Re: electrical problems with Galvanic Skin Response circuit ( GSR SCL )

Hi Steffen

I just checked my archive, although I have more than one version of the
schematic, they all show R25=1K.  You are correct that 1K gives a gain of
101.  I scratched out my one and only prototype of this circuit and see
that I used a 1003 resistor in that position, i.e. a 100K which gives a
gain of 2.

I'm not surprised at that documentation fault.  I design a schematic, then
the PCB, then build and test.  For a prototype I don't always go full loop
and update my schematic after making adjustments.  Sorry.  Often some
thumb-suck values are placed on the schematic with the intention to do
final calculations/adjustments at the time of testing/debugging.  Yup,
documentation is never much fun, is it?  And often the lack thereof comes
to bite you later on.

Also, the inverting and non-inverting inputs to the AD623 are wrong and
need to be swopped (pin 2 and 3).  The other PDF showing the SCL
calculations, also shows the basic circuit diagram, there the inv & non-inv
inputs are shown correctly.

BTW, that circuit is not the original Lightstone.  It is loosely based on
it.  The real Lightstone uses 2x AD623 instrumentation amplifiers.  The
AD623 is not as expensive as the AD620 and similar INAs but I decided to
change the front-end and use an opamp instead.  This was based on some
circuits that I found on the internet from which I further refined the
concept.

I did fully reverse engineer the schematic of the Lightstone.  Similarly I
reversed engineered the full circuit of the Emotiv EPOC, but I would not
publish these on an open forum.  I feel that that wouldn't be quite right.
I have, however, sent the circuit back-channel to one or two people in the
past, normally asking not to distribute further.  For those that follow my
occasional writings will know that I have even back-engineered the (analog
only part) of the QDS Excalibur 4-channel EEG amp - only from some color
photos I was given.  As you can imagine, back-engineering something like
this can be a lot of work but you learn a lot of things.  How to do things
and how not do do things. LOL.    In the case of the EPOC, I had to
desolder a 20-pin 0.05" pin header which holds the digital and analog
boards together.  That job in itself is not easy.  Having spent the time
reverse engineering something, I normally draw the schematic in CAD - that
is quick enough to do.  With SMT circuits, component values are not always
noted because SMTs mostly are unmarked and I am only going to lift them
with a soldering iron and measure if I consider it a critical value.

I only glanced at your message.  Implement the gain change and invert those
two pins and test again.  If you still have problems I can have another
look and see if I can give advice.

Regards
Stefan

------------------------------------------------------------------------------
Learn Graph Databases - Download FREE O'Reilly Book
"Graph Databases" is the definitive new guide to graph databases and their
applications. Written by three acclaimed leaders in the field,
this first edition is now available. Download your free book today!
http://p.sf.net/sfu/13534_NeoTech