Those 64 pin PLCC zif sockets are over $100. You need a minimum
of two. One on  the
programmer
and one on the proto board. It takes time to burn an EPROM. Then
you have to pull it
from the programmer and insert it in the proto board. You need at
least two PIC EPROM
chips.
One being erased and one being tested. It takes time to erase
them so one would
probably need three or four EPROM PIC's at around $20 each. I did
it this way
for years and years on 6502 systems until I wised up to using
static RAM in
place of EPROMS. I program by trial and error. Mostly error :-)

I still have not bought a PIC 17C756 EPROM programmer. Tracy has
one already
so I will let him burn the code after I get done developing it in
the Static RAM
based development system.

Perry Casson wrote:

My thinking is if you designed the hardware to allow for an optional EPROM.
Then once the start-up code is burned into the PIC that detects if the EPROM
is installed and if it is it executes that code and if it is not then it
continues to execute the code burned in the PIC.  If you did it this way you
once have the basic init code working and burned into the PIC you only have
to work with cheap proms for testing.

The advantages would be:

- no need for a ZIF PIC socket 'cause it would only need to be updated a few
times
- EPROM's and ZIF EPROM sockets and EPROM burners are real cheap
- You might need the EPROM address space someday to store more code or
lookup tables
- Your ram based system would need to reload the firmware from a PC after a
reset.
- 1 PCB layout for both a development system and the production PCB means
more people could be developers. I'm not sure on Tracey's plans with this
but if the firmware source is open to who ever buys the hardware those who
wants to add functions that are not planned for the base unit can do it.
(i.e. add oil temp and water pressure)
- After you have a full working revision burned into a standard burn once
(OTP)version of the PIC you can continue testing experimental firmware in
the prom.  Pull the prom out and your back to the standard firmware.

I see the have a flash rom version of the 16CXXX family I wonder if a flash
17Cxxx is far away?

enough of my ramblings,

Perry Casson

Good points Perry. I still make a lot of mistakes. I must have burned well over
a thousand EPROMs in my time. It is much easier and faster to push a key or two
and download the new software version to the ram in seconds.

No intention of using the RAM version in the flyable system. That would be
a different printed circuit board. Actually I will fix it so I can cut 
the development system part of the pc board off.

Yes the idea of a flash rom version of the PIC 17Cxxx family gives me pause. I am
sure
it is in the works. So this board is going to have short life. 
 
Paul Lamar
 
The Aircraft Rotary Engine Newsletter.              Powered by Linux.

http://home.earthlink.net/~rotaryeng/            http://www.linux.org