.)
Photos below—

Here is the main design page from my notebook—I’m making the lower of the two horns (the B-43a-2); the only difference is that it has a 60° bend around a ~10.5” radius followed by a 150° bend around a 12” radius, instead of two 60° bends around ~11.25” radii (in opposite directions), followed by a 90° bend around a 12” radius, meaning that it can be made in two pieces instead of three.

The "Neck" Former--

The wood skeleton cut out and sanded to a circular cross-section:

The ribs glued to the two halves of the skeleton of the first curve:

The halves of the skeleton glued together:

The first side of the former filled with expansion foam, after it started to collapse:

Detail of collapsed foam on the short curve:

Detail of collapsed foam on the large curve. The expansion foam pushed the end half-disc out, so I had to cut out some foam and reglue the wood, hence the gap:

The second side of the former filled with expansion foam:

The first side after a few more days—more collapsing of the foam:

As the expansion foam wasn’t working quite as anticipated, I sanded it down and filled in with paper mâché. The first application of paper mâché (paste) on the first half of the former wasn’t quite thick enough, so I had to do another after it dried; this is the first layer:

The first side after the second layer of paper mâché dried, before sanding (apparently, I didn’t get a picture right after sanding):

The finished former for the first curve. I sanded the paper mâché with 220 grit sandpaper after using an assortment of files and rasps to remove the excess material, then gave it two coats of shellac, sanded with 220 grit paper, and repeated the shellac-and-sanding process with 400, 600, and 800 grit paper.

The finished former for the second curve. The surface isn’t perfectly even, but it should be good enough--the orange areas are where I sanded the paper mâché too far and filled in with wood filler. Unfortunately, the paper mâché pulled in one of the wood half-discs on the end that the expansion foam had pushed out, but it is only about an ⅛” off at the most, so I left it alone, as I should be able to fill in the missing length on the horn with paper mâché paste:

The Bell Former--

The Styrofoam rings for the bell former, before gluing and sanding:

The formers for the two halves of the bell (prior to sanding) with the wood end discs glued and clamped in place:

The latch to keep the two halves of the bell together while sanding (to ensure a smooth flare):

The formers for the two halves of the bell after sanding:

The formers for the two halves of the bell after three coats of shellac (latched together):

The Horn--

The horn spigot—1.5” outer diameter DOM steel tubing (1 ⅜” inner diameter):

I bought a large roll of thick paper to make the horn, and finished two layers (overlapping paper strips about 50%) on the first section of former, but the layers didn’t laminate, I think because the paper was too thick and stiff for me to press it in place properly; so I had to tear it all off and start again with the same thinner paper I used for the last horn—so far, I have three layers done (overlapping about 75%), and the paper is laminating properly now.

The first section of the horn with two layers of the thick paper:

Detail of delamination, after cutting some paper away—I could feel that the overlapped paper was delaminated for most of the section’s length, which was confirmed by cutting some of the outer layers away:

The first section of the horn after tearing off the thick paper and applying a layer of thinner paper:

The same section after two subsequent applications of thinner paper (three applications total)—it took about three hours to do the third application, although I think I was overlapping somewhat more than 75%:

Unfortunately, the paper shrank a bit as it dried, so I need to add another partial layer at the large end to overlap with the next section of former by about ½", in order to align the the first section of the horn with the second section of the former:

More to come over the next few weeks!--I'm hoping to be able to test the horn before I go back to college in about three weeks, although I doubt I will be able to get the full ¼" thickness for the neck by then.

[poodling around]

My first paper appliqué horn has always seemed rather bass-weak, so I decided to try my hand at making something a little bigger: an EMG Xb Oversize copycat (sort of). Based on length, mouth, and throat measurements of Xa’s, Xb’s, and Xb Oversizes elsewhere on the forum, I believe this design should be about as big as an Oversize; it has a ~34” diameter mouth and is ~86.5” long from the throat to the mouth, with an 80 Hz flare. The “neck” isn’t bent quite the same as a Xb’s, mostly to reduce the number of different sections and make the construction less complicated.
I originally planned to use heavy-duty paper and Elmer’s Glue-All to make the horn, but instead switched to the same packing paper I used for the last one (more about that below); for ease of construction, it's a multi-piece design, like the last horn, except with the bell divided roughly in half in order to get the 34” mouth through my 29” door.
The bell portion of the former uses the same basic construction method as for the last horn, but it is split into two pieces, with wood discs on the ends, to provide a hard surface to cut against while trimming excess paper.
I had to dissolve the curved Styrofoam “neck” former out of the last horn, so for this one, the former has a wood skeleton filled with expansion foam and paper mâché—hopefully, it will release with some sharp taps from a hammer, instead of just crushing at the hammered end as the last one did. I also built a handle into the large of end of the former for the second (longer) bend, to help pull it out—we shall see how well it all works in practice; hopefully, the former does actually release, as it can’t be dissolved out like the last one.
I had initially planned to sand the expansion foam to the exponential flare and finish it with shellac, but within a few days after sanding, the foam had collapsed by up to half an inch, and after a few more days, it had begun to expand again, beyond the intended surface, so I sanded it all down to at least half an inch below the wood ribs, filled in with paper mâché paste, and gave it several coats of shellac to give a smooth surface and ensure that water from the horn couldn’t rehydrate the paper in the former. (I considered using general-purpose Bondo instead of paper mâché, but paper mâché is much less potentially hazardous.)
Photos below—

Here is the main design page from my notebook—I’m making the lower of the two horns (the B-43a-2); the only difference is that it has a 60° bend around a ~10.5” radius followed by a 150° bend around a 12” radius, instead of two 60° bends around ~11.25” radii (in opposite directions), followed by a 90° bend around a 12” radius, meaning that it can be made in two pieces instead of three.
B-43 Notebook Page.jpg

The "Neck" Former--

The wood skeleton cut out and sanded to a circular cross-section:
Former--A--backbone and half-discs prior to gluing.jpg

The ribs glued to the two halves of the skeleton of the first curve:
Former--B--60-degree bend with half-discs glued on.jpg

The halves of the skeleton glued together:
Former--C--Both halves glued together; support blocks in place.jpg

The first side of the former filled with expansion foam, after it started to collapse:
Former--D--Expansion foam on first side (after sanding, some collapsing).jpg

Detail of collapsed foam on the short curve:
Former--D--large end of 60-degree curve--collapsed foam.jpg

Detail of collapsed foam on the large curve. The expansion foam pushed the end half-disc out, so I had to cut out some foam and reglue the wood, hence the gap:
Former--D--large end of 150-degree curve--collapsed foam.jpg

The second side of the former filled with expansion foam:
Former--E--Expansion foam on second half, after sanding, some collapsing.jpg

The first side after a few more days—more collapsing of the foam:
Former--E--Collapsed foam on first half (after several days).jpg

As the expansion foam wasn’t working quite as anticipated, I sanded it down and filled in with paper mâché. The first application of paper mâché (paste) on the first half of the former wasn’t quite thick enough, so I had to do another after it dried; this is the first layer:
Former--F 1--Sanded paper mâché.jpeg

The first side after the second layer of paper mâché dried, before sanding (apparently, I didn’t get a picture right after sanding):
Former--F--Right Side with Unsanded Paper Mache.jpg

The finished former for the first curve. I sanded the paper mâché with 220 grit sandpaper after using an assortment of files and rasps to remove the excess material, then gave it two coats of shellac, sanded with 220 grit paper, and repeated the shellac-and-sanding process with 400, 600, and 800 grit paper.
Former--L--60-degree curve, shellacked.jpeg

The finished former for the second curve. The surface isn’t perfectly even, but it should be good enough--the orange areas are where I sanded the paper mâché too far and filled in with wood filler. Unfortunately, the paper mâché pulled in one of the wood half-discs on the end that the expansion foam had pushed out, but it is only about an ⅛” off at the most, so I left it alone, as I should be able to fill in the missing length on the horn with paper mâché paste:
Former--K--150-degree curve, shellacked 1.jpeg

The Bell Former--

The Styrofoam rings for the bell former, before gluing and sanding:
Former--M--Bell rings stacked prior to glueing.jpeg

The formers for the two halves of the bell (prior to sanding) with the wood end discs glued and clamped in place:
Former--G--Bell halves clamped.jpeg

The latch to keep the two halves of the bell together while sanding (to ensure a smooth flare):
Former--H--Bell halves--latch.jpeg

The formers for the two halves of the bell after sanding:
Former--I--Bell sanded.jpeg

The formers for the two halves of the bell after three coats of shellac (latched together):
Former--J--Bell halves shellacked, together.jpeg

The Horn--

The horn spigot—1.5” outer diameter DOM steel tubing (1 ⅜” inner diameter):
Horn--1--Spigot.jpeg

I bought a large roll of thick paper to make the horn, and finished two layers (overlapping paper strips about 50%) on the first section of former, but the layers didn’t laminate, I think because the paper was too thick and stiff for me to press it in place properly; so I had to tear it all off and start again with the same thinner paper I used for the last horn—so far, I have three layers done (overlapping about 75%), and the paper is laminating properly now.

The first section of the horn with two layers of the thick paper:
Horn--2--Two layers of tick paper.jpeg

Detail of delamination, after cutting some paper away—I could feel that the overlapped paper was delaminated for most of the section’s length, which was confirmed by cutting some of the outer layers away:
Horn--2--Detail of delamination 3.jpeg

The first section of the horn after tearing off the thick paper and applying a layer of thinner paper:
Horn--3--One layer of thin paper.jpeg

The same section after two subsequent applications of thinner paper (three applications total)—it took about three hours to do the third application, although I think I was overlapping somewhat more than 75%:
Horn--4--Third layer of thin paper.jpeg

Unfortunately, the paper shrank a bit as it dried, so I need to add another partial layer at the large end to overlap with the next section of former by about ½", in order to align the the first section of the horn with the second section of the former:
Horn--4--Shrinkage at large end of first section.jpeg

More to come over the next few weeks!--I'm hoping to be able to test the horn before I go back to college in about three weeks, although I doubt I will be able to get the full ¼" thickness for the neck by then.

This project of yours is fascinating !

I am not technically minded though (if that is the phrase).

So, you made the wooden skeleton and filled the out-side of it with expansion foam. Did you then cover this in papier mache and then remove the internal skeleton when the papier mache parts were self-supporting and 'strong enough' ? If so, how did you remove the skeleton from inside ?

I apologise if you have explained this already and / or it is self-evident.

Also, I hope my questions make sense

[Roaring20s]

I'm rooting for you!

James.

[Ethan]

I'm rooting for you!

James.

Thank you!

This project of yours is fascinating !

I am not technically minded though (if that is the phrase).

So, you made the wooden skeleton and filled the out-side of it with expansion foam. Did you then cover this in papier mache and then remove the internal skeleton when the papier mache parts were self-supporting and 'strong enough' ? If so, how did you remove the skeleton from inside ?

I apologise if you have explained this already and / or it is self-evident.

Also, I hope my questions make sense

Thanks! The paper mâché went on top of the expansion foam, and the wood skeleton remained embedded in both to make a solid piece--my hope is that having the wood running from one end all the way to the other will help it release from the horn, as the force of the mallet striking the narrow end should be transmitted along the whole length. Here's a close-up of the former, showing how the wood skeleton is still in place, with paper mâché filling in the gaps:

[Inigo]

You keep our mouths opened!!! Great!!!

[Ethan]

Thanks, Inigo!

I successfully extracted the former for the first section—it was firmly stuck, but popped out after hammering heavily on the end several times with a mallet and dowel.

The inner surface is a little winkled and there were some loose pieces of paper inside (which I was able to scrape out), probably because the Vaseline that I used as a releasing agent soaked into the first layer of thin white paper, so that layer didn’t bond to the next properly—the white paper is a bit smoother than the brown, which seemed as though it would be beneficial for the treble response if used as an inner lining layer, but if it causes wrinkling, it would probably be better to skip it.

Oddly, the shellac finish on the former seems to have deformed; it doesn’t feel as smooth as it did to start, and there are a few places were there seems to be a thick build-up of shellac. (It could actually be a combination of Vaseline and dried glue, but the orange color looks like shellac.)

As of now, I have two layers of paper on the next section—my plan is to do one more layer over just the large curve, then at least two over the whole before extracting the former, and then adding more layers as necessary.

For some reason, the paper is raised up from the former at the large end—I’m not sure why, but at any rate, I think I will have to cut it so that I can press it against the former, then add more layers on top to keep it flat.

Here are some photos—

The exterior of the first section of the horn:

The large end:

The interior looking through the spigot (I was able to remove the loose bit of paper on the upper left):

The interior looking towards the spigot:

The former after extraction:

Detail of shellac build-up, visible as a couple darker spots near the end of the former:

The horn with two layers of paper extended over the large curve (not trimmed to the end of the former):

Detail of paper not lying flat against the large end of the former (after trimming)--the lifted paper is visible on the right-hand side of the photo:

And here’s a video of the first section of horn in action before I started papering the next section—it sounded about as one would expect for something of its size, although the bass was surprisingly strong, audible down to about 100 Hz: https://youtu.be/48IxSfDBSUQ

[poodling around]

Thanks, Inigo!

I successfully extracted the former for the first section—it was firmly stuck, but popped out after hammering heavily on the end several times with a mallet and dowel.

The inner surface is a little winkled and there were some loose pieces of paper inside (which I was able to scrape out), probably because the Vaseline that I used as a releasing agent soaked into the first layer of thin white paper, so that layer didn’t bond to the next properly—the white paper is a bit smoother than the brown, which seemed as though it would be beneficial for the treble response if used as an inner lining layer, but if it causes wrinkling, it would probably be better to skip it.

Oddly, the shellac finish on the former seems to have deformed; it doesn’t feel as smooth as it did to start, and there are a few places were there seems to be a thick build-up of shellac. (It could actually be a combination of Vaseline and dried glue, but the orange color looks like shellac.)

As of now, I have two layers of paper on the next section—my plan is to do one more layer over just the large curve, then at least two over the whole before extracting the former, and then adding more layers as necessary.

For some reason, the paper is raised up from the former at the large end—I’m not sure why, but at any rate, I think I will have to cut it so that I can press it against the former, then add more layers on top to keep it flat.

Here are some photos—

The exterior of the first section of the horn:
Horn--5--First section after removal of former (exterior).jpeg

The large end:
Horn--5--Interior of first section 1.jpeg

The interior looking through the spigot (I was able to remove the loose bit of paper on the upper left):
Horn--5--Interior of first section, looking through spigot.jpeg

The interior looking towards the spigot:
Horn--5--Interior of first section, looking towards spigot.jpeg

The former after extraction:
Former--O--First section after removal.jpeg

Detail of shellac build-up, visible as a couple darker spots near the end of the former:
Former--O--Detail of deformed shellac on surface.jpeg

The horn with two layers of paper extended over the large curve (not trimmed to the end of the former):
Horn--6--Two layers over large curve.jpeg

Detail of paper not lying flat against the large end of the former (after trimming)--the lifted paper is visible on the right-hand side of the photo:
Horn--7--Detail of paper lifted off of former.jpeg

And here’s a video of the first section of horn in action before I started papering the next section—it sounded about as one would expect for something of its size, although the bass was surprisingly strong, audible down to about 100 Hz: https://youtu.be/48IxSfDBSUQ

Wow ! Fantastic sound on your excellent video ! To me, this sounds at least equal to an emg with an original horn already !

Excellent, and still fascinating. Can't wait for the next instalments ! Thank you so very much for sharing your adventure !

[Ethan]

Thanks! In person, though, there would be no competition between the horn stub and an EMG--even with bass as deep as it had, the stub sounded much less full than my Victrola Alhambra with its 200 Hz horn (which doesn't sound as full as my 100 Hz homemade horn). Here's the frequency content of the video of the horn stub--I believe everything below about 100 Hz is ambient sound, with the possible exception of the small 66 Hz peak:

I ended up tearing out the worst of the lifted-up paper and redoing it with the strips parallel to the horn’s axis instead of perpendicular to it (which I remembered, after doing two layers with strips wrapped around the horn, was how I did the first layers on the last horn); the replacement layers seem to be working well now, although there might be a small discontinuity that I’ll have to fill in on the inside of the horn, as it feels as though there are still places where the paper isn’t perfectly snug against the former.

Strangely, the horn seems to have bent between when I set the former for the 150-degree bend in place on the first section and now; instead of having the horn’s axis pointing very slightly downward at the large end of the curve, it seems to be pointing horizontal or ever-so-slightly upward—I’m not sure why; all I can think of is that the weight of the former in the center of the horn (when in position on the holder) pulled the middle down, changing the angle. I’m not quite sure how that would have happened, though, as I had a support below the first curve while doing the first couple layers of paper on the large curve, and after a few layers, the horn walls felt stiff enough that they should have been able to withstand the weight of the former, unless I removed the support before the innermost layers of paper were completely dry. Fortunately, it isn’t a major change, so I don’t think it should cause any significant problems.

The wall thickness where the two curves meet seems to be around 3/16” thick as of the fourth layer of paper, so I am hoping that I will be able to extract the former Sunday or Monday, after finishing the fourth layer and possibly adding a fifth, depending on the wall thickness at the large end of the section.

Some more photos—

Here is the horn with the lifted paper town off—a rather depressingly large area had to be torn out, but on the bright side, tearing it out proved that paper appliqué is quite sturdy—I had to use a pocket knife to cut the paper and peel back the edges of the strips before I could tear them off, and even so, it wasn’t terribly easy to deconstruct.

And the same part of the horn with one layer (out of two) of paper to patch what was torn out:

The whole section with three complete layers and a partial fourth (starting at the throat and continuing to a little past the middle of the large curve):

[Ethan]

Well, I finished the fourth layer Saturday and tried hammering out the former Sunday, by which I mean that I spent all of Sunday trying to hammer out the former. Conscripted both parents to hold it in place while I hit the end with a mallet and (for a brief while) a small sledgehammer—broke two ¾” plywood ramrods—accidentally smashed the wood flange on the spigot—accidentally loosened the metal spigot from the paper horn—went back to it in the evening after dinner—and by the end of the day, the former had only moved out about 1/16” to ⅛” at best. Tried again for a little while Monday—no luck. Cut off the 60-degree curve so I could hit the former directly (without a ramrod)—still no luck, but I found out the end of the former was slightly crushed from the hammering. Tried again to-day, with no better luck, so I cut the horn lengthwise along the outside of the curve, using a non-serrated knife to minimize loss of material—pried the horn off the former along the cut—and finally was able to pop the former out by slamming the narrow end against a concrete garage floor. (Methinks the former design needs some—er—slight—modification.)

It turns out some of the glue from the horn had stuck to the former despite using (I thought) a thick coat of Vaseline, and the shellac did the same clumping-thing as with the last section, hence its tenacity in the face of removal—rather a nuisance, but now I only have to deal with extracting the two straight sections of former for the bell, which should be much easier, at least by comparison.

I glued the cut along the back of the 150-degree curve, but will have to glue it again, as I couldn’t find a good way to clamp the two sides so used tape instead, and will need to add glue where there is currently tape across the cut.

Any thoughts on whether it would be better to re-attach the two curved sections by gluing the ends together and adding more paper over them (I plan to add about ⅛” more anyway, but wasn’t planning on doing it until fall break), or by attaching wood mounting rings to the adjacent ends, similar to how EMG made their Export horns?

More photos—

The first of the two broken ramrods, and the broken spigot flange:

The horn after cutting off the first curve and before extracting the second former:

Close-up of the cut:

Interior of the first section where the spigot came out:

Interior of the first section where I cut it off the next:

The second section’s former after extraction:

Close-up of the former—the paper mâché came loose between several ribs:

Close-up of the former—clumped shellac:

Close-up of the narrow end of the former where it was crushed by the mallet and sledgehammer—the two end ribs came loose:

The cut along the back of the 150-degree curve section of horn:

Interior of the same section prior to cleaning out Vaseline and shellac residue and crumbs, looking into the large end:

Interior of the same section prior to cleaning, looking into the narrow end:

The same section with the first application of glue along the cut drying, with the two sides held in place by tape:

[Roaring20s]

Thomas Edison’s quote, “I have not failed. I’ve just found 10,000 ways that won’t work,” is a perfect reminder that failure is not the end. It is just the beginning of a journey toward success.

You're becoming an expert.

James.