> There are plenty of projects where 6 and 8 pins are fine

Yes, of course, I was just making the point that many micros are
made with several alternate functions for each pin, which obviously
keeps the pin count down. For the reasons you cited, sometimes
we have to do the same (by adding our own h/w if necessary, eg
the old RC method to make an ADC out of a digital I/O)

> So even if a bigger PIC costs a little more than external hardware,
> it will get the job done reliably with a lot less futzing around trying
> to figure out why it doesn't work

The total beginner (or anyone just unfamiliar with a particular chip)
would be best advised to sample. I still have 4 DIP 4550 samples
here that are used for prototyping. Some projects originally proto-
typed with those chips have moved or are moving into SMT for
production. A lot of 4 x 4550 is relatively expensive to buy from
somewhere like RS and I'm very grateful that Microchip's sampling
service is so generous

I agree that if you have the luxury of excess pins or excess memory,
there's little point, unless there is a particular end goal (whether it
be a learning exercise or a need to consolidate pin usage so as to
downsize), of cramping yourself with a too-small chip

I'm working on something now that's based on an idea I saw in
a magazine. In that article an F88 is being put through the wringer.
Some clever pin multiplexing is being done, but given the amount
of I/O needed it would have been better done with something like
an 876 in the first place. I suspect that an F88 was chosen because
it's one of the few PICs you can buy at retail level, in this part of
the world anyway

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist