[geeks] Electronics

[Dave Kimmel]

On Aug 30, 2005, at 10:30 PM, Dave Kimmel wrote:

> Assuming that I don't give up on this I'll keep everyone updated as I
> finish more pieces of it.

Well, I haven't given up yet, so here's the latest bit of progress.

I figured out how to handle row selection without using a ton of  
output pins on the controller.  I've also hacked up a little program  
on the controller to run the row selection logic so that I could do  
some testing.

http://djk.smugmug.com/photos/34368392-M.jpg

On the left there is an STK500, hooked up to the breadboard so that I  
don't have to move the processor around every time I want to  
reprogram it.  On the breadboard, starting at the bottom there is an  
ATMega16, above that the row selection logic, and above that some  
ULN2003 high-current drivers.

The row selection logic deserves a schematic diagram, but I usually  
do those after I have a working prototype, so I don't have one yet.   
In a nutshell, I used two AND gates (the chip on the right of the  
middle row) to tie three 4017 divide-by-10 counters (the remaining  
three chips on the middle row) together and form a divide-by-21  
counter with decoded outputs.  The CPU drives this with two output  
pins, clock and reset.  The outputs of the counter are hooked to the  
ULN2003 drivers, which are then hooked to the rows of the display.   
Five columns of the display are hooked directly to the positive  
supply, through 330 ohm resistors.  It's not impressive, but it  
proves that I'm on the right track.

Bill, if you decide to get serious about programming AVRs, I highly  
recommend checking out the WinAVR package.  It's the GNU toolchain,  
plus a bunch of utilities, all in an easy to install package.  As the  
name implies, it's for Windows, but all of the important tools can be  
built for OS X quite easily.  The Procyon AVRLib is neat too, it  
gives you a bunch of useful routines for use in your AVR programs.

-- Dave Kimmel
    crisco_kid at shaw.ca