| ??? 08/08/04 12:08 Read: times |
#75639 - Triple redundancy? Responding to: ???'s previous message |
What about implementing an algorithm in 3 similar micros (preferably each from a different manufacturer to avoid repeatable manufacturer-created bugs) and pass each output through a "voting gate" of (A&B)|(A&C)|(B&C) ? Any two zeros = 0, any two ones = 1. Probability that one fails is real. Probability that two fail is minimal, but of one of two fails we don't know which one failed. Probability that two out of three fail in the same manner is marginal. |
| Topic | Author | Date |
| Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| Triple redundancy? | 01/01/70 00:00 | |
| Triple redundancy_Of Course! | 01/01/70 00:00 | |
| RE: Triple redundancy? | 01/01/70 00:00 | |
| RE: Triple redundancy? | 01/01/70 00:00 | |
| RE: Triple redundancy? | 01/01/70 00:00 | |
| Single point failure | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| RE: Reliability of BIT flags | 01/01/70 00:00 | |
| Correctness of any signal from outside | 01/01/70 00:00 | |
A sane portion of paranoia | 01/01/70 00:00 | |
| Thanks - The thread can close | 01/01/70 00:00 |