Electronic – Is a crystal needed for burning firmware on AVR

as long as you have the fuses set to use the internal oscillator and you are not using the XTAL pins for any reason on those chips, you should be A-OK. You should not set those XTAL pins to outputs however, but rather inputs without internal pullups enabled (i.e. DDRx = 0; PORTx=0). On a separate note though, I'm not so sure you'll be OK with more than one crystal per chip. The crystals are 'passive' components, and the AVR actually has some circuitry in it for 'driving' the crystal to resonate...

EDIT OK I just wanted to make sure you weren't trying to "share" crystals. As the others have stated, I can't imagine a (sane) reason why you would have to run your chips at the same clock rate. Be mindful however, that some of your register settings (for example timer prescales, baud rate settings, etc) have implicit coupling to your clock rate, and may need adjustment in otherwise identical firmware.

I believe I may have found a two part solution for my particular problem. I currently own an Arduino Diecimila (which has the ATMEGA168 with the problem) and also a new Arduino Nano 3.0. I went ahead and flashed the Arduino Nano with the ArduinoISP firmware. I then picked up 3 28 Pin ATMEGA328P DIP chips for $5.00 each at the local electronics store.

I used the ArduinoISP to flash one of the virgin chips with the Arduino Firmware using the above link with a crystal. I plan to now flash one of the other ATMEGA328Ps with the High Voltage Programmer firmware. This chip along with a bit of wiring on a breadboard should allow me to reset all of the fuses on the original ATMEGA168. Once that is done, I will be able to use the Arduino Nano to reprogram the chip.

Since the ATMEGA328Ps only cost $5 each, my end goal is to setup one 328 as an Arduino ISP programmer, setup the 168 as a HV Programmer, keep one 328 chip in my current Arduino Diecimila (upgrade!!), and I'll have one left over for whatever. I can then either switch chips on the Diecimila or build out a real board for the ISP and HV Programmer boards.

I still have not tested the JTAG ICE MKII yet with a logic/oscilloscope. Not sure if it is the root of the problem or not. I will address this issue after I build the HV Programmer.

I will report back on the success/failure of the HV Programmer. I just need to go get a few more 1K resistors to pull it off. Anyone ever tried the HV Programmer? Any other suggestions? If nothing else, hope this helps someone.