I've been promising myself a GAL programmer for quite a while now.
Unfortunately, until now, I haven't had the money to buy one.
When I found
MaWin's homepage (MaWin = Manfred Winterhoff, email@example.com) and read about the GALBlast, I instantly decided to
build one. That was it - there was no changing my mind.
After downloading a modified PCB for the GALBlast programmer from
I pulled out the Press-N-Peel and my PCB etching kit and etched the board.
After soldering in the parts, repairing a few solder bridges and calibrating
the programmer, I plugged in a GAL. It went up in a puff of smoke. Further
testing revealed that during earlier testing, I'd managed to short circuit
the output from the switch-mode power supply to one of the pins on the
74HC573 latch. Great. The latch was now acting as if its Latch Enable pin
was stuck active. The latch refused to keep the data outputs stable - instead
it acted more like a data buffer. After replacing the 'LS573 and repairing
the original fault (a small whisker of solder), I tested the programmer again.
It worked - and this time the voltage remained stable at 12V.
Well, that's it - the programmer works and it's been programming GALs
flawlessly since it was repaired. This time, I hope it stays like that.
I've been using Atmel's version of WinCUPL to generate the programming
data files for the GALBlast. To be honest, WinCUPL isn't too bad after
you've learned how to use it. The GALBlast software is pretty reliable,
too. Now all I need is some time to learn the more advanced features of
WinCUPL. But for now, all I need to know is how to get a GAL to work as an
address decoder, anyway.