On Fri, Oct 15, 2004 at 11:06:43AM -0600, tim_webb@agilent.com wrote:

> I strongly encourage you to use the MAX 232... or even simpler the MAX 233
> less components required and you can also request free samples from the
> manufacturer.  I have tested both parts in designs and they both have
> performed flawlessly for all of my projects.

Prices on the MAX233 are awful! $5 USD? OUCH!

But I could be convinced on the MAX 232. The TI version at digikey is 78 cents
each and 65 cents a pop for 25 of them. That's cheap enough to make it
usable.

> 

> I also suggest that you ditch the 16f84 for a Pic with a built in serial
> USART, it will make your life so much easier.

Definitely agreement on that.

>  You can even request free samples from Microchip too.  

Great for small project quantities.

>I believe learning is not about using old technology and bit banging, it is
> about learning how to utilize the newest technology at you disposal, it's
> quicker, it's easier, and its a lot more forward compatible as your project
> evolves.  Don't waste your time with old technology, move forward and use
> components that are not or will not be discontinued soon.

Some agreement and some disagreement here. I agree that new tech is usually
a good thing. However there's nothing wrong with old tech, except that it
generally makes life harder. I believe that you need to learn both. But the
order is important. For example right here in this thread. If you want to do
serial, you should learn how to use the USART first. Learn how to do it in
hardware. Why? Because the vast majority of the time you'll only need one
port and the hardware supports it. However bit banging can be useful when you
need a second slow port for example. So don't dismiss it. Just don't emphesize 
it as the primary option.

> In an R&D environment where I work, we are actively
> looking for the newest technology to push the envelope, using Obsolete
> components may leave you in a position where you will have to do a complete
> redesign then the obsolete parts are no longer available this is time
> consuming and not forward thinking.

Disagree here especially where Microchip is concerned. You can still get those
parts in quantity that are more than 10 years old. They are not going anywhere.

But even MChip discourages their use via pricing. A 16F84A costs 3 times as 
much as a 16F628A. And the latter part is jam packed with goodies. It's
even more obvious with the 16F88 which has everything known to PIC stuffed
in a 18 pin package (USART, comparators, A/D, multiple timers, sync serial,
PWM, nanowatt tech with precision internal osciallator) while the 16F84A has
basically nothing but a single timer and an interrupt. Yet the latter part
is nearly double the cost.

Discouragement at its very best.

>  If you are anything like me, I'd rather do the right thing the first time,
rather than have to redesigning a failed design later, I have better thing!

>  s to do, like search for newer technology for my next project, rather then
> do something over that should have been done right the first time.  Always do
> you research before you start a project, it pays off in the long run, you
> become a leader in the newest technologies available, before you know it
> others will come to you for advice, and you become a successful circuit
> designer.

Well here's where things fall down. The reason that the 16F84 is still so
popular is precisely because of this research. You hit google and type
PIC project or PIC programmer, and you're still going to get 10000 hits 
for 16F84 based projects without any reference to the newer stuff.

> And last but not least, this Pic list is one of the most resourceful places
> to get quick suggestions, pro and cons, and good advice.

Definite cosign on that one.

BAJ
_______________________________________________
http://www.piclist.com
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist