Sinclair Scientific

Sinclair Scientific

Sinclair Scientific calculator photographed c. 1974
Type	Scientific
Manufacturer	Sinclair Radionics
Introduced	1974
Cost	GB£49 + VAT
Calculator
Entry mode	RPN
Precision	5 digit significand, 2 digit exponent
Display Type	Light-emitting diode
CPU
Processor	Texas Instruments TMC-0805
Programming
Other
Power supply	4x AAA batteries (Scientific) 1x PP3 battery (Scientific Programmable)
Dimensions	50 by 111 by 19 millimetres (1.97 × 4.37 × 0.75 in) (Scientific) 73 by 155 by 34 millimetres (2.9 × 6.1 × 1.3 in) (Scientific Programmable)

The Sinclair Scientific calculator was a 12-function, pocket-sized scientific calculator introduced in 1974, dramatically undercutting in price other calculators available at the time. The Sinclair Scientific Programmable, released a year later, was advertised as the first budget programmable calculator.

Significant modifications to the algorithms used meant that a chipset intended for a four-function calculator was able to process scientific functions, but at the cost of reduced speed and accuracy. Compared to contemporary scientific calculators, some functions were slow to execute, and others had limited accuracy or gave the wrong answer, but the cost of the Sinclair was a fraction of the cost of competing calculators.

History

In 1972, Hewlett-Packard launched the HP-35, the world's first handheld scientific calculator.^[1] Despite market research suggesting that it was too expensive for there to be any real demand, production went ahead.^[2] It cost US$395 (about GB£165), but despite the price, over 300,000 were sold in the three and a half years for which it was produced.^[1][2]

In 1973, Texas Instruments introduced the TMS0803 chipset, the successor to the TMS0102.^[3] The new chip was highly customisable, with a programmable read-only memory, and variations of the 080x appeared in other Sinclair calculators.^[3][4] The TMS0805 variation was used in the Sinclair Scientific.^[3]

Clive Sinclair wanted to design a calculator to compete with the HP-35 using the 080x series of chips, and despite skepticism about the feasibility of the project from Texas Instruments engineers, Nigel Searle was able to design algorithms that sacrificed some speed and accuracy in order to implement scientific functions.^[4]

The Sinclair Scientific first appeared in a case derived from that of the Sinclair Cambridge, but it was not part of the same range.^[5] The initial retail price was GB£49.95 in the UK, and in the US for US$99.95 as a kit or US$139.95 fully assembled.^[6] By July 1976, however, it was possible to purchase one for GB£7.^[5]

The Sinclair Scientific Programmable was introduced in August 1975, and was larger than the Sinclair Scientific, at 73 by 155 by 34 millimetres (2.9 × 6.1 × 1.3 in).^[7][8] It was advertised as "the first ... calculator to offer a ... programming facility ... at a price within the reach of the general public," but was limited by having only 24 program steps.^[9]

Both the Sinclair Scientific and the Sinclair Scientific Programmable were manufactured in England, like all other Sinclair calculators except the Sinclair President.^[10]

Design

Sinclair Scientific

The Sinclair Scientific was severely limited by the custom Texas Instruments TMS0805 chip used, which could barely handle four-function arithmetic.^[11][4] The HP-35 used five chips, and had a been developed by twenty engineers at a cost of a million dollars, leading the Texas Instruments engineers to think that Sinclair's aim to build a scientific calculator around it was impossible.^[4] However, by sacrificing some speed and accuracy, Sinclair used clever algorithms to run scientific operations on a chip with room for just 320 instructions.^[4] Constants, rather than being stored in the calculator, were printed under the screen.^[4]

It displays only in scientific notation, with a five digit mantissa and a two digit exponent, although a sixth digit of the mantissa was stored internally.^[4] Because of the way the processor was designed, it uses Reverse Polish notation (RPN) for inputting calculations.^[4] RPN meant that the difficult implementation of brackets, and the associated recursive logic, was not necessary to implement in the hardware, but the effort was instead offloaded to the user.^[12] Instead of an "Equals" button, there is an "Enter" button that tells the calculator when a value has been entered, and then the operators are entered in after the operands.^[12] For example, to evaluate "(1+2) x 3", the sequence entered would be "3 enter 2 enter 1 + x."^[12]

To fit the program into the 320 words available on the chip, some significant modification was used.^[4] By not using regular floating point numbers, which require lots of instructions to keep the decimal point in the right place, some space was freed up.^[4] Trigonometric functions were implemented in about 40 instructions, and inverse trigonometric functions are almost 30 more instructions.^[4] Logarithms are about 40 instructions, with anti-log about 20 on top of that.^[4] The code to normalize and display the computed values are roughly the same in both the TI and Sinclair programs.^[4]

The design of the algorithms meant that some calculations, such as arccos0.2, could take up to 15 seconds, whereas the HP-35 was designed to complete calculations in under a second.^[4] Accuracy in scientific functions was also limited to around three digits at most, and there were a number of bugs and limitations.^[4]

Ken Shirriff, an employee of Google, reverse engineered a Sinclair Scientific and build a simulator using the original algorithms.^[11][4]

Assembly kit

The assembly kit consisted of eight groups of components, plus a carry case.^[13] The build time was advertised as being around three hours, and required a soldering iron and a pair of cutters.^[13][14] In January 1975, the kit was available for US$49.95, half the price at the time of introduction a year earlier,^[13] and in December 1975 it was available for GB£9.95, less than a quarter of the introductory price.^[14]

Giant Scientific

A version of the Scientific, with all the same functionality, was made to be 30 by 68 centimetres (12 in × 27 in), and was known as the Giant Scientific.^[5] It was powered by 240V AC, and used discrete LEDs for its display.^[5]

Sinclair Scientific Programmable

The Sinclair Scientific Programmable was introduced in 1975, with the same case as the Sinclair Oxford.^[15] It was larger than the Scientific, at 73 by 155 by 34 millimetres (2.9 × 6.1 × 1.3 in), and used a larger PP3 battery, but could also be powered by mains electricity.^[7][15]

It had 24-step programming abilities, which meant it was highly limited for many purposes.^[16][15] It also lacked functions for the natural logarithm and exponential function.^[16] Constants used in programs were required to be integers, and the programming was wasteful, with start and end quotes needed to use a constant in a program.^[16][17]

However, included with the calculator was a library of over 120 programs that that performed common operations in mathematics, geometry, statistics, finance, physics, electronics, engineering, as well as fluid mechanics and materials science.^[16][18] The full library of standard programs contained over 400 programs in the Sinclair Program Library.^[19]

References

External links

• Reverse engineering the Sinclair Scientific
• Assembly instructions for Sinclair Scientific kit
• Operating instructions for Sinclair Scientific (http://www.datamath.net/Manuals_Others/Sinclair_Scientific_GB.pdf alternative scan)