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

Steffen,

Rail-to-rail opamps cannot reach quite to the rails, I would have expected
perhaps within 100mV from the top rail, but I checked a few datasheets and
some devices come surprisingly close, really only a few mV.

The AD623 is supposed to be a rail-to-rail INA, but I see it performs
rather poorly in reaching the rails, the datasheet gives a worst case of
500mV below Vcc.

Your opamp output swing up to 4.7V sounds a bit low, but I would not worry
too much about that.  It might be attributable to measuring errors,
rounding off errors or lower than anticipated VCC.  For instance, the
LM78xx type of regulators have a poor output voltage tolerance.  You might
use a LM7805 which you assume gives VCC = 5.00V and if you measure, find
the voltage only to be 4.82V.  Obviously your opamp would then swing
proportionally lower than what you could expect from a true 5.0V supply.

I can't think of anything that could have your INA misbehave in the way you
describe.  Do you have decent decoupling capacitors over your opamp and
INA?  If you don't, there is always a possibility that a component might
oscillate at a high frequency. A newbie might possibly miss this even when
using a scope.  If an IC oscillates then you typically get these sort of
non-sense measurements.

Perhaps you should try to replace the INA with a new part.  Ideally from a
different batch.  The incidence of grey, copycat, etc. components is ever
present.  At work we had an incidence where an IC just didn't work.  After
desoldering from the circuit, it was found that all pins where
open-circuit.  The IC was cut open and it was established that it didn't
have the actual chip inside, it was simply an empty plastic package (!)
Thank you China-man.  I once received some Texas Instruments INA128
instrumentation amplifiers and each and every one in the packet of about 20
was faulty.   With the INA128s they where obviously faulty but sometimes
components seem to work - sort off.  So you start suspecting all sorts of
other components and debugging becomes a nightmare.

Electrostatic damage due to incorrect component handling is also a
possibility.  That too often means not total component failure but often a
limping component that just doesn't seem to perform right.

In mixed signal circuitry it is common to call the supply rail for the
analog circuitry 5VA, the A meaning analog.  The microprocessor on the same
PCB might also run from 5V and then is often run from a separate supply to
minimise noise.  So the micro's supply you would then call 5VD, the D
indicating that it is for the digital sub-system.  In a similar vain one
often calls a supply +VHT for some high voltage line (HT = high tension) or
you might give a label such as 5VRF for your RF tuner module supply,
3V3_LED for your display supply, etc.

The ISO is indeed an electrically isolated line from a DC/DC converter.

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