Commodore SR8120

		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 Caravan

Calculators: Handheld: Commodore SR8120 (aka SR8120D)

		Size (approx):	75mm x 145mm x 24mm(max) (w,h,d) Weight 114g excluding batteries
		Power:	9V DC, using 1 x PP3 size battery. Does accept an adapter (707 or DC-640 for USA, 708 for England and 709 for West Germany, 3mm plug, centre positive) through a socket on the top side on the far right. The manual recommends to remove battery before using adapter.
		Case:	Two piece matt plastic case in dark brown and beige. The keyboard surround is in matt black whilst the panel below the display is rough black plastic. The top area houses a black printed metallic brand sticker that sits in its own recess. Typical key colour scheme from Commodore and the small keys are very squishy but work well enough. Additional key functions are have green painted raised labels. The red plastic display filter is flat and heavily inset so that the case forms a light screen. Display is very clear and large though rather limited in viewing angle.
		Display:	12 digit red LED with bubble lens, the three on the far right are slightly smaller than the rest. 8+2 scientific mode. The twelfth digit is used for negative and error indication. There is a linear full length red magnifying lens.
		Features:	Standard four functions, logs, trigs, nth root, powers, pi, change sign, degree/radian conversion, double two-function memory
		Age:	1976
		Manufacturer:	Commodore Business Machines Inc., Palo Alto, California, made in USA, serial number 098880.
		Comments:	One of a large range of three scientific models (the others having 9 or 14 digits) in the "7 and 8" series case. The logic is quite sound with good precision and no major bugs. However it is let down by lack of recovery, basic memory and limited root/power function. On top of this the total lack of constant and cancelling of the shift mode is annoying

Components:	1 x cpu: Commodore GHU-03A 7549 (week 49 of 1975), 40 pin DIL, 0.6" width, grey ceramic 1 x IC: Commodore GHU-01A 7545, 16 pin DIL, 0.3" width grey ceramic 2 x IC: ITT 546A-5N 7531, 16 pin DIL 0.3" width black plastic (display drivers) 1 x 12 digit LED single unit bubble lens display 1 x capacitor
Boards:	The main CPU board (PCB 200800 Rev A TO-ITT ) rests above the keyboard assembly (KB-6373-01) loosely and is connected through a 21-way ribbon cable. The latter is held in place by four plastic lugs to the front.
Construction:	Remove the two screws from the back of the case. Gently push the two lugs in the lower section cuts outs and squeeze off the back. There are two inner lugs that can be seen from within the battery compartment. The back comes off fairly easily.

Logic comments:	(C) is used once to cancel the last number entered and a second time to clear the whole calculator, so it is really a (CE/C) key
	There is no constant on any function
	Scientific mode uses 8 digit mantissa and two (smaller) digit exponents
	Input overflow is suppressed, keying an ninth digit is ignored
	Negative numbers are shown with a "-" in the far left (twelfth) digit to allow full eight digit negative numbers.
	Overflow and underflow shows just "E" (negative or positive) in the far left (twelfth) digit and is not recoverable
	Divide by zero shows just "E" in the far left (twelfth) digit and is not recoverable
	The change sign function (+/-) can be used in mid number entry but not before
	Higher functions are accessed by a two key sequence using the (F) key which is not indicated and not cancelled by keying (F) again
	The two memories can be used to store (STO1) and recall (RCL1) only
	Overflow in memory is not allowed and retains the original number
	The (MR) key recalls the memory, the (MT) key recalls and clears the memory
	Memory store is indicated by the far left (ninth) digit's decimal point
	It is possible to store an overflow in the memory and later recover it by continual division by 10
	Powers can be seen to use the intervening logarithms to calculate the answer. Keying in (2)(Y^x)(1)(0)(=) shows "0.6931471" after the (Y^x) key which is the same as keying in (2)(ln) - then you would key in (X)(1)(0)(=)(F)(e^x). The same (with a division) goes for the nth root key. Whilst this works fine, negative bases are not permitted.
	The (d-r) key is used to switch between degrees and radian mode, the latter of which is indicated by the decimal point of the far right (first) digit. If there is a number of the display at the time this is also converted.
	The calculator has ten digits of internal precision/accuracy. However for powers it is nine, trigs are eight (except tan > 89°) which is six digits.
	No bugs found


The manual measures 71mm by 135mm and is printed in dark blue only. In English only it covers three models, the SR8120D, SR8140D and SR890D which have the same functionality but 12, 14 and 9 digits respectively. With 36 pages in all it has plenty of room for descriptions and instructions. Quote: "Thank you for selecting our new scientific calculator. It represents the finest achievement in solid state large scale integrated / metal oxide silicon technology."	This calculator-sized box measures 80mm x 150mm x 27mm (H x W x D) and is printed in three colours with a "jazz-tech" image. This particular example has printed "model" and "Serial No." on the top flap along with red stamped text. Made in the USA.	This image shows the main cpu board pulled away from the keyboard assembly - easily done as the long ribbon cable allows good access. Components (and hence cost) are kept to a minimum with the CPU, two display drivers, one extra IC, display and one capacitor. For its age this was quite advanced. The board really didn't need to be this long, it just made mounting easier.