Homemade diaphragms

[anchorman]

I have been working on a meltrope style back and aluminum diaphragms for a Victor exhibition soundbox given to me by a fellow collector a little while back.

The back was 3D printed, and I 3D printed some molds to make the rubber bushing in various sizes. I also 3D printed a form on which to emboss the diaphragm from aluminum. The aluminum used was from the top of a large tray that food came in at work. No annealing necessary, as it was already very soft and easy to form. A simple wooden tool was enough to make the ring shaped ridges, and stiffened the center of the diaphragm marvelously, while allowing it to flex around the edge.

It does not have as flat a response as the Columbia number 9, or the HMV 5b, and is not so loud. I think some of this is due to inferior springs used in the exhibition, and these will hopefully be replaced with point contact springs similar to EMG/expert soundboxes.

Attached here are some videos comparing the sound of this soundbox playing on an HMV 102, along with the 5b, and a Columbia No 9.


https://www.youtube.com/watch?v=KcSO93o-jdI

https://www.youtube.com/watch?v=06tB3-0b-_Y

https://www.youtube.com/watch?v=2md-DzMQ8Gg


It’s hard to get a real feel for the tonal changes with these recordings, as the levels are automatically adjusted when recording on my phone. Hopefully I’ll get a better recording setup at some point with manual control of the sound level. Regardless, there is some difference to be heard between them.

[anchorman]

The diaphragm in the picture with the form was the first attempt, and the one actually used came out a little better.

Some frequency comparisons made using a fixed tone disc from Columbia, spectrum analysis with one of the many programs available for iPhone.

The aluminum diaphragm is a definite improvement in the characteristics of both the mica diaphragm soundboxes. The nasty 1khz resonance of the exhibition is tamed with the aluminum diaphragm.

These tests were done on a friend’s cabinet model victrola, so they don’t reach as deep into the bass region as they did even on the horn on the HMV 102 portable.

I see now why EMG/expert chose this soundbox as their starting point, as the smaller diaphragm reaches more into the treble range than the larger mica diaphragm can do. I’m curious to see what better springs and needle bar can do to help in this regard.

I’ve been planning to build a large external horn machine for some time, and hope that these soundbox experiments might bring something to that machine when it is built. In the meantime, it’s fun to tinker and play with these things and see what I can learn.

[anchorman]

Spectrum of the modified exhibition, along with HMV 5b and the Columbia No 9 on HMV 102 portable

The 5b shows increased frequency response up to 4khz that is markedly stronger than the No 9. All of this is reflected in the Columbia sound vs the HMV sound. So much of the tonality of the machine is from the soundbox, to the point where the Columbia reproducer on the HMV 102 machine makes it sound rather like the Columbia 202. I don’t have an adapter for the HMV to fit the Columbia so I can’t yet make that comparison the other way around.

The modified exhibition is not nearly as flat in the current configuration. Hopefully I’ll be able to test the HMV 4 with mica diaphragm along with the exhibition with mica diaphragm on this same machine to better be able to compare results.

All of these using Burmese colour needles, as I am not wanting to trash my immaculate looking test records using steel needles. When I get the good machine built, and when I’m a little better at using the spectrum analyzer software, I’ll run the steel needle vs BCN test…

[anchorman]

I’m just noticing that I had the lowest dB level displayed by the SA program set differently for the Columbia and HMV 5b vs the other soundboxes, so comparing between the graphs is a little more difficult. But overall, the trends are clear.

[anchorman]

The reason this is showing all of these distinct peaks, is due to using a record that plays individual tones, rather than a gliding frequency tone. Lesser peaks are harmonics, most of which seemed to be 2nd harmonic.

The shape of the ridges chosen for this first attempt at a diaphragm were sort of based on pictures of EMG diaphragms, and sort of based on what felt right as I was making the model. Hopefully changing these shapes will give some insight into what is best.

[Hoodoo]

After watching the three videos, I have to say that I like the modified Exhibition best.

[CarlosV]

Interesting experiment! indeed the Expert soundbox is largely based on the exhibition design, but they managed to flatten its response and get rid of the 1kHz resonance. Maybe its heavy weight shifted the resonance to a lower range. I wonder what would be the effect of replacing the HMV 4 diaphragm with an aluminum one, as it shows a flatter response in its original mica configuration.

[Henry]

Since the input device is an iPhone microphone (!), I'm not lending much credence to the graphed results.

As for the Exhibition, cf. these data from Paul Edie: http://www.victor-victrola.com/Soundbox%20Article.htm

[anchorman]

Since the input device is an iPhone microphone (!), I'm not lending much credence to the graphed results.

As for the Exhibition, cf. these data from Paul Edie: http://www.victor-victrola.com/Soundbox%20Article.htm

I'm not posting them as "absolutes", but you can definitely see a trend happening, and results, though not measured precisely to half of a decibel are consistent from sample to sample, and they track with the sound that one hears playing from the horn. So still useful graphs to compare between one another. They would be utterly useless if we tried to compare them to graphs that someone else made under different conditions, but I'm not trying to do that. I'm trying to see and verify with my eyes what my ears are telling me, and for that they are good enough for now! I can hear that the HMV 5B is reaching up higher in the frequency range than the others I have, and I can also see that happening in the SA graphs.

[anchorman]

Interesting experiment! indeed the Expert soundbox is largely based on the exhibition design, but they managed to flatten its response and get rid of the 1kHz resonance. Maybe its heavy weight shifted the resonance to a lower range. I wonder what would be the effect of replacing the HMV 4 diaphragm with an aluminum one, as it shows a flatter response in its original mica configuration.

I'm going to try that with the HMV 4 at some point. I've got a few extras that a friend will be shipping to me from the UK, so will have plenty with which to experiment soon. The aluminum diaphragm makes more of a hump shape than the mica in the exhibition, which looks relatively flat if you could eliminate the 1khz resonance. It's possible it comes down to the springs. Hopefully I can make some better springs before too long and see how those help. It is hard to compare apples to apples, as there are so many variables at play, but you can still see some trends happening as adjustments are made. What a grand waste of time it all is given the technology we have now - but it keeps me from rotting my brain in front of the telly