14 Aug
2010
14 Aug
'10
4:20 p.m.
On 14/08/10 22:12, Alexandre Souza - Listas wrote:
... which brings us neatly back to "I need a GAL programmer that isn't
dependent on the host system running a single-tasking OS".
Relying on any PC to generate accurate timings is stupid unless you turn
off interrupts for a bit. Hope you're not expecting the clock to be
accurate when you've finished... and even then the timing might not be
that good (the DMA controller might decide to steal a few bus cycles
here and there).
If I was going to do it, I'd bolt a J-series PIC18F chip onto an FPGA,
then wire the FPGA up to a couple of gate driver boards. Hmm, sounds
suspiciously like the DiscFerret hardware, doesn't it? :)
--
Phil.
classiccmp at philpem.me.uk
http://www.philpem.me.uk/
What else do you need? The commands outlined in
openschemes seems to be
enough for creating another program, maybe I'm wrong :)
The point I was trying to make is that you said the TOP2008 had been
reverse engineered similarly; that page only has info on the TOP2005.
And quite frankly I don't need another badly-engineered Chinese brick on
my desk.
Reverse
engineering wouldn't be required. Tidying up MaWin's schematic
would be a good starting point, but the code is reasonably easy to
figure out. Porting the basic algorithms to Linux wouldn't be hard,
but getting the timing right would be.
That's why these timings must be done on board