Another Homemade Exponential Horn Project

[Inigo]

The longitudinal cut seems a good trick, not spoiling the resistance of the assembly. For similar reasons, I bet using two flanges for attaching the two horn sections as in export model send to be better for reinforcing it, as the bending effect on the curves would be strong when they are supporting them weight of the bell.
An external strut helping the large bell weight as other colleagues herein have shown is also useful.

[Ethan]

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.

True! Theoretically, it seems if you fail enough times, the solution—or a solution, anyway—really ought to appear simply by process of elimination. (Although somehow, I suspect it’s not actually quite as simple as that…)

The longitudinal cut seems a good trick, not spoiling the resistance of the assembly. For similar reasons, I bet using two flanges for attaching the two horn sections as in export model send to be better for reinforcing it, as the bending effect on the curves would be strong when they are supporting them weight of the bell.
An external strut helping the large bell weight as other colleagues herein have shown is also useful.

The external strut (or maybe a stout crane) is a must in this case—I have the B-39 (my first horn) resting on top of a bridge lamp that happened to be just the right height; when I had the same lamp supporting it right at the joint between the neck and bell, the horn started to sag, so I moved the support to about half-way down the length of the bell, and it seems to work fine. Currently, my plan is to make a wood support that attaches to the outside of the B-43 (the horn I’m currently making), although that might have to wait until next summer. I think you’re probably right about the mounting rings being stronger—they don’t look quite as elegant as a one-piece horn neck, but a horn with mounting rings all over it looks infinitely better than a horn lying in pieces on the floor because it wasn’t strong enough!

I’ve started papering the horn bell, and with four layers of thick paper overlapped 50-75%, the walls are about ⅛” thick, which should be thick enough for structural stability. I glued the mounting rings in place yesterday, but the glue isn’t dry yet, so I’ll have to wait to remove the formers—I’m hoping they’ll be ready by this evening, but they might have to wait another day for the glue to set completely.

The large end of the horn neck seems to have shrunk a bit where I had to glue the seam from cutting it off the former, making it noticeably smaller than the adjacent end of the next section of horn; conveniently, a piece of scrap wood from the center of the mounting rings between the neck and bell is just about the same size as the end of the bell former, so I stuck it in the end of the neck to stretch it back into shape until the mounting ring is in place on it and the glue is dried.

The shrinkage of the paper on the larger part of the bell pushed the former out about ⅜”, so I had to trim a little paper off the end; I’m not absolutely certain that I trimmed the end diameter right, but I’m hoping that it will at least be close enough—based on a visual approximation from the outside, the adjacent ends seem reasonably close.

More photos:

The former partially pushed out of the horn—about ⅜”:

The horn sections (the bell halves on the formers, the neck off), with mounting rings in place:

And here are the latest in-progress test videos, with the whole neck in place—the sound still isn’t nearly equal to the first horn, but it’s better than it was at the last stage, and I think it may have deeper bass than the B-39:
Playing the same record as in the first test video: https://youtu.be/qNzgfr1u9w4
Playing a 1929 Victor electric: https://youtu.be/Nm3UGuJ0rbg
Playing a 1927 Victor electric: https://youtu.be/xcXTCDI4kxA
Playing a 1926 Victor electric: https://youtu.be/iFrdhyxsRVE

I also came across a video I made but never posted of the B-39 at the same stage, playing one of the same records (although a better-condition copy—I was using more-worn duplicates for tests this time, as I didn’t know whether the tiny horn mouth would have harmful resonances): https://youtu.be/A75700_MsTM

In Audacity, it looks as though the B-43 neck has deeper bass (down to about 75 Hz, compared to about 90 Hz for the B-39), but the B-39 neck had stronger bass, specifically around 92 and 112 Hz; but as I had the camera closer when I made the video of the B-39, I’m not sure that videos represent the difference between their sounds accurately.

[BassetHoundTrio]

I am in awe! I have an EMG Xb Oversize, and am so impressed with your abilities. While I am terrible in areas mathematic, my ears can tell you that Balfour Davey got the math 100% right on his horn designs. Thank you so much for sharing all of the details, and if you are ever near Chicago, I hope you will let me know - would love for you to visit.

[chunnybh]

Very exciting. I can't wait to see the video of it all assembled.

[Inigo]

Yes... if half a horn already has such a sound, how will be the complete unit? it will blow the roof out!

[Ethan]

I am in awe! I have an EMG Xb Oversize, and am so impressed with your abilities. While I am terrible in areas mathematic, my ears can tell you that Balfour Davey got the math 100% right on his horn designs. Thank you so much for sharing all of the details, and if you are ever near Chicago, I hope you will let me know - would love for you to visit.

Thanks! The math for making a horn isn’t actually all that complicated; I just used the exponential horn radius equation given in Modern Gramophones and Electrical Reproducers, plus some simple circle-related equations for the bends. (The theory is much more complicated--I make no pretense of understanding the math of that.) I appreciate the offer--please do likewise if you're ever near Pittsburgh!

Yes... if half a horn already has such a sound, how will be the complete unit? it will blow the roof out!

The full horn doesn’t actually seem much louder than the B-39 (which is plenty loud enough)—much fuller sound and stronger bass, though!

Very exciting. I can't wait to see the video of it all assembled.

Voilà!:
"Cryin' for the Carolines" (before adding felt gaskets): https://youtu.be/BLKksEeJmv8
"Cryin' for the Carolines" (after adding the gaskets): https://youtu.be/8NOrqyLWsjU
"Nobody But You" (before adding gaskets): https://youtu.be/DNNFYv4wikg
"Did You Mean It?" (before adding gaskets): https://youtu.be/xRuMsBcJwc4
"To-Night's My Night With Baby" (before adding gaskets): https://youtu.be/yIwZCGyKPX8

The bell formers came out only semi-successfully—the mouth former popped out quite nicely after a few whacks with a mallet, but the rear bell former stuck; the foam compacted when struck and the wooden end disc broke after several minutes of unsuccessful hammering. At that point, I figured there wasn’t much point trying to get the rest out intact, seeing as the narrow end was already so smashed up, so I removed the former by breaking it apart from the inside; it seems the glue from the horn had adhered to it in a few spots (again…), and it was apparently enough to prevent the former from coming out. I suspect that I didn't have enough Vaseline on the former, as I used up most of what I had on the larger section, which came out much more easily. I used an old credit card to scrape as much of the remaining Vaseline as possible from the interior of the horn, then scrubbed the surface with alcohol wipes to remove any residue--it seems to have worked quite well; the inner surface isn't tacky at all now.

The paper on the mouth section of the bell is rather wrinkly and doesn’t lie flat everywhere, and there are quite a few places where the innermost layers (the first ones to be applied) delaminated, so it needs at least some regluing, and/or possibly a thin coating of something that would fill in the voids or flatten the raised paper—as it’s only at the largest diameters, it seems as though regluing alone might be sufficient, but smoothing the surface would probably make it easier to apply a decorative paper, which I hope to do over winter break. Either way, fixing the bell will have to wait until fall break at this point, as I'll be going back to college in a couple days.

The paper shrank away from the mouth, the same as it did on the last horn, so I had to trim off about ½” off the length, for a final mouth diameter of about 33.5”, which is about what I planned.

There were some slight gaps between the bell halves between the bolts when I first assembled the horn, so after I made the first tests, I dismantled the whole, added felt gaskets between all three joints, and reassembled it—as there doesn’t seem to be a noticeable difference in sound quality, I would guess that the gaps were small enough and close enough to the mouth not to cause much loss of bass.

Sound-wise, it's a major improvement over the last horn; not every record suddenly has a lot of bass (although some do!), but the sound is generally much fuller, and possibly clearer as well; and surface noise seems slightly less noticeable on some records.


More photos--

The complete horn--the B-39 technically could support itself and the cabinet could hold it up without quite tipping over, but at least with the current neck thickness (~3/16"), the B-43's bell is too heavy for the neck to support safely, and the whole thing is too heavy for the cabinet to hold up, so some sort of external support is necessary--a bridge lamp is currently fulfilling the role:

The wrinkled paper inside the mouth:

Shrinkage of of individual layers of paper at the mouth rim--each layer was initially trimmed right to the edge of the former, not staggered as pictured below after the shrinkage (which also made the horn contour slightly concave right at the rim, hence the need to trim some off):

[Watanabehi]

I am in awe! I have an EMG Xb Oversize, and am so impressed with your abilities. While I am terrible in areas mathematic, my ears can tell you that Balfour Davey got the math 100% right on his horn designs. Thank you so much for sharing all of the details, and if you are ever near Chicago, I hope you will let me know - would love for you to visit.

Thanks! The math for making a horn isn’t actually all that complicated; I just used the exponential horn radius equation given in Modern Gramophones and Electrical Reproducers, plus some simple circle-related equations for the bends. (The theory is much more complicated--I make no pretense of understanding the math of that.) I appreciate the offer--please do likewise if you're ever near Pittsburgh!

Yes... if half a horn already has such a sound, how will be the complete unit? it will blow the roof out!

The full horn doesn’t actually seem much louder than the B-39 (which is plenty loud enough)—much fuller sound and stronger bass, though!

What an amazing work! Thank you so much for sharing this process.
Very exciting. I can't wait to see the video of it all assembled.

Voilà!:
"Cryin' for the Carolines" (before adding felt gaskets): https://youtu.be/BLKksEeJmv8
"Cryin' for the Carolines" (after adding the gaskets): https://youtu.be/8NOrqyLWsjU
"Nobody But You" (before adding gaskets): https://youtu.be/DNNFYv4wikg
"Did You Mean It?" (before adding gaskets): https://youtu.be/xRuMsBcJwc4
"To-Night's My Night With Baby" (before adding gaskets): https://youtu.be/yIwZCGyKPX8

The bell formers came out only semi-successfully—the mouth former popped out quite nicely after a few whacks with a mallet, but the rear bell former stuck; the foam compacted when struck and the wooden end disc broke after several minutes of unsuccessful hammering. At that point, I figured there wasn’t much point trying to get the rest out intact, seeing as the narrow end was already so smashed up, so I removed the former by breaking it apart from the inside; it seems the glue from the horn had adhered to it in a few spots (again…), and it was apparently enough to prevent the former from coming out. I suspect that I didn't have enough Vaseline on the former, as I used up most of what I had on the larger section, which came out much more easily. I used an old credit card to scrape as much of the remaining Vaseline as possible from the interior of the horn, then scrubbed the surface with alcohol wipes to remove any residue--it seems to have worked quite well; the inner surface isn't tacky at all now.

The paper on the mouth section of the bell is rather wrinkly and doesn’t lie flat everywhere, and there are quite a few places where the innermost layers (the first ones to be applied) delaminated, so it needs at least some regluing, and/or possibly a thin coating of something that would fill in the voids or flatten the raised paper—as it’s only at the largest diameters, it seems as though regluing alone might be sufficient, but smoothing the surface would probably make it easier to apply a decorative paper, which I hope to do over winter break. Either way, fixing the bell will have to wait until fall break at this point, as I'll be going back to college in a couple days.

The paper shrank away from the mouth, the same as it did on the last horn, so I had to trim off about ½” off the length, for a final mouth diameter of about 33.5”, which is about what I planned.

There were some slight gaps between the bell halves between the bolts when I first assembled the horn, so after I made the first tests, I dismantled the whole, added felt gaskets between all three joints, and reassembled it—as there doesn’t seem to be a noticeable difference in sound quality, I would guess that the gaps were small enough and close enough to the mouth not to cause much loss of bass.

Sound-wise, it's a major improvement over the last horn; not every record suddenly has a lot of bass (although some do!), but the sound is generally much fuller, and possibly clearer as well; and surface noise seems slightly less noticeable on some records.


More photos--

The complete horn--the B-39 technically could support itself and the cabinet could hold it up without quite tipping over, but at least with the current neck thickness (~3/16"), the B-43's bell is too heavy for the neck to support safely, and the whole thing is too heavy for the cabinet to hold up, so some sort of external support is necessary--a bridge lamp is currently fulfilling the role:
Horn--15--Complete.jpg

The wrinkled paper inside the mouth:
Horn--15--Wrinkly paper in mouth.jpeg

Shrinkage of of individual layers of paper at the mouth rim--each layer was initially trimmed right to the edge of the former, not staggered as pictured below after the shrinkage (which also made the horn contour slightly concave right at the rim, hence the need to trim some off):
Horn--15--Shrunken paper at mouth rim.jpeg

[Inigo]

As colleague emgcr demonstrated with his fiberglass horns, you'll see how noticeably the sound clarity improves (yes, yet more and better) once you've trimmed the mouth to a perfect smooth circumference... Seems that an adequate smooth and stiff shape at this circumference is very important for the sound pressure waves to separate cleanly from the horn mouth and project a detailed sound into the room.

[JohnM]

I’m a bit late to the party on this, but I think that much of your form-making might be simplified by using medium-density urethane foam board (one brand name is SignFoam), commonly used in the sign-making industry. Anything that may be done to wood — carve/rout/sand/etc, may be done to MDUF and it may be glued together with a urethane glue (‘Gorilla Glue’).