[sdiy] My SMT projects

[Tony Clark]

   Since the discussion of surface mount technology cropped up, I thought 
I would share what I've been doing with it.  My idea was to make compact 
modules that had every feature I could think of built right in.  Somewhat 
inspired by Grant's fabulous work with the Wiard modules:  Lots of 
functionality in a small package.
   The first module, M101A is a classic VCO module, built around a 
modified ASM-1 core.
  
http://www.physicsenterprises.andrews.edu/~clark/diy/images/m101a_01.jpg

   The VCO has 9 waveform outputs:  Sine, Triangle, Tangent, Sawtooth, 
Square, Pulse, double frequency (2x) Triangle, 2x Saw, 2x Square, and 2x 
Pulse.
   I ended up with this many different waveforms due in part to the 
waveshaper circuitry I used to get the sine waveform.  I find it 
different since it isn't necessary anymore to use two VCO's tuned one 
octave apart!

http://www.physicsenterprises.andrews.edu/~clark/diy/images/m101a_02.jpg
http://www.physicsenterprises.andrews.edu/~clark/diy/images/m101a_03.jpg

   These images show a mostly completed PCB for the VCO.  I had run out 
of electrolytics at the time, so the bottom 2/3rds wasn't finished at the 
time of the photograph.  The top 1/3 of the PCB contains the VCO core, 
the other circuitry are the waveshapers.
   As this was the first of these modules, I did some things differently 
than I do them now.  I put all of the resistors on the bottom side of the 
board.  If I were to do it now, I would put as much as I could on one 
side and only resort to putting components on the bottom side if space 
became an issue.
   PCB statistics:  249 components, 588 connections

   The second module I did was a VCA board:

http://www.physicsenterprises.andrews.edu/~clark/diy/images/m201a_01.jpg
http://www.physicsenterprises.andrews.edu/~clark/diy/images/m201b_01.jpg

   I used one PCB to make two distinct modules.  M201A is a quad VCA 
while M201B is a dual VCA with main outputs and soft-start circuitry to 
eliminating thumping on power-on.

http://www.physicsenterprises.andrews.edu/~clark/diy/images/m201a_02.jpg
http://www.physicsenterprises.andrews.edu/~clark/diy/images/m201b_02.jpg

   In the above PCB pics, you can see that M101A uses the top-half of the 
PCB, while the lower-half is used for M101B.  On this board I was able to 
put all of the components on one side.
   PCB statistics: 307 components, 588 connections

   I have one more module, M301, which is a variation on the Moog ladder 
filter (variable pole) plus an onboard ring modulator.  I don't have any 
pictures of it up currently, but I may be able to add it in the next day 
or two.  It is the most complex board I have done to date with over 318 
parts and 685 connections!  For that board I did have to utilize the 
backside to put a few parts, but I limited it to the large 1210/1206 
resistors.

   All of the PCB's use 10mil traces with 10mil separation.  I think the 
latest board uses 8mil traces with 10mil separation.  I'm really pushing 
the limits of the PCB house that I get my boards from.  They guarantee 
8mil/8mil, and although their boards have been great, I did have a 
bridge that occurred on one board.
   Building boards this complex hasn't been without it's problems.  All 
of the boards have had at least a couple of errors (mainly due to 
schematic entry snafus), so I learn something new each time.

   Some other things of interest:  The boards are all completely CV or 
GATE controlled.  That made it more of a challenge, but more rewarding as 
literally everything can be modulated!
   The connectors are VME 96-pin and are great!  I like the fact that 
since the PCB isn't physically wired to the front panel, I can unplug the 
PCB and put in another one (if one should ever fail) without any hassle.
   The front panels are custom punched 1/8" anodizing grade aluminum with 
engraved legending.  The letters are hand-painted and then the whole 
panel is finished with a clear matte lacquer.  
   My first panels were hand lettered and finished with a hard 
polyurethane.  But it was extremely time consuming and frustrating.  In 
the end that kept me from making any more modules until I had found an 
adequate solution to my panel making problem!  I'm pretty happy now 
about the way they look although the lacquer is a bit fragile.  I need a 
good alternative that's more durable.  If anyone has a suggestion, let 
me know!

   Well that's about it for now! BTW, anyone who would like to see my 
modules in person should attend the AHMW show on April 6th.  I plan to 
bring them along with my Modcan and Fenix (plus something special...)  
   Details can be found here:

   http://www.ph.k12.in.us/~drew/ahmw2002.html

   Cheers,

   Tony

-----------------------------------------------------------------------
         The E-Music DIY Archive - New Site Coming Soon!
-----------------------------------------------------------------------

"We wouldn't want to ship something that doesn't work"
   - Carl Stork, general manager, Microsoft Windows division
     Excerpt from EE Times April 2, 2001