[hpsdr] Hermes Build Experience

[Rob Frohne]

Greetings,

I thought I would provide you with a short update on my experience 
building Hermes.  There are some notes and links to photos on the Wiki 
<http://openhpsdr.org/wiki/index.php?title=Hermes_Board_Build_Notes> 
too.  I wanted to test our stencil technology out with Hermes, because 
it is a little more challenging than most of what we have done in the 
past here.  The basic method is seen in this video 
<http://www.youtube.com/watch?v=0FHvikd3BkY&feature=youtu.be>.  We make 
the stencil on an LPKF engraving machine.  The solder stencil is used to 
lay on the solder paste on the board.  The components are then placed 
with tweezers or what have you on the pasted board, and then it is baked 
in the reflow oven or toaster if you don't have a regular reflow oven.

We had to modify the gerbers that Abhi provided because they were just 
for solder mask, and the holes need to be smaller than that. We then 
engraved one for the back and one for the front.  The back was done 
first, and was quick, about an hour to place the parts and bake it.  It 
worked very well.  The front was more challenging.  It took about 
fifteen hours to place the parts.  It is like putting together a puzzle, 
except you don't try pieces for fit, you find them and then carefully 
place them.  Terry Long, N8AB, provided some help with finding the more 
common parts through his documents, but the longest part of the process 
is still like playing "Where's Waldo?" 
<http://whereswaldo.com/index.html#findwaldo/map1>  I wish we had the 
original design files in Kicad or some open source format, because then 
you could quickly use the computer to search for and locate the parts.  
As it is, the gerbers are vector graphics files (including the text) so 
there wasn't an easy way to find the parts. Terry's files 
<http://people.wallawalla.edu/%7ERob.Frohne/Hermes/1010_HERMES_PROD_BOM_N8AB.xls> 
were searchable, because he labelled the parts.  There were some errors, 
which I hope Terry will fix for us.  It took me about a week (where the 
solder paste was just sitting drying on the board) to fit in the fifteen 
hours to place the parts.  I was a little worried that the paste would 
harden, but that didn't turn out to be a problem.  I probably wouldn't 
leave it sitting around much more than a week though.  There is always a 
risk something might bump it and the parts go everywhere.  I had mishaps 
placing the FPGA and the LTC2208, and both of them had to be slid over, 
smearing the solder paste a bit.  There were about 60 solder bridges, 
mostly on the FPGA, but none evident on the LTC2208.  It took about an 
hour and half with very tiny solder wick (about 1/16") and a good Metcal 
iron to remove the bridges.  Having the right wick and a good iron is 
important.  You need to use a new piece of wick on each bridge.

I am still waiting for some Mouser parts that I neglected to order until 
this week, so I haven't powered it on yet.

The cost of the board and parts was about $450 without the LTC2208 or 
the Altera FPGA which I had on hand already.

I am planning on making several more stencils, and passing on the ones I 
used, so that others can build their own the same way I did. If you are 
ready for one (and will pass it on within a couple of weeks to someone 
else on the list), send me a personal email.  I'll send my stencil set 
on, and over the next couple of days, we will make a few more sets.  I 
will send John Melton one or two to send around Europe as well.

73,

Rob
KL7NA

-- 
Rob Frohne, PhD, PE
Professor
EF Cross School of Engineering
Walla Walla University
100 SW 4th Street
College Place, WA 99324
(509) 527-2075

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20121218/1a7dd317/attachment-0003.htm>