| I suppose this doesn’t *need* the ATMega, but they’re only a few bucks, so what the hey. This uses an old GI SP0256-AL2 speech synthesizer IC containing all 59 English phenomes. It sounds very computer-like (even though the original documentation boasts “natural sounding speech”. The DIP switches allow the user to specify a unique device ID for the network. Because it makes noise, I added brown-out detection — something I probably should add to all the devices eventually. The PWM speech output is filtered though a 5KHz low-pass, then sent through a volume control pot, then on to a low-power amplifier. The same 4-pin input is at the upper left, and the 2-pin speaker connection is at the lower right (above the big cap). Bonus: A programmable LED nestled between the ICs! (Oooo!) |
