??? 05/29/05 02:34 Read: times |
#94071 - Square Root Algorithm Responding to: ???'s previous message |
Neil,
I presume the objective is compute an RMS quickly. Try implementing Newton's algorithm for computing square root. Stop the iteration when - (1) the root no longer changes significantly (you define), (2) implement a fixed number of iterations by doing a bit of pre calculation on paper to know how many iterations are needed over the range of input numbers, Speed conversion of the algorithm by starting the algorithm with a seed that varies according to the input value. This is not a lookup, it simply varies the initial guess. Cheers, Bruce |
Topic | Author | Date |
RMS Calculation | 01/01/70 00:00 | |
Going back to basics... | 01/01/70 00:00 | |
How Many Samples Per Second? | 01/01/70 00:00 | |
Square Root Algorithm | 01/01/70 00:00 | |
RMS for sinusoidal signals | 01/01/70 00:00 | |
Clarification | 01/01/70 00:00 | |
I beg a question... | 01/01/70 00:00 | |
Full o' holes | 01/01/70 00:00 | |
Bang on target...![]() | 01/01/70 00:00 | |
thats true... | 01/01/70 00:00 | |
RMS Value | 01/01/70 00:00 | |
Please elaborate more... | 01/01/70 00:00 | |
Elaboration | 01/01/70 00:00 | |
Correction | 01/01/70 00:00 | |
DC+AC RMS | 01/01/70 00:00 | |
DC contribution to total rms value | 01/01/70 00:00 | |
Thanx Anywayz... | 01/01/70 00:00 | |
square root by Newton's Rule | 01/01/70 00:00 | |
Lookup table | 01/01/70 00:00 | |
better option...as of now | 01/01/70 00:00 |