On 31/01/2010 19:49, Geoffrey Reed wrote:
On 1/30/10 8:09 PM, "CSquared"
<csquared3 at tx.rr.com> wrote:
> Ah yes, 6850 and ACIA, thanks for the memory jog. I had completely
> forgotten both of those. I can't remember now precisely why I felt the
> need to reset the ACIA with a software controlled line, but ISTR I
> *really* needed to occasionally, and a power cycle was the only way. I
> do remember the little system had several of the ACIA's attached to one
> 6809 and I controlled a number of unique hardware devices that way.
Hmmm, possibly It is because I am more of a
hardware guy anymore than a
software guy, but couldn't you make an adapter pcb for the ACIA chip,
that
had a relay and a few other componants on it, like a 555 set as a
timer, and
when the 555 is triggered, it would power off your acia for a preset
amount
of time by tripping the relay so the VCC pin no longer gets voltage,
then at
the end of the timing cycle it returns the relay to normal position
and the
acia comes back on and you set baud, etc....
I am really puzzled by this. The ACIA has no hardware reset line, but
it's very easy to send it a master reset command in software. In fact,
you have to do that to clear the hardware reset condition after a power
up, because the ACIA stays in the inactive/reset condition until you
program it.
Charlie, you weren't using GTE 68C50s, were you? They did have a few
weird bugs.
I'm pretty sure they were Motorola parts. As I told Geoffrey, it has
been a very long time - like probably 40 years - and I've forgotten a
lot of the details, and just sort of remember the pain. I think the
ACIA would somehow get into a state where it would not pay attention to
the master reset command.
Over the years, because of this and other similar experiences, I sort of
developed a philosophy that every peripheral of an embedded controller
should have a software controlled reset line that could unconditionally
reset the device independent of the state of the device. Without this
capability, I don't believe it is possible to create a truly robust
system. By robust I mean a system that can continue to run for months
with no intervention needed to clear an error condition. Another
example of the sort of experience that led to this philosophy are the
modems I mentioned to Geoffrey. A reset command to the modem has no
effect if the firmware in the modem is off in the weeds somewhere.
Later,
Charlie C.