Hidden/unused ultrahigh frequencies on a 2 minute cylinder.

[dk6400brian]

I just bought and received an Edison Home Model B with a Model C stylus. Couldn't hold it back. I just had to clean and restore it, carefully. More on that subject later.

Almost finished with the restauration project, I had a look at the stylus on the Model C reproducer tonight, under magnifying glass.
I'm very nearsighted (-6 on both eyes), so if I take my glasses off I can see objects clear in a distance of 6-7 cm/2½-3" with a resolution way under one millimeter. Coupling that with a tiny magnifying glass I have a resolution around the 0,005-0,01mm mark. Something in that range. Perhaps it's even lower. I certainly can distinguish tiny details.

Now. I had a look at the stylus which is very round to say the least, but I made another discovery that I'll like to share:

I inspected the grooves on a 2-minute Columbia cylinder, (57564 "Lær Mig Nattens Stjerne", Vilh. Michelsen, Tenor) with this setup. The grooves are in some places completely worn to pieces, but others aren't. To my astonishment in the bottom of some of the grooves, there were unused fine microdetailed hills and dales, as if the stylus were too big to reach them. They were clearly high frequency harmonics that came in pairs and repeatedly.

In order to roughly determine the frequency I had to do some math:

A 2-minute cylinder is 2.25" in diameter, equivalent to 5.715 cm. (Forgive me for having to switch to metrics)
The circumference is therefore 5.715*pi = 17.95 cm.

Groove length per minute: 17.95 cm * 160 rpm = 2872.67 cm/min
Groove length per second: 2872.67/60 = 47.88 cm/sec or 478.8 mm/sec (Rounded to 480 mm/sec)

Wavelength of selected frequencies:

31.25	Hz	15.36
62.50	Hz	7.68
125	Hz	3.84
250	Hz	1.92
500	Hz	0.96
1000	0.48
2000	0.24
4000	0.12
8000	0.06
16000	Hz	0.03	mm

Width of a 2-minute groove is 1/100th", equivalent to 0.254mm

I estimated the width of the unused high frequency hills and dales in the bottom of the groove to be 1/5th of the width of the groove.

My best guess is, that the unused hills and dales do have a frequency of:

0.254mm/5 = 0.0508mm
480mm/0.0508mm= 9448.82 Hz

In my best judgement, the Columbia cylinder would have no problem in containing even 16000 Hz with a wavelength of 0.03mm. There's certainly some potential in wax as a medium.
Unfortunately the Model C stylus is too wide to reach the bottom of the groove, so that's perhaps one of the reasons why the reproduced frequency response is limited. Perhaps by experimenting in a better stylus shape, I'd say High Fidelity is within reach.

I also discovered one other thing during this. The recorder at Columbia that originally cut the grooves were a bit too wide. The top of the groovewalls is not straight, but follows vaguely the hill/dale pattern if the amplitude/volume/dynamics is too high. Two grooves next to each other containing high volume has a sharp top of the groovewalls or the top has been carved/cut away.
I can tell the difference of a worn groove and an intact one. If a wornout stylus has destroyed the grooves, then all grooves would've been in a severe state. They are not in this case.

Anyway. Finding such high frequencies on a cylinder just had me at hello
I'll try and think of a setup in order to combine my magnifying glass with a digital camera, so this can be documented in some way.

Cheers
Brian (New member)

(edit) Something went wrong in creating the matrix for wavelengths....sorry. Hope you can figure it out.

[lucius1958]

You should very definitely document this!

BillS

[dk6400brian]

I've taken pictures of it. The dark lines in the harmonics are certainly 5 times smaller than the width of the groove.



One other thing.
Have you discussed if turning a recorder stylus (the sharp end of a 0.01" glass rod) 180 degrees, so the end of the glass rod sort of floats on the bottom of the groove, instead of diging into it, would give a better representation of the recorded groove ? It's just a thought being a newcomer

(edit) I'll send you a pm with the link to the full picture. Please tell me what you think of it.
(edit#117) Link: lHarmonics1

[refseries]

Dear Brian

This is very interesting. Let me have a look at some Columbias under my microscope. If I find any of these strange grooves I will record the cylinder and look at the recording on an oscilloscope to see if they show there. Certainly a ball two minute stylus won't pick up anything much over 5KHz.

Keep you posted

Keith

[dk6400brian]

Hi Keith,

I had just been a Phonograph owner for a day or two, when I couldn't resist digging into the grooves, how things are recorded and reproduced.

Now 10 days later I've come up with a design for mechanical 2 track stereo recording unit and the corresponding reproducing unit on a 100 TPI cylinder and found candidates for materials for it to happen. Both units neatly designed to fit into the carriage. For now it's just a category 5 brainstorm, but eventually it must happen. Without ideas, even Edison would have flipped burgers.
I can't find material candidates for 2-track stereo on the 200 TPI cylinders, so it will have to rest.

The design of the original Edison recorder and the reproducer is intelligent in all its simplicity.
With the recorder, the positive phase of a soundwave will leave a dale in the groove and when the reproducing stylus picks it up, it will all be in phase, due to the linkage/suspension of the stylus arm, forcing the diaphragm upwards, when the reproducing stylus hits a dale.

I can't help to think that if Berliner hadn't invented the disc grammophone, we would have had highly evolved cylinder phonographs by now.

The idea of having the sharp end of a sapphire stylus in direct contact with the diaphragm cut the grooves has proved to be brilliant, but there are setbacks. The shape and size of it leads to inter-cutting between grooves and that must be the hills and dales one can find on the surface by looking at a cylinder with the naked eye. Also the position of the recording stylus limits dynamics due to slew-rate.
But the sharp end is smaller and much sharper than the reproducing sapphire stylus which would explain why I found such high frequencies.

The size and shape of the reproducing stylus. It's way to big in my opinion and thereby hinders it to pick anything up above 5KHz. It is also floating on top of the groove where it will pickup the inter-cutting and scratches/dirt.

I will be most grateful if you would post your results. With modern equipment a wax cylinder wouldn't have any problem in containing the full frequency range and due to the size of the sharp end of a recording stylus I think it's true that it in fact did record high frequencies 100 years ago, never to be heard until now.

Will you be recording using a setup with a mandrel and a modern turntable arm equipped with a 78 rpm stylus ?

Kindest regards
Brian

[refseries]

Dear Brian

On re-examining your picture I would guess that your high frequencies are about 5-6KHz, but even so this is at or above the upper limit of a conventional ball stylus. The use of a ball was a balance of frequency response against wear; too sharp a stylus will wear out a record at first play, while too rounded will severely limit upper end frequencies. The early soft wax cylinders wore quickly, which I guess mitigated in favour of a fairly generous radius. Once the four minute indestructibles appeared, stylus radii reduced, and while Edison’s Amberol stylus was a narrow disc some Columbia styli I have are quite sharp – sorry but I can’t quote sizes. That said one should never use a four minute/indestructible stylus on a two minute wax record, as the recorded track will be ruined at first play.

Having said all that a ball stylus will ‘skate’ over higher frequencies in the wax, so these should be recoverable with a suitable stylus, and as long as tracking weights are very low this should be achievable without trashing the record. For normal recording purposes I use styli with original sizes supplied by Expert Stylus, in the interest of protecting the cylinder. However I have modern styli as well and can use these for indestructibles if needs be. You can see my humble efforts on sdrv.ms/18nSKPR, construction details being shown in the pdf.

It will be interesting to see what high frequencies can be found, if any. The limit will be the recording process, and as far as I am aware no wax cylinders were electrically recorded. In fact ignoring modern cylinders the only electrically recorded cylinders I have heard of are the very last Edison blue Amberols, the sound quality of which is very good indeed (pity about the content, though...). As wax cylinders or wax masters for indestructible cylinders used cutters driven by either acoustic vibration or by some form of pantograph, the mechanical inertia of the system will put an upper limit on recorded frequency. While I am no physicist I would have thought that your suggested figure of 5KHz is not far off.

Having said all that, let me look at my Columbias and see what I can find. Incidentally, the record you mention is quite rare – I produce a cylinder database for the CLPGS and have very few Columbia cylinder titles from continental Europe in the 5xxxx range. Do you have any more? If so, I would be very interested in the numbers/titles/performers.

Cheers

Keith

[borrilabs]

Cylinders are also capable of very low frequencies too! In the late 1990s and early 2000's I did quite a bit of electrical cutting of cylinders, I used a Fairchild magnetic transducer, driven by a 30 watt Western Electric valve amp, this ran the stylus of a floating weight style Edison recorder from the late 1890's. I had a device called the Edisonia pickup, to play them back, and also it had both a 2 and four minute stylus. I made some cylinders that were so wide in frequency response, that they sounded like a normal 45 rpm record, you could hear kick drums, bass guitars and they would make sub woofers pump, and also had dazzling highs that were the same as any flat modern record! I also made some experimental acoustic records that were surprising too. I made a diaphragm of green Victor record sleeve paper, in 3 steps, and the larger back piece, outside was two thousandths, while it got closer to the center it increased to 9 thousandths. I used RTV blue silicone for the dampening material, the diaphragm floated in it, and the stylus assembly was glued to the diaphragm and none of this touched the side of the cup, so it was free floating. I recorded at Madison Square garden with this recorder and could record those low bass drum thuds, used by Slick Rick, and it played back with that booming bass you hear the annoying cars play, so wide range is fully able to be done, and so can stereo be achieved on cylinders. Most collectors do not like some of the crazy experiments I have done over the last 15 years. So you know I am not full of hot air. The photo of the Archeophone, the white and pink cylinders were experimental electrical high fidelity recordings of LedZepplin, The Beatles, and some Orthophonic records dubbed to cylinders, the year .... 2000-2001! They are made on my own blanks, at the time a aluminum steartate, aluminum and paraffin compound. Henry C was supposed to be playing period records, but was demonstrating his remarkable player with my electrically recorded cylinders. Another photo shows the electrical recorder I made for high fidelity cylinder cutting, this also dates from 2000.

Attachments:

[borrilabs]

Here is some changes I made in the blanks from 2000- now. The pink and white ones were ones in a style I made from 2000-2011. The new style came out in may of 2012, with a single spiral and July of 2013 were double core spiral cylinders. The formulation is an aluminum soap with Sal soda and Ceresine. The combination of the compound is key, and if not done right will not yield good wax, even if all the ingredients are weighed correctly, it is heated to a frightening 520 degrees F, at or near its flash point. My early batches was made in 2 lb batches when I started out, but I make 15 lbs of compound at a time now. Cylinders can not be made from fresh wax, it has a special annealing process, in order to work, also a certain amount of scrap wax is added to catalyze and homogeneous the batch. Also if you even make cylinder wax, realize that a little bit of wax expands 4 times the batch size as it is a foaming off process. If you make the wax in the wrong container, it will be a nice 500 degree volcano, and as it erupts like Mount Vesuvius, spraying hot wax all over, and catching fire as the small hot amounts hit the oxygen rich air.

Attachments:

[kevin6569]

Wow. That electric recorder looks cool...where can I get one?

[borrilabs]

The electric recorder is something I made and designed. Fairchild lathe heads can be found on ebay at times. In the early 2,000nds interest in vinyl records declined and record cutting equipment was cheap. With the increased interest in record players and turn tables the value of record making equipment is increasing.