Re: L293D
I'm not an Atmel microcontroller expert, but here's what could be happening, if you Atmel were a PIC *.
- Most microcontrollers have tri-state digital I/O pins. The states are high, low and high-impedance (aka high-Z). When the microcontroller wakes up, the pins are usually in the high-Z state.
- Since pins are in the high-Z state, the microcontroller is not driving the L293D.
- Something else is driving the L293D. It could be, for example, a stray current on the board.
Would this be consistent with your observations?
What happens if physically remove the microcontroller and power up the circuit?
There's a Yahoo group, which has good discussions of the motor control hardware:
**broken link removed** OSMC stands for Open Source Motor Controller, which folks in this group have actually built to high standards.
------------------------------
* I'm also making some assumptions about your procedure and schematic. It would help if you post your schematic of snippets from it.