electrical problems with Galvanic Skin Response circuit ( GSR SCL )

Hi

I'm not very experienced in electronic, but nevertheless i want to 
improve my knowledge and so i try to build an GSR/SCL-device, according 
to Stefan Jungs circuit diagram to put it together with the digital 
board of the openeeg device.
  https://sourceforge.net/p/openeeg/mailman/message/30429870/
Thanks Stefan.

  I choosed the DC skin conductance level measurement as you can see in 
attachment circuit.png.
In the original Lightstone-device (attachment Lightstone.pdf) the 
reference voltage is connected to Vin- and the amplified voltage from 
OP2336 is connected to vin+ of the AD623. In Stefan's circuit this is 
reversed, and since the device didn't work i changed the pins like in 
the original circuit. Was this a bad idea?

Now i've got 3 problems:

1.
According to the AD623 datasheet the 1k resistor R25 between pin1 and 
pin8 makes a gain of 101 (calculation at attachment circuit.png).
Isn't it to much gain?
Supposing a Skin-resistance of 500k i got a V0 of 0,72V.
So the difference will be: V0-Vref = 0,72V-0,5V=0,22V.
V1=0,22V*Gain = 0,22V*101 = 22,22V
22,22 V is to much for the Atmel AVR ad-converter.
Where is my fallacy?


2.
R25=25k --> Gain=5
I get 1,8V, 50Hz mains hum when i'm connected with electrodes at my 
fingers to the circuit.(attachment hum.jpg)
When i connect a 500k resistor at the input there is no hum available.
Is this normal and do i have to build a band-stop filter somewhere in 
the circuit or could something other be wrong with it?
I dont understand why there is absolutely no milliVolt of hum when i 
connect a resistor.


3.
A bit confused, i decided to make some measurements with a resistor 
instead of my skin (attachment measurement.png).
So i simulated  the skin-resistance Rscl with a potentiometer from 100k 
to 900k.

To get an idea what is wrong with the gain of the AD623 i made three 
measurements with Rg=1k, 10k and 50k.
The values for V0 are fine and match to the theoretical formulas for OPamps.
At first appearance V1 looks good, but if i calculate the gain with 
V1/(V0-Vref), the gain changes at a wide range.
Shouldn't the gain be constant dependent on the value of R25 or is this 
an override-effect, because the output voltage v1 should be much higher 
in some cases.
To see how big the gain theoretically  schould be, i calculated the gain 
in the last row with the formula from the AD623 datasheet.
Gain = (100k/R25)+1.
Only in some cases the theoretical gain and the calculated gain from the 
measurements are equal.


Many questions, i hope you understand my english and can give me some hints.
Thanks

Steffen

------------------------------------------------------------------------------
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