Screwed up fuses at its worst with an AVR controller.

[TerminalJack505]

It always makes me nervous changing the fuse settings. Especially when changing the clock source and/or programming/debugging interface-related bits. I always double-check everything and hold my breath until I've reestablished communications to the MCU.

I'm glad they changed how the XMegas work regarding the clock source, as you allude to in the video. The XMegas always start-up using the internal clock and switch to the external clock based on _software_ instead of fuse settings.

[englsprogeny]

Mike, It seems to me that I read once on the AVR usergroup that once you set fuses wrong, you are done with that controller. I get your logic that you should be able to reset the fuses through ISP. Are you able to (at this point) read the fuses through ISP? Let me know how this goes.

[realedna]

First time for me to see somebody using a sinosoid waveform as a clock for microcontrollers. That's probably not a good idea, especially for reviving a misfused chip. You should stay in the digital world and use a proper square waveform.

[linuxgeek81]

@realedna Don't worry, I tried a square signal as well, didn't work. Ever noticed how a crystal swings? ;)

[linuxgeek81]

@RSole52 No way, needs a ZIF-Socket for the controller and I've only got one for ATmega64 and ATmega128 controllers. Building something is not worth the trouble here.

[Elamin008]

maaan I wish I'll be like you one day

[linuxgeek81]

@Elamin008 To epically fail? :D

[Flankymanga]

shit happens :/

[linuxgeek81]

@Flankymanga I knew it was the fuses, the fuses for the AT90CAN differ from the regular AVRs, that's why I fowled them up. Now it works as expected. One dead beat, though. In the end we always know better ;)

[orbiter8]

Looks like an absolute nightmare you got there Mike. Hope you get it sorted out soon.

[linuxgeek81]

@orbiter8 Yeah, I threw the bug away and replaced it, got one from another circuit. This was really a dead case I'm afraid oO"