??? 07/27/07 16:05 Read: times |
#142447 - Missing clock Responding to: ???'s previous message |
"1. I can put an N-bit (4096?) counter on the oscillator"
No problem as long as the oscillator starts. One nice feature I like to see in a reset circuit is the missing clock detector. If the cpu isn't in a deliberate power down condition, but the oscillator supplying the system clock stops, then the reset should trigger after some set time period (in microseconds). The N-bit counter wuld fit nicely with a missing clock, but the counter itself must have another time source. Some controllers like the MSP430 and ARM7/9 have multiple oscillators to ensure the system clock always has a source, even if the primary crystal oscillator fails. This is nice if only to have a way to log the xtal failure with some indicator light. This also handles the problem of the internal watchdog counter losing its clock so it never times out. "3. configuration bit set the polarity" Not terribly useful. Most ICs use an active low for the external reset, and supervisor ICs usually have an open collector low reset output. I can't recall ever using an 8051 in a design where the reset polarity was an issue. |
Topic | Author | Date |
Open questions on reset improvements | 01/01/70 00:00 | |
do you have fuses? | 01/01/70 00:00 | |
level | 01/01/70 00:00 | |
SiLabs | 01/01/70 00:00 | |
opinions | 01/01/70 00:00 | |
RTP? | 01/01/70 00:00 | |
RTP = Erik\'s place (look at his \"signature\") | 01/01/70 00:00 | |
Geography | 01/01/70 00:00 | |
Research Triangle Park | 01/01/70 00:00 | |
Missing clock | 01/01/70 00:00 | |
also in \'51\'s | 01/01/70 00:00 | |
(some?) SILabs '51s has that one | 01/01/70 00:00 | |
It all depends ... | 01/01/70 00:00 |