From: Patent Tactics, George Brock-Nannestad
Hello, John Erdmann asked, and Juan Moreno, Lou Judson, Steven Smolian, and
Bob Olhsson replied - I feel provoked to do the same.
> Hello Everyone,
> With regard to the three record speeds, is it just a coincidence that
> 33 + 45 = 78? Can anyone enlighten me as to the history and/or the
> physics behind these choices?
I shall reproduce my reply to the same question in New Scientist 1 October
1994 p. 85
"From 1894 to around 1930 there were many different record speeds ranging
from 65 to 90 rpm, each case being a compromise between the playing time and
tneed for a clean cut in the original wax. The Victor company used 76 rpm for
many years for its recordings but instructed buyers to reproduce at 78 - the
record's durability was improved that way. The standard of 78 rpm arrived by
default, although the actual speed depended on the electrical mains
frequency. Constant linear speed, or varying the rpm was commercialised but
did not prove to be a success (until the arrival of the CD).
The speed of 33 1/3 was introduced in 1927 after theoretical analysis of
the compromise between signal-to-noise ratio and playing time (3 minutes per
radial inch) by J.P. Maxfield of Bell Laboratories for sound films produced
on the Vitaphone system. And it was a professional de facto standard before
it became commercialised by CBS in 1945. It has been suggested that 78 minus
33 equals 45 was the reason for the emergence of 45 rpm records but, in fact,
Maxfield's analysis still applies: the 45 "single" was RCA's equivalent of a
10-inch, 78 rpm record, only smaller."
----- to avoid later discussion about 78, 77.92, and 78.26 I would just
- cogwheels can only have a whole number of teeth, which make up the
fractions when calculating the transmission ratio.
- synchronous motors can have any number of poles
- a 60 Hz motor with two poles runs at 3600 rpm
- a 50 Hz motor with two poles runs at 3000 rpm
- gearing will reduce the speed to as near 78 rpm as you can get, and in the
case of 50 Hz it comes out as 77.92 and at 60 Hz it comes out at 78.26. Both
sets of decimals are rounded.
You can also use slow-speed, multipole synchronous motors.
----- in the semi-old days speeds were measured by stroboscopes: the light
source would light up twice per period, i.e. 120 flashes per second for 60 Hz
and only 100 flashes per second for 50 Hz mains. This means that with a
stroboscope having 77 stripes you would obtain 77.92 when there was no
apparent movement of the stripes. Similarly you needed 92 stripes for 60 Hz
to get 78.26.
33 1/3 rpm required 216 stripes at 60 Hz but only 180 stripes at 50 Hz
45 rpm required 160 stripes at 60 Hz, but when you used 133 stripes at 50 Hz
you obtained 45.11 - so US 45s were reproduced faster in Europe, probably as
a revenge for all the US 78s being reproduced too slow in Europe.
The only place where these speed differentials matter is where time
accumulates, such as at the end of a half-hour radio program. You want
neither over-, nor under-run - silences are banned on the radio.