| ??? 02/03/11 15:02 Read: times |
#180981 - I didn't employ anything Responding to: ???'s previous message |
I just got the task to write new PC software for existing hardware, and did not have any fun at all with the task since any single collision did result in a 100% requirement for powercycling the unit - the individual boards didn't have any watchdog.
By the way - careful with using the name "Max". Maxim have a large number of components in their portfolio. Many of them don't have anything with RS232 or serial communication to do. Your main problem is garbled data accidentally being accepted as valid data, because the two units starts to send at the same time, resulting in valid stop bits. And you can't rely on parity either since two toggled bits will result in same parity. The only protection that gives a reasonable probability of catching errors is if the two units sends data with 16-bit or better CRC. A larger CRC allows larger burst errors while still guaranteeing that the error gets caught. |
| Topic | Author | Date |
| Paralleling Max232 output | 01/01/70 00:00 | |
| Revommedation On Paralleling Outputs | 01/01/70 00:00 | |
| Multiple Mux | 01/01/70 00:00 | |
| Adding another brain is another option. | 01/01/70 00:00 | |
| RS232 is *not* a bus | 01/01/70 00:00 | |
| Two UART | 01/01/70 00:00 | |
| how about a hub | 01/01/70 00:00 | |
| Is there really such a beast? | 01/01/70 00:00 | |
| external UARTs | 01/01/70 00:00 | |
| There once was... | 01/01/70 00:00 | |
| use 422/485 | 01/01/70 00:00 | |
| Can work well but big potentials for trouble | 01/01/70 00:00 | |
| Thanks, Peter.. | 01/01/70 00:00 | |
| I didn't employ anything | 01/01/70 00:00 | |
| some comments | 01/01/70 00:00 | |
| You really need reliable protocol | 01/01/70 00:00 | |
| collision detection | 01/01/70 00:00 | |
Collision avoidance | 01/01/70 00:00 |