idea for improvement

[spiritracker]

Love the minipov. I'd like to see more color, though. Would it be possible to create a minipov4 with multi-color LEDs and modify the software to allow assigning of color to the individual blinks?

thanks,
Greg

[sparr]

Yes, it would. Get to it.

Seriously though, I am working on a design for a narrower RGB minipov variant, for poi-like usage. Unfortunately my PCB-making skills leave a lot to be desired.

[mtbf0]

keep in mind that an rgb led has four leads, one of which connects to ground and the other three of which must be connected to pins on your microcontroller. that means that to control eight leds you'll need 24 outputs. also the green and blue leads will probably need upwards of, say, 3.4 volts. not really a project for an attiny2313 with a pair of aa's.

take a look at the spoke-pov. 32 leds through shift registers.

[sparr]

I am considering a few options for the 8xRGB setup. Three shift registers would work, but would slow down the refresh significantly, which becomes a factor when trying to do PWM. I am leaning instead towards a 3x8 multiplexed configuration, with the AVR directly driving the anodes of each 'row' and driving an open sink FET for the cathodes of each 'column'. This allows for far faster updates (three PORT* writes instead of three SPI transfers) and eliminates components (one quad FET and no resistors*, instead of 3 shift registers and 3 resistor networks).

* My naive experimentation concludes that the AVR alone works for LED current limiting when driving a single LED from a single pin. With 3 column multiplexing that will be a max of 33% duty cycle at ~40mA, which any 20mA-nominal LED should handle just fine.