Curious -- what exactly is "Antwort" that I occassionally
see in the subject line of these messages?  I thought I
was up to date on the piclist and internet jargon, but I
may have missed this one.

Cheers,
-Neil.




-----Original Message-----
From: pic microcontroller discussion list
[mailto:PICLIST@MITVMA.MIT.EDU]On Behalf Of Matt Pobursky
Sent: Thursday, June 20, 2002 5:11 PM
To: PICLIST@MITVMA.MIT.EDU
Subject: Re: Antwort: [PIC]: cheapest way for adding output pins ?


More counterarguments... ;)

On Thu, 20 Jun 2002 16:39:07 -0400, Bob Ammerman wrote:
>A comment was made on this thread that the shift register
>solution required
>many packages vs a CPLD or some such. Here are some
>counterarguments:
>
>1: The packages involved are quite small (SO16) and don't take
>up all that
>much space (how big is a 100 pin CPLD).

About twice the space of (2) SO16's for a 14mm 100pin TQFP.

>2: Since you only have to route a serial and two clock signals
>to each shift
>register, so you can distribute them on the board to be close to
>their point
>of use.

That's true. If you are already working with a large physical
board then many IC's close to their termination makes sense.

>3: The lower density may permit the user of a 2-layer instead of
>4-layer (or
>more) board.

I routinely use 100+ pin CPLD's on 2 layer boards -- although 4
layers does make power/grounding easier and better. One other
factor though is that with the CPLD you have a lot of lattitude
in assigning pinouts such that you can "octopus" out from the
pins directly to their destination by judicious pin assignment.
In a current design I have a I/O expander type CPLD design that
has 48 digital I/O lines and zero vias on their traces.

>Also, the 0.32 price each at 100 drops quite a bit in larger
>quantites. How
>about 0.1875 each in SO16 qty 10000 on tape and reel? That is a
>total of
>0.1875 * 13 = $2.44 parts cost per board.

Parts cost will always be lower with discrete logic. I learned
that long ago. But sometimes the other physical and performance
related characteristics make the CPLD route a good choice too.

>Sometimes simple is just what you need.

Agreed.

Matt Pobursky
Maximum Performance Systems

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu