I have a 4 x 8
matrix of RGB 8-segment displays, (short for a 7-segment 'digit' plus a dot pixel, for which each segment (as well as dot) contains: 1x red LED, 1x green LED, and 1x blue LED).
To drive them, I will be using the TLC6983
[DataSheet]
. This driver multiplexes the LEDs using 16 scanlines, however, I need to determine the scanline groupings.
Question
What are the tradeoffs between scanline configurations when dividing the matrix into differently shaped subsets? Does simultaneously enabling LEDs localized in a compact area (and then not at all for an increased duration) improve or worsen characteristics such as:
- Thermals
- EMI
- Anything else
Example scanline configurations
qty | spatial relation |
shape | @ | segments / digit |
---|---|---|---|---|
2 | adjacent | rows | @ | 1 |
4 | adjacent | columns | @ | 1 |
2 | adjacent | digits | @ | 8 |
4 | distributed | rows | @ | 1 |
4 | distributed | columns | @ | 1 |
2 | distributed | digits | @ | 8 |
16 | distributed | digits | @ | 1 |
I assume that scanning on distributed shapes (example: rows 1,3 then 2,4) is thermally better than adjacent shapes (example: rows 1,2 then 3,4), but maybe if everything is cycling at such a high rate, it doesn't matter.