Go Down

Topic: Problem requiring reflash? (Read 233 times) previous topic - next topic

FrederikJ

I'm having a problem with some AVR 168p's which sometimes requires a reflash to perform properly.

The problem seems to randomly appear from time to time. I have several 168p's in a system and when I turn the system on, each should seperatly generate a pulse on an LED, to verify that they have been turned on an initialized correctly. Once in a while, however, this is not the case. I've then tried removing the card with the given controller, and powering it on seperatly from the others. Still no response. I tried measuring voltages around the board, to make sure that the IC was being sufficiently supplied. I found no discrepancies.

However when I reflash the software to the card, it works flawlessly again.

I looked through the code and made sure that there is no writing being done to neither the EEPROM or Flash memory. So in my mind there should be no difference between turning the power off and on again, and completely reflashing the card with the same software.

This is becoming a major source of frustration for me, as it seems completely random and seemingly impossible to track down the error.

Does anyone have any experience akin to mine or any suggestions for tracking down the error?

Please let me know if any additional information is required. I'm sorry if this was posted in the wrong forum, since parts of my question talks about my specific project, but I felt that this was more heavily related to the IC itself and the project was just there to provide context.

AmbiLobe

Wires are the first place to look.
Historically, discrete wires have caused more problems for circuits than all integrated circuit transistors and wires combined. Use an ohm meter to check your wires.
I am going to get going.

Go Up