Homebrew PCBs

Brian Schmalz wrote:

>Carl,
>	Thanks for the interest - it's a very small USB PIC based
>development board - sort of like a Basic Stamp in that it sits down in a
>breadboard. Example code supplied so that it shows up as a virtual
>serial port on your PC, Mac or Linux. 
>
>	If you want a little bit more information, check out the pages
>I've put up at http://www.greta.dhs.org/UBW - it's not much, but it's a
>start.
>
>	I would love nothing more than for Olimex or SparkFun to
>actually build and stock these boards. But I have no idea how many they
>would really sell.
>
>*Brian
>  
>

  Mainly sending to the list since a few others might like the PIC 
pattern, just skip the rest of the message after that if you don't have 
interest in how to sell prototype boards without much hassle..  
Considered not sending since enabling isn't in my best interest, but 
this type of pattern is common enough anyway so it's not that big of a 
deal..

  For Brian, if you want to do this, I am placing a board order around 
the end of this week or beginning of next, my board is simple and done 
just doing tweaks for size and fit..  They allow combined shipping, if 
you forward $80 (the $69.99 and ~$10 will add on my shipping..) by 
Paypal to me and the board pattern by the time mine goes out, it will 
economize things for you by another $20 for the 200 sq inches total.  In 
fact since your board is simple I will finish it off for you and add a 
232 pattern and maybe another desirable item or two that won't increase 
board space.  My stuff is making money, so you can have the ecomization, 
but I'll keep 3-4 boards for my use, and send your order on 2 day 
priority with insurance once it gets here.  Takes 3-4 weeks for them to 
get here though.  Should be able to make 40 boards easily enough, maybe 
50+ with a little squeezing, and they part them and come in a nice 
stack..  I would fund it myself but every spare cent is going into items 
for an upcoming convention, $80 now for me is hundreds less worth to 
sell then, so it'd be a much later date before I'd do a set of these 
even to sell.  Most of the details are below..

Alan



Original below..

  Sparkfun etc are nothing special, DIY.  It is about 1/10th of the work 
you seem to be guessing, if you have the foresight to set things up 
properly.

  First off, the boards are reasonable, tight design like I favor 
myself, which happens a bit more often now with the deals on per area 
boards.  But the idea of each package variation seperately is off a 
bit.  I threw a pic up here:

http://www.modellights.com/images/temp/Picsockets.jpg

  That takes care of 28 pin SOIC and DIP and 40 pin DIP all in one 
whack.  28 pin DIP PIC right side lines up very well with 40 pin DIP 
right side, puts portC(4-7) on portD(4-7) but that's of little 
consequence.  Needs a second pass at a higher resolution grid, moving 
the angles from the SMDs on a diagonal to give better clearances for the 
SMD pin fanouts.  Left side of the DIP 40 connects to everything, right 
side of the SMD and fan back out and you've connected to everything 
important.  Port C and D pins 4-7 out of the SMD are for pin functions, 
two traces wrap them around to the 40 pin RX and TX for serial functions 
coming from the 40 pin layout.  Simply cut those two traces if using a 
40 pin and everything's available.  Extra cost per board and size for 
people wanting to use the 28 pin chips is way offset by the huge 
reduction in boards required and inventory cost.  PLCC and TQFP can 
similarly be done together, but hardly worth it for a prototyping 
board.  With free PIC samples, people can get one of these packages for 
their testing.
  Pay attention to just how much this gets you.  For $100, covers things 
to 50 or so depth for all three packages regardless of which is most 
popular.  Vs. $140+ for a design with a mixed board of 3 seperate 
layouts, that will need a reorder once the most popular hits about 15 or 
20 orders.  Or $240+ for 3 seperate designs which will take a long time 
to pay off, and some of the least popular package may never even sell.  
Much less hassle and not much waste, even though a 1"sq plus waste is 
built in for each board for the 28 pin packages.


  If you want to do something, but it seems too hard, you don't decide 
to not do it, you rearrange it to make it easy.  While it isn't that 
hard to pull out an address etc and do things up for a single case, it 
becomes work when you start having to do it on an odds and ends basis 
with everyone putting their address in different places etc.  Make the 
person wanting the stuff do the work, since they are most familiar with 
their address etc and will only have to do it a single time.

$5 or $6 for board.
SASE enclosed with proper postage for a board, 2-3 stamps will likely do it.
Signed agreement page to cover any false expectations.

Long time frames so it can be a batch job for you and not too big of a 
hassle.  Have a code on your incoming address like ATTN: USB board etc 
and throw them in a shoebox and go through once every week or two.  
Simply open envelopes, pull out the money, agreement, and SASE, stuff in 
a board, close the SASE and stick it in the out going pile.

No tech support beyond the site.  Questions may be looked at eventually 
and enough of the same may be addressed on the site etc.
They must send email with filter subject after mailing payment to start 
the clock.
No secondary emails will be checked looking for missing boards etc until 
after 6 weeks older than the starting email.
If they don't have a board by 2 months etc assume the first payment was 
lost in the mail they can chase it with a second attempt.  If both end 
up arriving before one is sent out, first money will be sent back unless 
clearly marked they want a second board.  If both arrive but seperated 
and one is already out, simply send the second board.

If they've made two attempts and seen nothing after 3 months, then they 
must assume their payments are either lost in the mail, or you've sent 
the boards and they got lost getting to them.  Either way, they must 
assume the mail is simply too unreliable from them to you etc and they 
must write it off as a lost cause and make their own.  No refunds since 
it's likely you never got their money, or did send them a board and it 
didn't get there or could show up much later.  Maybe after a second 
email with no board but you have their agreements and got their money on 
at least one, an extra try sending a board out, but pad the initial 
prices enough to cover one now and then..

You don't want the boards free or too cheap, it will pull in a lot of 
people who will want it but likely never even have a real interest in 
using the boards, simply depleting your inventory for little reason.  
Still, make a few bucks so maybe just order more often and make their 
money too.  Maybe offer 2 for $8 or so and another stamp or two on the 
SASE as well.  Long time frame with agreement and above the cost of the 
board will weed out most freeloaders and cheats etc.  Anyone sending 
complaints or wanting a free board, send a copy of their signed 
agreement and tell them to have their lawyer clue them into their own 
lack of reading or thinking ability.  If you have nothing from them, 
they never sent anything or it didn't make it to you..


  It can be done, and it can be worked into a system where you simply 
open their envelope, take the money, and stuff a board in the inner 
envelope.  $5 a board will cover the costs of ordering from someone like:

http://search.ebay.com/_W0QQfgtpZ1QQfrppZ25QQsassZtigeryeshQQssPageNameZSTRKQ3aMEFSQ3aMESOI

200 sq inches should net a good amount of boards, after 20 you start 
making a little bit, simply enough to cover the hassle now and then, you 
won't get rich even though the per time is high.

  If you haven't guessed, I've looked at doing this myself, especially 
since so many others consider it a hassle and I know it can be done 
quite easily with the proper setup.  But many of my own designs I don't 
want out there.  And even this, I'd build it, seems like a great thing 
to put out there, then a week later I'd make it into a USB version of my 
IR bootloader system, and then damn this is too useful for the people 
who would end up competing against me, so I don't want to offer it or 
even just the USB board..  Best to see it ahead of time and not even 
think it lol..


  Of course the real utility of your site isn't the board or even the 
bit of modded software, but that you've pulled the links and got it into 
a working system on both ends so others don't have to hunt everything up 
and reinvent the wheel getting started.  Should speed up making my USB 
bootloader a bit, but will be June before I can get on with it anyway 
with too much happening right now.