14-04-2020, 06:09 AM
When I did my electronics degree from '74 to '77 we were not allowed to use a calculator in exams until the final year. That was really because it was only then that a significant number of students actually had a calculator. So it was log tables, and techniques that are best forgotten. Like filter synthesis by doing long division of complex polynomials in an exam.
It was only post degree that I bought an HP34C. That came into its own because it had numerical integration and solving for roots of a function, which saved my bacon when the only other option was punch cards on the University mainframe.
Going back to Sinclair for a moment, one of the guys bought a QL ("Quantum Leap"). Like any new toy he put in a test programme that he could check the result with a calculator. The result was tantalizingly different. Turned out that common constants were blown into ROM (like e, pi etc). One of those was root(2), and whoever filled in the ROM data put a slightly incorrect number. If you put in root(2.0001) the machine used an algorithm to calculate the result, and got it right. AFAIK Sinclair never owned up to that or corrected it.
It was only post degree that I bought an HP34C. That came into its own because it had numerical integration and solving for roots of a function, which saved my bacon when the only other option was punch cards on the University mainframe.
Going back to Sinclair for a moment, one of the guys bought a QL ("Quantum Leap"). Like any new toy he put in a test programme that he could check the result with a calculator. The result was tantalizingly different. Turned out that common constants were blown into ROM (like e, pi etc). One of those was root(2), and whoever filled in the ROM data put a slightly incorrect number. If you put in root(2.0001) the machine used an algorithm to calculate the result, and got it right. AFAIK Sinclair never owned up to that or corrected it.
