As the Serbian language is quite complicated, with "heavy" and long sentences, it needs larger X×Y alpha matrix. But, there are two unique details - all letters are used (there are no dummy letters), and some words are vertically aligned - like in the crosswords.
GPS receiver L80 is used, so there is no internal timebase, no battery and no adjustment keys. MCU is PIC18F2525. There is the phototransistor which is used to dim LEDs automatically.
Details
This is the letter arrangement - it was a huge job to solve it, some words (blue characters on the drawing) are aligned vertically, although all letters are used horizontally.
Yes, a lot of "naest"s :) (it's something like "teen", but for all numbers from 11 to 19). But it was necessary to repeat them all if you want to get perfectly clean and grammatically correct output, with 1 minute resolution.
looks like you could have saved some naest here and there, as long as you could live with splitting of a word :)