thanks for your input and sorry for my belated reply.
I already read several threads concerning "software reset" after connecting USB and how to prevent this (but I would go for the 10µF capacitor method e.g. described herehttp://electronics.stackexchange.com/questions/24743/arduino-resetting-while-reconnecting-the-serial-terminal
, instead as it is easier and less 'lossy' because no current is drawn during normal operation). Some people write that this sort of reset would only occur on Mac OS X and Linux machines while Windows would not trigger a soft-reset during a simple USB handshake
(= just plugging the cable) but only when trying to actually access the serial port (e.g. via HyperTerminal). Hence, I thought there might be another hardware-related problem which triggers a reset when connecting my Duemilanove board to the PC via USB...
Nevertheless, after finally finishing my project, yesterday, I now found out that there is yet again a problem while un
plugging the USB cable: The LEDs on the Arduino go off for about half a second (despite my 'stabilization' of the 5V line with a 1000µF capacitor) before the external (permanent) power source kicks in... Hence, even if I would be able to prevent the soft-reset during establishment of a PC connection, unplugging the USB cable would definitely cause a hard-reset because of power loss for a certain time.
As I never need to power the board via USB, I now thought of using a little 'trick' to prevent any interference with my external power source: Would it be okay to simply cut the two power lines of the USB cable (VCC / GND) and only leave the data lines (D- / D+) intact ? If I can still transfer data with only those two data lines connected, I guess this would solve the unplugging issue (if it is not ground-related).
Maybe there is somebody out there who already did this and can share his / her experience, so I know if this makes sense or if there is a better way to go.
Thanks for any feedback !