From: sleeper75se (sleeper75se_at_yahoo.se)
Date: 2001-12-16 17:42:32
Hi Joerg and everybody,
here are a bunch of cost-optimisations I think we could make. We can
save perhaps $10 here, assuming we have the PCB's (that are $15 each
for euro-size) factory made.
All resistors and capacitors that can be replaced with surface mount
parts, should.
Not only are smd-resitors smaller, they are also 10-30% cheaper (at
my local reseller anyway).
I feel that size 1206 is easily solderable by a beginner, and
resistors and capacitors are cheap so the psychological barrier (that
is, fear of failure) should be low. We can even make a tutorial ...
This excludes filter capacitors, because smd caps have very low
precision.
If we do this, we can perhaps cut the board size in half, halving PCB
manufacturing costs. $7.50 saved...
Suggestions for the gain stage (IC9A, IC5B):
The gain can be trimmed on the PC side, in software.
If we remove P1/P2 and set R37/R40 = 10K, we'll save at least $2.
Suggestions for the lowpass filter:
For the surface mount option to work the resistors will have to be
replaced with more common values:
Replacements values
-------------------
pole 1-2: 13.7K => 13K and 332K => 330K
pole 3-4: 38.3K => 39K and 357K => 360K
pole 5: 21K is optimal - needs two resistors (20K + 1K).
For G=1.02 @ 50Hz (rather than G=1.00) use only one of 20K or 22K.
The gain errors caused by lowering pole 1-2 values are offset
somewhat by rasing values in poles 3-4.
So, what do you think?!
Also, if you have the time, please give me your thoughts on biosemi's
artikel7 from an earlier post.
Regards,
Andreas
This archive was generated by hypermail 2.1.4 : 2002-07-27 12:28:33 BST