Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
07/11/06 16:31
Modified:
  07/11/06 16:35

Read: times


 
#120028 - I just realize you have a fundamental error
Responding to: ???'s previous message
once the internal watchdog bites, there is no guarantee whatsoever that the uC will function correctly.

fron t6he datasheet
13.3. Power-fail Reset
When a power-down transition or power irregularity causes VDD to drop below VRST, the power supply monitor will drive the /RST pin low and return the CIP-51 to the reset state (see Figure 13.2). When VDD returns to a level above VRST, the CIP-51 will leave the reset state in the same manner as that for the power-on reset. Note that even though internal data memory contents are not altered by the power-fail reset, it is impossible to determine if VDD dropped below the level required for data retention. If the PORSF flag is set, the data may no longer be valid.


What you need to do is to have a 'split' power supply: two diodes, one feeding a smallish capacitor with a bleeder resistor that supply an external watchdog connected to an uC interrupt, the other feeding a large capacitor supplying your uC. That way you can get some time after power fail where the uC has enough Vdd to save the values. If the internal supervisor resets before you have saved all, you are lost.

Erik

List of 10 messages in thread
TopicAuthorDate
help pcf8583 interfacing with C8051f005            01/01/70 00:00      
   switch case??            01/01/70 00:00      
   What about the Bit-banged version?            01/01/70 00:00      
      bit vank version            01/01/70 00:00      
         that is a mistake, you will get faster processing            01/01/70 00:00      
   use CodeArchitect            01/01/70 00:00      
      while code arch code support c8051f005            01/01/70 00:00      
         It is not "code arch", it is CodeArchitect and the            01/01/70 00:00      
   I just realize you have a fundamental error            01/01/70 00:00      
      and I did a thing I often do ...            01/01/70 00:00      

Back to Subject List