AVRISP-MKII timeouts

[Chris McClelland]

Hi Dean,

Been a long time! Still a very happy LUFA user!

I spent a frustrating day yesterday assuming I had a problem with wiring or with my shonky stripboard soldering, but I concluded that the COMMAND_TIMEOUT_TICKS in the AVRISP-MKII clone firmware is too short. The LUFA master branch consistently succeeds in programming my atmega88 designs, but consistently fails to program my atmega162 designs. Admittedly these are from tubes I've had lying around for 14 years, so perhaps the timings have improved on modern devices. I narrowed it down to a very minimal setup (AVRISP firmware with correct board settings and F_CPU running on a fresh Minimus board, and a fresh atmega162 with default fuses, only connections to the programmer and a decoupling capacitor).

Are you happy to get a GitHub PR on abcminiuser/lufa for this? With a bit of work I can find the timeout threshold where my device starts to fail, and set it a bit higher than that.

Chris

[Chris McClelland]

Ooops, ignore me. Setting the avrdude cycle duration to a shorter value fixed it.