| +Home | Museum | Wanted | Specs | Previous | Next |
Monroe 920 Desktop Calculator
Updated 3/21/2002
The Monroe 920 is yet another example of Monroe leveraging the electronic calculator expertise of Canon to provide the 'guts' of their calculators for them. The 920 is clearly a near-clone of the Canon 120 and 1200, with the only small differences between the Canon and Monroe models. The only major difference, other than subtle keyboard layout differences and color schemes, is that the Canon machine does not provide a constant (K) key as this machine does.
Inside the machine, all of the circuit boards have Canon logos all over them. Another example of this, although somewhat more subtle, is the Monroe 950, also in this museum. This particular example of the 920 was made in the late part of 1969, and has a maintenance sticker on the back that lists "12/12/69" as the date of installation. Date codes on the IC's in the machine range from 6846 (late 1968) through 6930 (Q3 '69), which seem to correlate with the installation date.
Monroe 920 with the case opened up
The Monroe 920 is a small-scale integration
IC-based calculator. All but one of the 66 IC's in the machine are
Texas Instruments-made DTL (Diode-Transistor Logic) devices in the
SN39xx and
SN45xx-series. A single very early 7400 TTL (Transistor-Transistor Logic)
Quad Two-Input NAND IC was included for whatever reason. This is one of the
earliest uses of a 7400-series TTL device found in an electronic
calculator I've come across so far. The 920 was designed with a fairly small
compliment of different chips. The devices (with the device type in
brackets, and the number of devices in parenthesis following the device
part number) used in the machine are: SN3920(6), SN3925[Quad 2-Input NAND](32),
SN3931[Triple 3-Input NAND](11), SN4553[DUAL 4-Input NAND w/expander](4),
SN4554[Quad 2-Input NAND](12), and the lone SN7400 TTL device.
Close-up of Monroe 920 Circuit Board The 920 uses an acoustic delay line
device for working storage. The delay line appears to be very similar to the
delay line used in the Monroe 950. The
delay line has a total delay of 494 microsoconds, with a tap near
the middle of the line providing a delay of 240 microseconds.
For more information on the use of delay line technology in early electronic
calculators, I recommend reading an interesting article by Nicholas Bodley
on the internal workings of the ground-breaking
Friden 130 calculator. You can jump to this article by clicking
HERE. Like most electronic calculators of its
time, the 920 uses Nixie tubes as the display technology. Each Nixie
tube contains the digits zero through nine, along with a decimal point.
The Nixie tubes are driven by discrete transistor (Toshiba 2SC780) drivers.
It is interesting to note that the Nixies are multiplexed on this machine,
which makes the 920 one of the earliest examples of a multiplexed
non-CRT-based display that I've seen. Another oddity of this machine is
that the anode drivers for each Nixie tube consists of a transistor and a small
signal transformer. The transformer design simplifies the decoding
circuitry necessary for driving each of the twelve Nixie tubes as they are
multiplexed.
The Bottom Circuit Board The calculating brains of the machine are made up of two fairly good-sized
circuit boards stacked one atop the other across the bottom of the machine.
The top circuit board contains the Nixie tube displays and driver circuitry,
master clock oscillator which uses a 2.0MHz crystal for generating the
timing for the machine, and the divider chains that divide the 2MHz clock
down to the fundamental clock rate of 250KHz. A crystal-controlled
oscillator is used because it allows for accurate and stable timing of the
signals being launched into the delay line. Also included on the top circuit
board are the read and write amplifiers for the delay line. The top board
contains 21 of the 66 IC's in the machine. The bottom board seems to contain
most of the calculating and state logic, with the remaining 45 IC's on it.
The keyboard connects to the bottom board via an edge-card connector, and the
two boards are connected together by edge-connectors which are hand-wired
together.
The Top Circuit Board (with Power Supply/Delay Line Module Set Aside) The Monroe 920 operates very typically for machines from this vintage.
It uses adding machine style logic for add and subtract, with a key with
a white "=" legend for adding and one with a red "=" legend for subtracting.
For multiplication and division, the problem is entered an algebraic form,
with either of the "=" keys used to calculate the result. The calculator
uses fixed-point decimals, with a three position switch on the keyboard
panel selecting 2, 3, or 4 digits behind the decimal point. As with just
about all Canon calculators, the "C" clear key clears the entire machine,
and the "CI" key just clears the display, and is used for correcting
input errors. The machine also features a 'push-on/push-off' "K" key
for allowing a constant multiplier or divisor. A View of the Power Supply and Delay Line Module The 920 has no notion
of negative numbers (much the same as the
Brother Calther 412 which is of similar vintage and technology).
A negative result shows up as the 10's compliment of the negative number.
e.g., 10 - 15 results in 9999999995.00 (with the decimal point set at
the '2' selection). Pressing the red "=" key with such a number in the
display will immediately compliment the number (with the example
returning 5.00). The machine does detect overflow, and indicates an
overflow or 'input error' (such as two keys pressed at once) by lighting
a neon indicator positioned at the left end of the display in the
shape of a left-facing arrow (another Canon trademark). When the
machine overflows, it locks up, and requires a press of the "C" key
to continue. The 920 doesn't know that dividing by zero is impossible,
and locks up tight without lighting the overflow indicator. Pressing
the "C" key here will return the machine to normal. Interestingly,
pressing the "CI" key when the machine is churning trying to divide by
zero results in the machine unlocking, but doing strange things
ranging from lighting up all of the decimal points but doing nothing,
to accepting numeric input, but not performing any math operations.
Again, pressing "C" when in this weird mode will restore the machine
to normal.
Detail of Magnetic Reed Switch Keyboard The 920 uses a magnetic reed-switch keyboard that is very reliable, and
was the predominant keyboard technology of the time. The picture above
clearly shows how this technology works. The fact that there is no
physical connection between the key and the switch itself makes these
keyboards last a very long time. This is why many of these very old
calculators still have keyboards that work as they did when they were new,
whereas later calculators that used other forms of 'direct contact'
keyboards have gotten 'bouncy' or have quit working altogether.
The 920 is quite fast for a machine of its vintage.
With a main clock rate of 250KHz, it is substantially faster than
earlier all-transistor designs. Additional chains of counters are
clocked by the 250KHz master clock to generate the various operational
states of the machine. 9999999999.99 divided by 1 takes less and 1/3
second to give a result, and addition and subtraction
give virtually instant results. The machine does not blank the display
during calculation, so some action of the Nixies can be seen on
operations that take longer to perform. The machine also lights up all
decimal points when it is busy calculating.
