Using an AtTiny85, I'd like to drive an AlfaZeta C30NR flip-disc component to indicate dead batteries for a small project. It is a semi-permanent electromechanical indicator that uses two pins, SET and RESET, and depending on the direction of a current pulse, the state of the component toggles.
In order to accomplish bidirectional current, I could connect an output pin to SET, and another output pin to RESET, allowing current to be tugged in either direction by the microcontroller. It doesn't seem that there would be issues, as the voltage going between the pins would not exceed their specs, and outputting high on both would have a net 0 effect.
In this particular case, it is obviously a poor idea as the current draw required by the component is 250mA where the absolute maximum per I/O pin is 40mA. It actually doesn't seem possible here, as a current divider between the sink output and ground would require a resistor (preventing one output from driving the component with the minimum current while protecting the other), but I'm more interested in the theoretical aspects.
What drawbacks would there be to this configuration? What alternatives would be better design?
Naturally, one option is the C30ND, which uses a common ground for either input pin. I imagine another is to use transistors to convert the outputs into open collectors.
A pulse from PB2 turns the indicator on. A pulse from PB1 turns it off. The pulse must be [email protected] for 1.5ms, but PB2 and PB1 can only receive 40mA. Adding parallel ground, and a resistor on PB1 means that PB2 can safely use a 250mA current, but then PB1 cannot give 250mA of current (it is limited by a resistor and a direct path to ground).