# Measurement of Galvanic Skin Response

1. Apr 6, 2013

### arunks91

1. The problem statement, all variables and given/known data

I have attached the photo of the circuit. My doubt in the circuit is that, I could not exactly figure out what the diodes do. The description says it is used to set the input of the op-amp to 1.6V above v-. But how exactly does it achieve that? Can somebody show it mathematically?

Here is also the link of the website where I found the circuit. The circuit in question is the last circuit in the page.
http://produceconsumerobot.com/truth/
2. Relevant equations

3. The attempt at a solution

File size:
5.8 KB
Views:
164
2. Apr 7, 2013

### rude man

Unless the LED conducts in the reverse direction the three diodes conduct no current and therefore there is no voltage drop across those diodes.

I would need a data sheet to determine the reverse breakdown voltage of the diode. Or maybe it's actually two diodes wired internally back-to-back. These are designed to emit one color light with current in one direction and another color when current flows in the opposite direction.

Your question makes sense only if the LED conducts in the reverse direction, in which case there will be a total of about 1.6V across the diodes and then obviously the + input to IC1A is 1.6V above the output.

If the LED current is in the indicated forward direction then there can be no current through the three diodes and therefore no voltage drop across them. They effectively don't exist.

Or maybe the diagram shows the LED hooked up backwards.

3. Apr 7, 2013

### Staff: Mentor

And the input circuit would seem to have a problem. You are applying a DC voltage, and AC coupling the result through to the first opamp? Where did you get this circuit?

For a GSR measurement, you need to apply a human-safe (small, high impedance) AC voltage, and measure the resulting AC current...

4. Apr 7, 2013

### rude man

Well, there is a V+ going to one of the skin electrodes so an ac voltage could be developed going into the input cap: V_in = (V+ - V-)R4/(R4 + R_s) with R_s= skin resistance modulated by some ac method. As for V- that would appear to be some kind of positive feedback ....

The signal they're looking for is actually ac, see the proffered link.

EDIT: Nope, not the heartbeat, obviously. Just a sudden change in skin resistance when a question is asked, looks like. They're obviously looking for the (relatively) sharp falling edges. The rest of the waveform is too slowly changing to get thru their hi-pass.

Last edited: Apr 7, 2013
5. Apr 7, 2013

### Staff: Mentor

Yeah, but GSR is a [STRIKE]DC measurement[/STRIKE] DC value measured with an AC excitation. Or a slowly-varying measurement on the order of 1Hz as the person sweats more. Professional instruments use a very low-level AC excitation to measure the AC impedance to display as GSR.

EDIT -- and those instruments need to be qualified under UL-544 Medical Devices for use in the US. I don't know the applicable medical safety standards for the EU and Rest of World...

6. Apr 7, 2013

### willem2

Did you see the connection to V- at the top of the circuit? The wires crossing between R3 and the diodes must be connected (and there should have been a dot to make it clear), so C2, the diodes and R3 produce a reference voltage of (V- + 3 diode drops), that is tied to the positive inputs of both opamps.

7. Apr 8, 2013

### arunks91

Yea that was my doubt as well, but people across many forums say that the circuit seems to work. The circuit is in the link I provided. Although when I rigged it up, the LED just keeps glowing and really doesn't change state at all. Will it work if I make the input an AC signal? And also what exactly does the capacitor parallel to the diodes do?

Last edited: Apr 8, 2013