| ??? 08/30/11 11:27 Read: times |
#183536 - Just as there is a timeout Responding to: ???'s previous message |
Terminating the slave response with the 9th bit set should be ok. Especially if the slave always used a fixed value that does not collide with the address used for any other slave in the system - preferably a hardcoded value like 0 or 255, or the slave's own address.
This would work very well, as long as the master side does have a timeout handler, to recover if the last transfer gets garbled. So master can switch data direction instantly when all is well, but the bus gets a short idle period if last byte no longer had the high bit set for some reason. |
| Topic | Author | Date |
| Multiprocessor Communication 8052 | 01/01/70 00:00 | |
| where is bottleneck? | 01/01/70 00:00 | |
| That's the usual approach | 01/01/70 00:00 | |
| one comment | 01/01/70 00:00 | |
| 9th Bit - How ? | 01/01/70 00:00 | |
| how to use bit 9 for data bytes? | 01/01/70 00:00 | |
| one form of 9th bit use | 01/01/70 00:00 | |
| One byte | 01/01/70 00:00 | |
| One Byte !!! | 01/01/70 00:00 | |
| re: 1 byte - MDB | 01/01/70 00:00 | |
| Strong work | 01/01/70 00:00 | |
| Just as there is a timeout | 01/01/70 00:00 | |
| my reason | 01/01/70 00:00 | |
Protocol should preferably support dry-counting for EOP pos | 01/01/70 00:00 | |
| Methods in brief | 01/01/70 00:00 | |
| At least 1 packet less (sic) | 01/01/70 00:00 | |
| Neither! | 01/01/70 00:00 |