🚀 go-pugleaf

Rich wrote:
>
> I am working on a small project involving a large number of digital
> ins and outs.  I have been looking at using a couple of Intel 8255s,
> but was wondering if there are any better options out there.
>
> Thanks for any info.
>
> Richard
>
> --
> ri...@earthlink.net

Basically no.  Though these parts are old, they're really pretty good.
There are modern, faster, CMOS, etc versions of these parts, but they're
the same 8255 chip.  Being able to program input, outputs and irq is
pretty nice.  The downsides to these parts is that they're big, somewhat
expensive in low volumes and can't sink/source much on their outputs.

Some other options might be:
1. If the I/O's are spread out, just use 74xx574's for outs and
74xx540's for inputs.  These are octal jelly bean parts that are very
cheap.  There are even transciever versions if you need to get fancy.

2. Several companies make big version of these bus parts in widths to
20+ bits which can make things simpler.  Check out Pericom and National.

3. You could use a medium sized FPGA to get lots of I/O.  It would
certainly not be as cheap as an 8255 but you could put some custom stuff
in it like address decoding, power up defaults, etc.  Something like an
Altera 7064 is pretty cheap and has a fair amount of I/O.

4. You might consider using a hybrid approach with a small FPGA routing
signals to 74XX574's etc.


Good Luck,
John

ri...@earthlink.net (Rich) wrote:

>I am working on a small project involving a large number of digital
>ins and outs.  I have been looking at using a couple of Intel 8255s,
>but was wondering if there are any better options out there.

Richard since you were unspecific about the amount and spped of the IO
generalisation is hard to come by.  The 8255 or the z80 PIO are hard
to beat.  For some tasks I've used the 8741/2 series slave cpus as
they were fast enough and could do some processing for my tasks.
These parts offer a minimum of 18 lines of IO and can use an expander
device (8243) to get more in chunks of 16 pins.  The down side is they
must be programmed as they are MCUs and not as cheap as the 8255.

Allison

Real address is: Allisonp @ world DOT std DOT com
  This was done to discourage some of the junkmailers.

Rich wrote:
>
> I am working on a small project involving a large number of digital
> ins and outs.  I have been looking at using a couple of Intel 8255s,
> but was wondering if there are any better options out there.
>
> Thanks for any info.
>
> Richard
>
> --
> ri...@earthlink.net

Hi Richard,

You may look at some EPLD/FPGA...  We use some Xilinx chips to
interface CPU with digital I/O ports.

On the other hand, if you have board space, why not resort to the
good old logic latches ( HC types are cheap and work nice ).


--
         _________   _________   _________   ___         _________
Michel  /  ______/\ /  ___   /\ /  ______/\ /  /\       /  ______/\
       /  /______\//  /\_/  / //  /______\//  / /      /  /\_____\/
      /  ____/\   /  ___   / //  //_   /\ /  / /      /  ____/\
     /  /____\/  /  /\ /  / //  /__/  / //  /_/___   /  /\___\/
    /________/\ /__/ //__/ //________/ //________/\ /________/\
    \________\/ \__\/ \__\/ \________\/ \________\/ \________\/
Mont-Briant

In article <VA.000000...@port1-annex1.webrider.net.au>, Jim
Francis <j...@webrider.net.au> writes
>The 8255 is still an excellent PIA in many respects. However there
>is one very important point which is not obvious from the data
>sheet and can rule out use of the part for certain applications.
>
>I can't remember the exact details but basically you cannot turn
>an I/O pin from input to output (or vice versa) on the fly
>without resetting the whole port and causing all the pins to
>glitch. I had this problem while interfacing the 8255 to an LCD,
>trying to turn a single bit around to read the LCD "BUSY" completely
>stuffed things up. There was no way around this problem for my
>application and I had to use a different part.
>
>If you think this may be a problem for you, I advise a _VERY
>CAREFUL_ reading of the data sheet.
>
>Regards
>Jim

This is the biggest limitation of this device, the pins are not bit
programmble for I/O direction. Ports A and B can each be programmed to
be an input or an output port - all 8 bits. Port C is nibble
programmable, bits 0-3 can be all inputs or all outputs as can bits 4-7.

However it is possible to use the 8255 to drive an LCD display - we do.
Anyone interrested should go to the literature section of our website
and download the dev552.pdf file, the technical manual for our 80C552
based module. It contains a brief desciption of 8255 programming in its
most common mode and also 8051 assembler code for driving a *standard*
1-4 line alphanumeric LCD.
--
Kindest Regards, Gerry             |  We manufacture processor modules
ge...@devantech.demon.co.uk        |  using 80C31, 80C188EB 68302 CPU's
http://www.devantech.demon.co.uk   |  also assemblers & C compilers.

The 8255 is still an excellent PIA in many respects. However there
is one very important point which is not obvious from the data
sheet and can rule out use of the part for certain applications.

I can't remember the exact details but basically you cannot turn
an I/O pin from input to output (or vice versa) on the fly
without resetting the whole port and causing all the pins to
glitch. I had this problem while interfacing the 8255 to an LCD,
trying to turn a single bit around to read the LCD "BUSY" completely
stuffed things up. There was no way around this problem for my
application and I had to use a different part.

If you think this may be a problem for you, I advise a _VERY
CAREFUL_ reading of the data sheet.

Regards
Jim

       If it wasn't for blinds it'd be curtains for all of us
     Jim Francis - Melbourne - Australia - <j...@webrider.net.au>