I am building a 16x16 single color (blue) LED matrix. I have purchased several MAX 7221s for this and future projects, but now I'm stuck on a conceptual problem:
Should I be controlling this 16x16 matrix with four MAX7221s as if it was really four 8x8 matrices or does cascading two MAX 7221s allow me to address the entire 16x16 matrix? If it's the former, what's the point of cascading the chips together via the method in the MAX72xx data sheet?
What I was expecting was: to cascade two MAX7221s, giving me 16 digit and 16 segment lines by which to control the muxed 16x16 matrix. I have taken a short glance at the LEDControl arduino library I plan to use and the wording seems to agree with my idea. However, I'm reading about several projects where they mention using one MAX7221 per 8x8 matrix in the overall display. It seems like this type of configuration would require more work in the firmware to, say, turn on a whole row of LEDs because I'd have to also specify which quadrants we mean to turn on the row in.
So, which is correct? Does cascading two MAX7221s let me control a muxed 16x16 matrix, or do I need to cascade four MAX7221s and 'fake' a 16x16 matrix by controlling four independently muxed 8x8 matrices? I am obviously missing some detail along the line and would hate to have to build an entire 16x16 matrix just to find out I'm wrong and need to break it up into four 8x8s.