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

Hi Stefan,


> 1. A high skin resistance gives a low output voltage, a low skin resistance
> value gives a high output voltage.
>
> 2. The above point can be proven by inspection - if skin resistance is
> infinitely high the opamp acts as a simple unity gain amplifier and the
> same 0.5V Vref will be applied to the + and - inputs of the INA, the
> differential voltage is thus zero and the INA's output is zero.
>
> 3. If the skin resistance is 0 ohms, the opamp's gain will be infinitely
> high and it will saturate at it's Vcc.  The gain of the INA should be
> adjusted to then give a FSD value compatible with your ADC's maximum input
> voltage (also FSD).

This is the case for 1. and  2., but for 3. the Output is not 5V, but 4,7V.

> 6. Do not randomly guess resistance values.  Choose a combination of
> resistors, calculate the expected voltages at the opamp and instrumentation
> amplifier's outputs and then verify with a multimeter or scope if the
> measure values correspond with the theoretical values.  First ensure that
> your opamp works as expected and then check correct operation of your INA.

I didn't guess totally randomly, i calculated with an Excel sheet values 
and tried to get an output voltage over the full  scale, but while i 
studied the formulas i came to the conclusion, that it is impossible to 
get a full scale output voltage from 0,5 to 5V if the skin resistance is 
alternating between 800k and 900k however i  choose the combination of 
resistors.
Maybe it is not an error, but willed so from the designer of the 
lightstone as you describe at 9.

> 7. To do point (6) consider the opamp as just a simple non-inverting
> amplifier.  The gain is set by  R24 and SCL.  The "input" voltage is simply
> Vref, which I have fixed in my design at 500mV.

The opamp works fine and the calculations correspond with the 
theoretical values.


> 8. The INA is simply a difference amplifier.  You take the output voltage
> of the opamp, subtract Vref (0.5V) and multiply by the INAs gain.

The INA doesn't work as accepted. I operated the INA standalone with a 
voltage divider made of resistors and the behaviour is always the same.
The output of the INA is  to less.

f.e:

R25=1,47k-->Gain=69

+Input:0,560V
-Input: 0,486V
difference:0,074V

Output Voltage: 2,235V
Gain:30,2
Gain_theoretical:69

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
with increased power supply of 6V:
+Input:0,847V
-Input: 0,566V
difference:0,281V

Output Voltage: 2,79V
Gain:10
Gain_theoretical:69

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+IN to ground (single-ended)

+Input:0V
-Input: 0,714V
difference:0,714V

Output Voltage: 2,54V
Gain:3
Gain_theoretical:69

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Maybe my INA is defect?
Btw, what does 5VA ISO mean? VA is technically Watt?
Does ISO mean isolated for safety reasons?



>
> 9. Although the skin resistance always tends to be high, the circuit was
> designed to measure down to nearly zero.  This is also what the lightstone
> does.  If you don't like that operation, you would need to shift the INA's
> DC offset by applying a low impedance source, DC offset voltage to pin 5
> (Vref of the INA).

I will test this, when the problems from 8. are solved

Regards 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