Jeff Liebermann wrote:
On Fri, 30 Sep 2011 09:10:01 +0100,
lid (Adrian Tuddenham) wrote:
Any Dictabelt model
I just tore apart many boxes of junk that I haven't seen for 20 years,
but didn't find my Dictaphone or belts. Sorry. I think asking in the
rec.antiques.radio+phono
newsgroup might be more productive in finding someone with a working
Dictaphone machine.
I have a working Dictablet machine (in addition to the one I have
cannibalised to make the archival player). That is how I know the
approximate speeds. What I do not know is the manufacturer's specified
speed. For a number of reasons, the speed I have measured on my
machine may not be the same as the specified speed.
What I really need is the manufacturer's written specification (which
might be found in either an internal works document or a decent workshop
manual). Somebody somewhere must have worked on those machines and will
know what the speed was supposed to be.
I am constructing a playback machine for archival work, which needs to
be able to cope with any Dictabelt. There was a standard speed which
applied to all the normal models and all belts, so as to allow
interchangeability of the belts world-wide.
The basic machine was designed for 60 c/s mains, but there were
adaptations to allow the machines to run at the correct belt speeds on
50 c/s. Some special machines were also designed to run at half speed
for logging purposes . I have already encountered some belts which were
recorded on unadapted half-speed 60 c/s machines working from 50 c/s
supplies - and, no doubt, there will be others which were recorded at
different non-standard speeds for various reasons.
That explains the strange errors in apparent belt speed. I would have
expected South Africa to use the same standards as UK (230V 50Hz).
It did, but they were using a machine which had been adapted to run at
half speed on 60 c/s - and were running it on 50 c/s. The speed was 17%
below half speed.
What I need to know is the manufacturer's specification for the standard
belt speed which was supposed to allow interchangeability between all
the machines. I can then calibrate the speed control on my machine to
show the %age speed variation from the norm, so as to deduce information
about the circumstances under which each belt was recorded.
If that's the situation, you may need to handle a fairly wide range of
belt speeds. I suspect that it might be easier to setup a servo
controlled motor, external belt tensioner, and stylus attached to a
linear actuator. In effect, a belt recorder implimented mostly in
software. That should give you control over speed and head position,
which should also solve the track skipping problem. It would also
allow you to run at some high speed (i.e. 15 ips) and slow down the
playback in software.
My experience suggests that it is better to allow the stylus to follow
the grooves, rather than being rigidly fixed where the groove ought to
be. That way, belts which wander across the faces of the rollers don't
cause the stylus to skip. The leadscrew of the machine drives the
playback head carriage, but the head swivels freely to adjust itself +
or - 1/8" from the central position set by the carriage.
At present I am using the machine's existing motor with the two sets of
stator coils driven in quadrature from a pair of high voltage amplifiers
by signals from a 4-phase oscillator. As well as providing variable
frequency, the system has to control the current through the motor
windings so as to gives adequate mechanical torque but not saturate the
iron and give rise to excessive hum harmonics in the playback head. The
motor will work over a speed range from 20 to 70 c/s if the phasing and
currents can be correctly maintained.
It may be possible to put a toothed wheel and magnetic sensor on the
drive roller shaft, which would be used to control the motor drive
frequency so as to compensate for slippage in the rubber idler and drive
belt and the motor slip frequency. That way, the drive roller speed
would be accurately controlled and the only discrepancies would come
from the very slight slippage of the recorded belt on the roller.
Another way of specifying the speed would be to count the drive roller
revolutions. Off-load, it seems to be spot-on 120 rpm, so perhaps that
is how the manufacturers laid down their standard.
Good luck. Sounds like an interesting project.
It is.
--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk