Also How were cylinders plated and pressed

[VRCM]

Also How were cylinders plated and pressed?

[cd4cutter]

For anyone interested in the early days of acoustic phonographs, the following books are very helpful:

"Edison Cylinder Phonograph Companion" by George L. Frow

"The Edison Disc Phonographs" by George L. Frow

"The Compleat Talking Machine" (note the spelling) by Eric L. Reiss

"Look for the Dog" (Victor Talking Machines) by Robert Baumbach

"Columbia Phonograph Companion, Vol. I" (cylinder graphophones) by Howard Hazelcorn

"Columbia Phonograph Companion, Vol. II" (disc graphophones) by Robert Baumbach

There are many other good references available, but these are the "bibles" of the phonograph collector. Every sound or audio-related museum should have these reference books on the shelf. Some are now out of print but can usually be found at used book sites.

The Frow cylinder book contains several appendices, one of which has considerable information about the Edison Business (dictation) machines.

The subject of cylinder manufacture is covered in some more obscure texts, one of which is "Edison Blue Amberol Recordings, 1912-1914" by Ronald Dethlefson. This book contains a section written by one of the employees at the Edison Phonograph Works describing the manufacture of the Edison Ediphone School Records. These were made the same way as the earlier Blue Amberol cylinders which were made of celluloid, one of the first "plastic" formulations to gain widespread use.

The first cylinder records were duplicated for commercial sale one at a time in a process called "pantographing". A master wax record was played with a stylus which was mechanically linked to one or more slave styli which each inscribed a new wax recording as the master was played. This process was very inefficient and did not produce very many copies before the master was worn out. Edison was one of the developers of a better method of recorded cylinder manufacture which he called the "gold molded" process. It is similar to the way disc records are manufactured.

Here is a greatly simplified description of the process: The master wax recording was deposited with a gold surface to make it conductive. This was put into an electroplating tank which built up a layer of metal over the wax. The wax was then melted away, leaving a "negative" version of the recording in the plated metal. This item (after strengthening with additional metal support) was used as a "stamper" or mold. It was put into a heated molding machine. A specially formulated wax (more accurately described as a "metallic soap") was poured into the cavity and allowed to cool. Upon cooling, the wax record shrinks sufficiently to allow it to be pulled out of the stamper mold. The Blue Amberol records were made in a similar process. The recorded stamper/mold is made as before. Then a tubular piece of thin celluloid is inserted into the mold. The molding machine contains a central expandable rubber bladder within the mold. The mold is heated up and the bladder is expanded which presses the celluloid tubing against the recorded mold. After cooling, the bladder contracts and the molded celluloid record shrinks slightly which allows it to be removed from the mold. Most record manufacturers of celluloid cylinders (there were others besides Edison) then strengthened the thin molded celluloid tube by installing a hard core material inside the tubing. Edison poured a plaster mixture inside the tubing, U.S. Everlasting used an asphalt-like material, and Albany Indestructibles were made by inserting a cardboard core reinforced with a metal ring on each end.

Very little detailed documentation exists on early record manufacturing techniques because the manufacturers treated this information as trade secrets. So little or nothing was ever written about the processes that was made available publicly. This attitude was maintained well into the 1970s - you will not find much written about the specific details of record compound formulation or about the specifics of the record press heating and cooling cycles. So, sadly, much of this know-how will vanish when the old-timers who worked in these manufacturing areas themselves vanish.

[VRCM]

WOW. That is great information. I can't believe it actually shrinks enough to be pulled out of the mother. I will check to see if our library already has them (doubt it though); if not I will search around and start to collect them. Thanks for the info.

[Steve E.]

Peter Dilg on Long Island still makes small runs of cylinders from molds! Wizard Records is his label. and indeed, the cylinders shrink upon cooling.

[VRCM]

Does this happen to lps when they are pressed?

[cd4cutter]

No, disc records are not designed to shrink after they are molded. I used to cut test records that required some very accurate diameters be recorded on them, so I conducted some experiments to determine what shrinkage may be taking place in vinyl record molding. The results showed that very little to no shrinkage occurs when molding vinyl disc records.

But shrinking was a requirement of any material suitable for molding cylinder records because that was the only way that the record could be released from the cylindrical mold. So special compound formulations ("wax" or celluloid) were required to allow the cylinder to shrink a predictable amount while still maintaining proper geometrical ratios in all dimensions. This also required that the mastering be done just slightly "bigger" than the finished product was supposed to be. So if the shrinkage was, say, typically 2 percent (I'm not sure if this is the typical number), the recorded pitch of a "2 minute" record had to be about 2 percent coarser (at 98 lines per inch) and with a slightly wider groove (about 6.012 mil radius) to result in a finished record that would shrink to 100 lines per inch and have a 6.00 mil groove to play on a standard cylinder player with a 6 mil stylus and that used a feedscrew to advance the playback reproducer at a set pitch of 100 LPI. Note that the speed of the recording did NOT require such fudging. Most cylinder records were designed to play at 160rpm. The rotational speed remains the same even when the record shrinks because it still takes the same amount of time for the recorded groove to make one complete revolution regardless of its diameter.

[Steve E.]

This also required that the mastering be done just slightly "bigger" than the finished product was supposed to be. So if the shrinkage was, say, typically 2 percent (I'm not sure if this is the typical number), the recorded pitch of a "2 minute" record had to be about 2 percent coarser (at 98 lines per inch) and with a slightly wider groove (about 6.012 mil radius) to result in a finished record that would shrink to 100 lines per inch and have a 6.00 mil groove to play on a standard cylinder player with a 6 mil stylus and that used a feedscrew to advance the playback reproducer at a set pitch of 100 LPI.

That is CRAZY!!!!! wow.

[shawnborri]

Edison master machines recorded at 97.3333 tpi to make masters for the master, mother, working mold process. I own an original studio recording head and the master feed screw, the one I have though is set up for 194.33333 for four minute, the master machine had a gear on the mandrel and this feedscrew attached at the back of the machine to make four minute records, while the 97.3333 tpi was used for two minute cylinders. The recorders used on professional cylinder (and diamond disc) lathes of the vertical cut kind, were similar to a Columbia floating recorder, except they were heavy with an advance ball to set the depth into the wax. I now have about 30 hours of recording time on this head, learning how to adjust it, and work with different horns. Say a band record, a 5' horn works great, while a 30x6 cone horn with no bell, is good for most instruments, and male vocalists, while a 25x3 works well for female vocal, violin and talking records.