Home
New Stuff
Author
Links
Guest book
Web-log
Adverts
Bric-a-brac
Calculators

Components

Ephemera
Events
For Sale
Glossary
History
Hit or Miss
Radios
Transport
Ultra
Valves
Wanted
Metal Puzzles

Clocks
 

Calculators: Handheld: Sperry Remington 803-B 

Size (approx):

118mm x 180mm x 50mm (max) (w,h,d)

Weight 500g excluding batteries. 

Power: 7.5V DC, 5 x AA size batteries.  Accepts an adapter (1.2W) through a top side socket on the left.  There is a red ribbon pull to aid battery removal.  The on/off switch is on the left hand side just below the brand label.
Case: The main case is a two-piece off-white glossy plastic construction which is heavily wedge shaped.  A large front panel of matt black plastic covers the whole front.    At the top, in its own recess is a silver printed label with the brand and model number.  The neutral plastic display filter is deeply inset and tilted. The keyboard is surrounded by a brushed aluminium panel, printed with black text for the switch labels.  They keys are long travel, hollow sounding but easy to use.
Display: 8 digit blue VFD display with a ninth digit for negative and error indicators
Features: Four function with switched constant and switched fixed decimal modes.  Semi-reverse polish notation logic.
Age: 1973
Manufacturer: Remington Rand Office Machines Division, made in Japan.  Serial number 923412 on battery compartment cover. 
Comments: Big wide calculator that just about fits in the hand but is more likely designed for continuous desktop use.  at 1.2W though it would have eaten batteries at a high rate.  Unusual early "small zero" (see below) display notation but the display is bright and easy to read.  The logic is not so good with no recovery, archaic semi-RPN input and several bugs.  Possibly similar to some of the Casio AS-8 series calculators.

Components: Main Board:
1 x cpu: Hitachi HD3276P 3B (February 1973), 28 pin DIL  0.6" width black plastic
1 x IC: Hitachi HD3252P 3F (June 1973), 16 pin DIL, 0.3" width black plastic
1 x IC: Hitachi HD3113P 3D (April 1973), 16 pin DIL, 0.3" width black plastic
1 x IC: Toshiba TM4352 3-E (May 1973), 16 pin DIL, 0.3" width black plastic
9 x 1 digit VFD  single glass tubes
9 x diodes
11 x resistors
7 x resistor arrays
Keyboard:
4 x capacitors
Power Supply Board:
6 x transistors
8 x diodes
13 x capacitors
13 x resistors
1 x transformer module; Fuji MCT 0604 3D-04m (date code April 1973)
Boards: The keyboard assembly (8H-E4A) is fixed to the front of the case using sturdy screws and plastic brackets.  The main cpu board (8H-1B) sits lose in the case and is connected through a 24-way gold plated connecting block.  A third, power supply board (8H-SC) sits lose on top of the battery compartment.
Construction: Remove the four screws accessible from the back.   The front section will then lift off easily, hinging to the right

Logic comments: The Clear key (C) is used to clear an input number and the (AC) key to completely reset the calculator
Overflow on the input of a number is suppressed, keying in a ninth digit is ignored
Overflow shows zero and "E" in the far left (ninth) digit and is not recoverable
Divide by zero shows zero and "E" in the far left (ninth) digit and is not recoverable
The selectable constant (K) function operates on multiply and divide only
Negative numbers are shown with a "-" in the far left (ninth) digit thereby allowing full eight digit negative numbers
The logic is semi-RPN: to do 3-6 key in (3)(+=)(6)(-=)
The results of divisions in floating mode do not have suppressed trailing zeros hence (1)(/)(2)(+=) will give "0.5000000"
You can switch between floating [F] decimal mode or fixed 0/2/4 digit
It suffers the divide to negative zero bug: with (K) switched on and (F) mode, key in (1)(+=)(2)(-=) to give "-1" then (/)(1)(0)(+=)(+=)(+=) etc. to eventually get "-0.0000000"
It suffers the pseudo fixed decimal bug: in (F) mode key in (1)(+=)(.)(0)(0)(0)(+=) will give "1.000" which remains a three digit fixed decimal number until you need more or use divide

display

The image left shows the front of the case removed and the keyboard connecting block (green block to the bottom right) unplugged.

As would be expected with the age of calculator, loads of ICs, early single digit VFD tubes and no expense spared in the construction.  The display is held together by strong metal plates and even the black and red cables coming from the adapter socket have their own board plugs.

However, it is surprising that the boards are lose in the case and only held down by some plastic pillars on the front section..

The board design numbers with there circled revision letters tend to imply that this was a Casio built calculator.

Note the unusual way that zero is represented - by only using the lower half of the digit.  The main theory about this is that it allowed any particular segment to stop working and you could still identify the number.