Power-up problem

Go To Last Post
3 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi.

I have two designs using a 2323 and a 2343, and in both cases I have some "boot" problems. When the circuits are connected to a clean power supply, everything starts up nicely. But when I try them in the environment they are ment (with an ugly power-supply), they sometimes have trouble starting up. When they first start, they work perfectly, no matter what I do to provoke them, so it seems to be a "booting" problem. I also have activated the watchdog, but it does not seem to help.

I have been thinking about a power-up reset solution using an RC network (ref. previous post on power-up reset), but would like to know what makes the AVR "hang" like this?

Regards,
Ole-Johan Ellingsen

admin's test signature
 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Ole,

You can refer to previous postings on this argument on this forum, I use 2343s clocked by internal RC oscillator, and they hang on power-up, as you can see at the ERRATA SHEET in the section DEVICE of this site.
No, I just tried power resets, I tried a DS1233 yet ( but it means recover from power failure, not cares about micro's failure), no solution seem to come for now.

Let us know if something can cure this

Fabrizio

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

AVR parts are well known for requiring external reset circuitry to power up correctly (despite Atmel's claims to the contrary). Our testing with several AVRs and various circuits show they literally "go crazy" at low voltages and execute a lot of unintended code. The longer your design spends in the 1 - 2.2 volt range, the more likely your AVR will not power up (or down in the case of EEPROM corruption) cleanly.

An external supervisor IC, properly connected and rated for the right voltages should solve your problem. Although we've still seen EEPROM corruption even with an external supervisor IC, we've never seen the device not power up. Is there anything else in your design that could be randomly causing power up problems? For example, I've encountered LCD initialization routines that were written incorrectly and got stuck reading the busy flag of the LCD indefinitely.

Again, this is something that countless other manufactures have solved with a solid on chip BOD, how come Atmel can't manage this too with the AVRs?

admin's test signature