anyone out there

[Andrew K. Mattei]

Bruce Plecan wrote:
> 
> I'm hoping it's due to the really cleaver guys working onna emulator design,
> that the guys here at CSH, HQ can build.

Mr. Chatterbox (me) has been busy lately... Trip, parents in town,
etc...

However, I have spent some time trying to learn about the CPLD as
discussed in Ludis' email. Unfortunately, I'm having a tough time
locating a "basic theory of use and operation" guide on PLD's. Back when
I was in college taking digital design (a mere 8-9 years ago), they
never even mentioned PAL devices! We did everything with discrete TTL
chips/gates... And we liked it! Well, sort of. I've spent time on the
Xilinx site, the Lattice site, and some others, and am coming up dry.
Unfortunately, it's not part of my "professional job duties" to work on
this stuff, so I can't effectively do it at work. ;)

The information I've gathered thus far has given me just enough to get
my imagination running. However, my mental stumbling block kicks in when
I start thinking about thrashing bits around and filling / retrieving
addresses. I think about sending the commands (like xFFF, xFFD, xFFC)
and then sending data to fill addresses. That's where my black box
fails... :( To create a logical expression based on several inputs, and
then have those inputs perform different functions later on... OK,
really basic question. I "assume" some kind of clock is going to be
connected to the CPLD - and that based on sequential commands (like
those above), and a certain logical gate layout, we can vary outputs.
:-/ How, I don't exactly understand. Of course, *it's not the point of
this list to teach CPLD's*, is it? ;^D

Of course, I have delinquent thoughts on a (shrug...) low tech parallel
port based dumb (no uC) semi-emulator. I have all the parts for it in my
junk bin, and it would just be a matter of building it. I say
semi-emulator because it's not a true "real time" data changer, it's
almost just a pair of flash chips in parallel with buffers, selectable
Bank "A" or Bank "B", that you send a binary image (or pieces thereof)
to. Pretty simple/basic/old stuff. Premise of operation is that you
could have two different images loaded, switch between the two (manually
or electronically), have one bank as "experimental", etc... Load time
wouldn't be too terrible (minute or so?), but advantage would be that
you wouldn't have to ever go back to an EPROM programmer again. Just
have a DB-25 cable hanging out of the ECM to plug in to the parallel
port. I do believe that this deviates from what the original intent of
the "EPROM Emulator" was for, though. I believe that the "original
intent" was to have a near-real time changeable data bank...

I know, I just need to get off my arse and actually *do* something on
this, right? ;^D You see, it's either work on the car downstairs (the
one I've removed every bolt from), or sit upstairs and solder... Both
take lots of time. And I'd really like to get that d*mn car rolling
again... ;)

Cheers!

-Andrew