From: "Frank Bemelman"
References: <firstname.lastname@example.org> <6qjF9.160687$QZ.26861@sccrnsc02> <3DE670F5.6BF51A27@gravitywell.org> <email@example.com> <firstname.lastname@example.org>
Subject: Re: ? Binary to BCD conversion chip ?
Date: Fri, 29 Nov 2002 10:40:33 +0100
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
Organization: EuroNet Internet
NNTP-Posting-Date: 29 Nov 2002 09:40:46 GMT
"Tony Williams" schreef in bericht
> In article <email@example.com>,
> Frank Bemelman wrote:
> > Hmm... you could try a binary counter, incremented by whatever clock is
> > available, use the same clock to increment a dual bcd-counter, compare
> > the *binary* result with your input, and when both are equal, inhibit
> > the clock to both counters. Add flipflops and latches to make it
> > all foolproof ;)
> That's similar to JF's more standard suggestion.
> Load the BIN counter with the binary unknown,
> zero the BCD counter, count down the BIN, count
> up the BCD, when BIN=0, BCD= right answer. Only
> about 5 or 6 chips. :-)
Yes, it is terrible. At the moment I am designing a
board that is going to be terrible as well. Nothing 'fits'
and I need the craziest stuff to 'connect' everything
together, lots of half-used ic's etc, really beyond
the point of being funny ;)
Now, all of a sudden, I read my own text above, and notice
that I used the word 'inhibit'. When choosing labels in
my schematics, I always use 'disable'. Seems like a
rather old-fashioned word, 'inhibit'. Never mind.
> The cost of doing BIN-BCD explains why early uPs
> often had a BCD arithmetic facility, or the odd
> little hooks for it.
8051 has a few BCD instructions I believe, but never
knew what to do with it ;)
(remove 'x' & .invalid when sending email)