woodelf wrote:
Well I do most of my hobby development work at 3am
and don't want to turn on the PC. :)
Since I plan to do 512x8 proms stand alone is the better
idea for me.
Looking at the 2816 data sheets you have listed 10^4 or 10^5
cycles -- I take that to be write cycles and any number
of read cycles is that correct the way I read it?
If you are expecting to turn the crank on the microcode
development more than a few times, consider designing
the interface to the microstore in such a way that you
can piggyback a small SRAM in place of the PROM. Use
latches with tristate outputs to drive the address to
the microstore (assuming you aren't pressed for speed
in the final design).
Then, when you want to twiddle the contents of the SRAM,
tristate the latch outputs and enable address buffers
on your "piggyback SRAM board" so that you can force the
address to whatever suits you. Drive the data you want
into the RAM and generate a write cycle.
I.e. make a PROM emulator.
When done "programming", disable the address buffers on
your piggyback board and reenable the tristate latches
*in* the processor.
<shrug> If you are the type that likes to do iterative
designs (instead of getting it nailed down on paper
ahead of time), this can be a real time saver. And,
bipolar PROMs get expensive if you're burning a new
set every hour...