Greetings all!  I was an electronics hobbyist back in the late '70s
and early '80s but have been pretty much a software guy since then.  A
few weeks back I stumbled onto the SX and was shocked that nowadays a
50 MIPS microcontroller can be had for five bucks.  I immediately went
out and popped for the SX-tech tool kit.  It should be arriving any
day now.

In the meantime, I've been researching other microcontrollers, mostly
the Microchip PIC.  I must say it has been quite an effort trying to
figure out the product line, which chips are history, what to use
where and so forth, but I'm beginning to see the method to the
madness.

I've found an incredible amount of information about the PICs on the
web, perhaps almost too much.  The number of different PIC programmers
alone is astounding.  And this is where I have my newbie question.
I'm sure this is a FAQ, and if someone can point me to a single
reference I'd be happy to go off and start reading.  (Also, are
archives of this list prior to the recent switch to majordomo
available, and if so, where?)

So, here's what I know (or think I know) and what I don't...

It seems as if the only features the fancy (i.e. expensive)
programmers have are the ability to program everything in a single
socket and then test the programming at multiple voltage levels, but
it also seems as if these features aren't really necessary for the
hobbyist.  It also seems as if they can program the older non-serial
units, but I doubt I'd be dealing with any older chips at this point.
The speed at which the chips are programmed seems to be more a
function of the software than the programmer itself.  Is there more
that I'm missing here?

It doesn't really seem as if the very simple programmers based on
AN589 or COM84 really do anything different that the fancy-pants
programmers, so long as you get the programming signals on the proper
pins.  The one difference I do see is that signal timing is handled by
the PC, which on a slow machine (or perhaps even a not so slow
machine) is non-deterministic in a multi-tasking environment.

It seems as if the electrical characteristics of the AN589 circuit are
standard for all recent PICs, and it is the software dataflow that
really determines what chips can be programmed.

Then there is the issue of in-circuit programming.  It seems as if
in-circuit programming uses the same programming functions and signals
as the "standard" programming, and doing in-circuit programming is
more a matter of designing the embedded circuit with proper isolation
to accept in-circuit programming.

Now, because I'm new to this, I'm going to be doing a lot of
re-programming on a breadboard.  (I even found my old breadboard and
stash of old components!)  It seems silly to constantly be moving the
PIC back and forth between the breadboard and a programmer -- at the
very least that would involve the purchase of two ZIF sockets.  It
would seem to make much more sense to wire the five programming pins
from the breadboard to a socket which can then be plugged into either
the programmer or a loopback to the breadboard.  In reality, since I
have just one rather large board, the programmer would likely live on
the same breadboard to start out, epsecially if it was a simple one.
The advantage of this "jumper" approach seems as if would simply
eliminate the need for all of the isolation circuitry.

This *seems* as if it would be a very simple and elegant solution to
development and prototyping, yet I have found only a few references to
this method. (Wisp and TLVP) Furthermore, it seems as if this method
will easily allow the use of the LVP, eliminating the need for the 13v
supply & associated circuitry.  Are there issues involved that I am
missing?  Are there cases where this doesn't work?

Right now, I'm leaning towards using Byron Jeff's design for the
Trivial Parallel Port programmers (both the low voltage and high
voltage varieties) due to the low component count.  If this turns out
to be a good solution, I'll probably use pp06 for programming because
of it's wide device support.  (That does mean I'll be writing my own
parallel port driver, but given the modular nature of the program this
seems trivial.)  Again, this assumes the electrical characteristics
for the different devices is the same, and it's the format of the data
stream to the device that determines what devices can be programmed.

I do have a specific project in mind which will require a fairly large
number of simple PICs, and it looks like the 12F629/16F630 will fit
that bill at $1.60/$1.80ea.  (This is the specific reason for my
interest in PIC instead of SX!)  Although these devices are not billed
as LVP devices, it seems as if they do not need the 13V of the more
typical HVP devices.  They seem to be HVP flash memory, but include an
internal high voltage generator.  It seems as if anything from 8.5V to
13.5V will do the trick.  If I'm understanding this correctly, it
eliminates the need for that pesky 13V source!

Many thanks in advance for any comments, advice and suggestions in
these areas.  Confirmation of where I'm right would be appreciated as
well as note of where I'm in error!

-p.
_______________________________________________
http://www.piclist.com
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist