AT90S1200 with 00 in signature bytes

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

This one is funny. In the course of making my own LPT based serial programmer software for AT90S1200, I emited an "invalid" command (unfortunately I "lost" the command). Since then, this particular device always returns the signature bytes all as zero. The device works correctly, I can program it and read from it without any problems, but the sig bytes are always read as 0.
Anyone has any idea about what happened? This one is just for curiosity, since the device works perfectly.

Embedded Dreams
One day, knowledge will replace money.

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

sig damaged, device works 100%, no problem, except with some programmers, won't work, you have to overrule the device. Sometimes the sig kan damagae by have a load on the ISP i/o's and start dumping data. A bit strange.

admin's test signature
 

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

Seems that the signature is not hardcoded but stored in Flash or EEPROM memory. There it can probably be damaged by all events that are dangerous for user data too (running the device with low VCC, reset while write, etc.).

I had a 1200 with signature 00h 90h 01h. Some times later it return 00h 00h 00h and now it is completely dead.

Possible reason: I have sometimes connected the programmer while AVR power was switched off. If one of the programming lines is high, it sources current into the AVR Pin through the Pins positive clamp diode into target system VCC. In my case this was a significant load (Target system with AVR + Programmer (target powered) = 20-30mA@5V). Atmel does not specify maximum clamp diode current. Therefore I think this situation should be avoided.

Micha

admin's test signature
 

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

In one application note from ATMEL (the "zero cross detector"), they mention that the current on these diodes should not go beyond 1 mA.
I have already programmed my sig-all-zeros device several times and so far it continues to work perfectly.
cheers

Embedded Dreams
One day, knowledge will replace money.

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

I have a similar problem, with two at90s1200. It works fine the first time, but after I shut the power down it stops responding. And even more wierd is the fact that beside the sig bits being zero, I can read the device but I can't write it. I know that this sounds like the lock bits being set, but I have been very carefull not to modify them, ( I only did serial programming using ATprog and sercon2-mini). I used the same procedure a lot of times on a 8535 and it worked fine...until this one erased it's device code too, but this one is capable of writing. Ps I am using a 11.0592Mhz XTAL. Could this be too fast for serial downloading?
Thanks