Theatre Design & Technology - Feb 1969 - 6

disconnecting a Jones plug between the keyboard and the
punchers and replacing it with a special harness that connected it directly to the readers. In this way accurate copies
could be made at high speed.
Pulse Recording on Magnetic Tape. A second major aspect
of the programming concerned the insertion of properly synchronized cue pulses on a second track of a magnetic tape
recording. (The first track was used for the audible program
material.) Th is was very simply done by keying a 1000 Hz
fixed frequency oscillator into one of the recording inputs of
a standard stereo tape recorder while listening to the program material that had previously been recorded on the other
track.
Keying the cue pulses was not as simple an operation as it
might first appear. To begin with, there was a limit to how
fast the readers could execute a read cycle. In the machines
available to us the cycle time was about 1/ 16th of a second so cues had to be separated by at least this much
time. We needed a means of keying the oscillator that
would limit the maximum number of cues per second to less
than 6, which would give a pulse of controlled duration but
which could be initiated instantly by an operator.
The problem was solved very simply by using the reader itself. The readers, I ike the punchers, were equipped with a
fast acting solenoid tripped clutch which would normally allow only one revolution unless the solenoid continued to be
to be energized. The clutch shaft was equipped with a commutative type 7 -position rotary switch. Since one revolution
of the clutch shaft was equivalent to one reader cycle, it
meant that by definition one segment of the rotary switch
could be used to generate pulses of approximately 1 / 1 7 x
1/6 or 1/ 42-second duration that could not be more closely spaced than 1 /6 second. In recording, the operator simply
pressed the armature of the clutch actuate solenoid as he
listened to the previously recorded music or other material
in order to record clean 1000 Hz pulses of proper duration
and separation.
Program Playback: Initiating th e Cue. In performance
there were two ways of initiating cues, i. e. advancing and
reading the punched tape. The first of these was the playback of the 1000 Hz pulses recorded on magnetic tape. Two
systems for converting the 1000 Hz pulses to 18 VDC to
operate the reader actuate solenoids were tried. The simplest
scheme (the one finally used) was to rectify the amplified
output of a standard 30 W audio amplifier and use this to
operate a D.C. relay. Though this performed fairly well, luck
would seem to have contributed substantially. To begin with,
any audio signal of sufficiently large amplitude could have
operated the relay. Thus the system was sensitive to any
large electrical disturbance: switch transients, power line
surges, etc. Second, it was sensitive to amplitude variations.
Obviously the level of recorded pulses had to be monitored
fairly carefully at the time of recording but any of a number
of factors could cause level variation in playback. Third, it
was difficult to locate a relay sufficiently sensitive to follow
the rapid-fire short duration pulses and yet handle the
relatively heavy current of the solenoids. 1
The second means of initiating a cue was from the Stage
Manager Control Unit which contained some relays and two
push buttons that could be operated during performance. One
of these was the "Initiate" button which allowed the readers
to go through one cycle. As with the punchers, a "one
1. All of these problems were to have been solved by a tuned frequency
limiting amplifier operating an S.C.R. Unfortunately an unanticipated
ground loop burned out the entire unit shortly before opening but bench
tests proved quite encouraging. The amplifier was impervious to extraneous Signals in excess of 400% of the signal pulse. One caution
should be noted about S.C.R. 's, however. In D.C. circuits S.C.R.·s are
self latching, so some means of momentarily shunting the anode-cathode
circuit would have to be provided. Also. because of the self·latching
characteristic we probably should have introduced a slight delay in the
S.C.R. gate circuit. again with the goal of reducing the possibility of a
momentary transient's firing the S.C.R. As mentioned, this circuitry has
yet to be proved under performance canditons.

One of two specially modified tape readers. Note the rotary
switch on top which integrated reader and cue storage functions.

shot" configuration was used to prevent slipping into the
next cue. The other push button was an "Automatic Cue
Override." This was essentially a panic switch which would
cut off the pre-recorded cues from the magnetic tape while
allowing the stage manager to continue to initiate cues
manually.
Counting the Cues. Inevitably there had to be a means of
counting events happening as rapidly as 6 per second.
Hence, one of the functions of the Stage Manager Control
Unit was to supply a count output to actuate a counter. The
counter consisted of a series of stepping switches connected
to indicator lamps. Actually there was no particular logic to
this, apart from their availability. Probably a simple solenoid
operated Veeder counter would have done just as well. However, the unit did have the advantage of a high potential
count rate (over 1000 cues in 49 seconds) and since the
input to the "units" counter was an S.C.R., the counter
would register pulses of very short duration. The counter input was arranged in such a way as to be easily reconnected
into the Tape Punch Group to allow cues to be counted as
they were written.
Reading the Cues. The readers required no substantial
mechanical alteration. The sensing switches were rewired,
however, to provide simultaneous output from all five hole
positions instead of the sequential arrangement used originally. The rotary switch was still very useful. It provided the
proper electrical integration between the cue storage (see
below) and the reader cycle. These units were extremely
reliable but proper initial adjustment of the tape guides and
back-up plate was found to be critical.
Cue Storage. One characteristic of the reader that caused
additional compl ication was that the fingers of the sensing
switches pressed against the tape only momentarily during
the read cycle. Thus, the projector shutters, for example,
would have been open only for the 1 /6 second duration of
the ready cycle. To store this information I resorted once
again to the use of relays wired in a self-locking configuration. Each of the sensing switches was connected to the
coil of a relay in a separate rack. The sensing switches
themselves were only energized by the second segment of
the reader rotary switches. The first segment of the rotary
switch was reserved to operate a relay that would open the
latching circuit of all the relays in the memory bank. Thus,
at the beginning of each read cycle the memory would be
erased and then reset by new information read directly from
the tape. Since the time lapse between erase and reset was
about 1 /40 second, no flicker could be observed in a shutter that was programmed" on" continuously.

[]ITO THEATRE DESIGN AND TECHNOLOGY

FEBRUARY. 1969

Table of Contents for the Digital Edition of Theatre Design & Technology - Feb 1969