At 03:10 PM 1/24/2009, Tony Vandiver wrote:

> > That's a waste of money unless you're absolutely sure that's the 
> final product
> > you want and are not in the prototyping stage. If you're still in the
> > prototyping stage like I am, it's not worth it to get a PCB made. What
> > if there are errors in my design? There goes $50
>That's not completely true.  For your $50, you get a lesson - a very
>good one usually.  Do you think your first try is going to work for
>boards that you etch yourself - if not, consider all the time and effort
>involved in redesigning/rebuilding/reetching/redrilling/resoldering?  Do
>you think that the design that you have etched (and designed accordingly
>to make etching feasible) will be compatible with real manufacturing
>processes?  I buy boards all the time that I think are "close enough"
>and don't worry about what's not right.  The best way to get a board
>right is to kludge an almost right prototype to make it work, esp
>compared to building one from scratch.  Lot's of people disagree with my
>approach, but I contend that it's fastest and let them work on their
>designs/layouts for days while I'm working with my real prototype and
>sending it to the customer for evaluation with a few jumpers on it.
>Let's say you have to cut and patch two or three traces - compare how
>this board with a few jumper wires goes together with how a homebuild
>goes together.

Well put Tony. Often assembling a home-build board is a real chore. 
For even the simplest boards you should check the continuity between 
pads because it is so easy to get hairline voids.

The $50 is in reality an hours work at even moderate salary rates. It 
is very easy to lose an hour on even the simplest board. Generating 
the design should be as much about thinking and checking as it is 
about doing. But then it can also be about how you work. I tend to 
use a mixture of demo boards, little SMD mounting boards and white 
prototyping block. Pretty much anything can be mocked up that way and 
the design concept proven.

The wire jumpers are essential. It is not unknown for a board to get 
to the production phase with an error in it. But it should never be 
the end of life. Even if you are building the board 'homebuild' 
style, thinking carefully about the design can save a lot of wasted 
time. Here, and in many other places where hobbyists gather, I find 
people routinely not reading data sheets. Manufacturers produce data 
sheets and application notes for good reason - they have the 
information you need to use the parts. The data sheet should be 
carefully read BEFORE you do anything else!

John


>How much money you have or don't have is irrelevant.  The question is
>what's the best value, and if I could save $25 for my business by doing
>something different like building boards in house, I'd do it and
>likewise most companies would do it, but it just doesn't make economic
>sense because time /is/ worth money.  Yes, this means that you're
>probably going to have to buy barebones prototypes twice (or in some
>cases 10 times), but when you're done, you have a repeatable design
>that's ready for manufacturing.  That being said, manufacturing is
>typically my goal, but it may not be yours, and if that's the case, your
>priorities can/may be different.
>
>My point is that a lot of us have been down the road you're going down
>before, and we're trying to tell you that it's not pretty.  We've
>learned our lesson, and are trying to teach it to you by passing along
>this information, but unfortunately, the only true way to learn
>something is to do it - so go for it, and later when someone asks for
>your advice and you know the answer through years of experience, try not
>to quiver when they tell you that you don't know what you're talking
>about - they too will need to live through it.
>
>Thanks,
>
>Tony
>
>--
>http://www.piclist.com PIC/SX FAQ & list archive
>View/change your membership options at
>http://mailman.mit.edu/mailman/listinfo/piclist
>
>
>--
>No virus found in this incoming message.
>Checked by AVG.
>Version: 7.5.552 / Virus Database: 270.10.13/1912 - Release Date: 
>1/23/2009 6:54 PM

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