I can't comment on whether your breadboard programmer circuit is correct because I'm a beginning breadboard user without a lot of expertise. And I'm a beginner with AVRs too.
My perception is -- you are not providing a schematic for it and it is difficult for me to identify the devices from the picture. Your breadboard needs a larger footprint than a small PCB and physically bumping the breadboard could change the circuit or cause it to malfunction when a component comes loose or detaches itself from the circuit. Others in this forum are far more expert than I so listen to them first... Bob smcx wrote: > Well, that might be a problem. But in ponyprog, I always perform a write > without an erase cycle, and I never get a verification error. Nevertheless, > I'll try to make the programmer again and try by tomorrow. By the way, I was > using the programmer constructed on a breadboard. Is it necessary to > construct the circuit on PCB? (Here is a > http://3.bp.blogspot.com/_83WgROTnvFo/SQXziI7buYI/AAAAAAAAAM4/U2FwTh4AOiQ/s1600-h/DSC00993.JPG > picture of my programmer. > > > Peter-35 wrote: > >> smcx schrieb: >> >>> And here is the output from AVRdude (I was using AVR8-Burn-O-Mat as a >>> GUI) >>> >>> C:\WinAVR-20071221\bin\avrdude.exe -q -u -C >>> C:\WinAVR-20071221\bin\avrdude.conf -p m8 -P usb -c usbasp -D -E >>> noreset,novcc -U flash:w:E:\Labsoft\ATMEGA8\main.hex:a >>> >>> >> The -D option your frontend passes to avrdude is the main Problem here I >> think. It disables the erase-cycle avrdude normally performs before >> writing to the avr. When the old program is only overwritten with the >> new one, it is no wonder that the verification fails. >> >> regards, >> Peter >> >> >> >> _______________________________________________ >> AVR-chat mailing list >> AVR-chat@nongnu.org >> http://lists.nongnu.org/mailman/listinfo/avr-chat >> >> >> > > _______________________________________________ AVR-chat mailing list AVR-chat@nongnu.org http://lists.nongnu.org/mailman/listinfo/avr-chat
