I want to measure the electrical resistance of my fingertip. I'm building a thimble with a few sensors and one is this resistance probe. My multimeter tells me the resistance is around 2 MΩ. A simple voltage divider might not work. So I tried this circuit:
A voltage divider feeds 1.2 V to an opamp. The voltage at A1 is calculated as:
1.2 * (1 + X / 2MΩ)
In the end I just need numbers for a machine learning thing, so I'm satisfied with what I have. I read A1 on my Arduino analog pin as 1023, when not touching X. And I get values around 800 when I touch X.
But how accurate is this? 800 corresponds to 3.9 V. That leads to X being 4.5 MΩ which is quite plausible. I don't know much about opamps, so I don't know what values I have to look at in the datasheets. Are the resistors in the MΩ magnitude too large for opamps? And how much lower can the reference voltage go in my voltage divider?
[EDIT]
Fixed some numbers there. I calculated the voltage wrong.
You even can short-circuit X and get the same voltage as provided by your voltage divider at the + input of the opamp. Rule of thumb. the opamp tries to keep its inputs at the same voltage.
It will go down considerably (10k) with wet fingers or harder press. Also check the resistance between your hands.
Yes, I have noticed how the measured values vary a lot. That's why I have to focus on getting a good design on the thimble, which should fit perfectly on my finger. It squeezes gently around my ring finger, when I put it on.
The uA741CP has some offset inputs, which I haven't used. I haven't really found info on how to use them. Do they have to be grounded or set to +5 if not used or can they "float"?
What do you mean? It already does run nicely. 5 V and GND is connected to the opamp, as well as + and - and out.
Between two spots, 10 mm apart from each other. They are two 1 mm metallic pegs touching the skin of my finger, when the finger is in the thimble. By designing the thimble I wish to create a testing condition that will be as constant as possible from time to time. All involved sensors will be at exact the same spot on my finger. Pressing with a constant force.
So I try my best with the resistance sensor, knowing that I may not reach what I want, a stable reading significant to my purposes. Even if I manage to get a stable, repeatable value, the value itself might not correlate to blood pressure at all. What I'm doing is namely measuring how much red, green, blue and infrared light passes through the finger tip, what's the temperature and electrical resistance of the finger tip. All measured values and their rate of change might correspond to measured blood pressure. My project is about finding out that.
That's not a requirement but a meaningful modification of the circuit. Nowadays, 50 years after the 741, 5V rail-to-rail opamps are available and almost industry standard.
But I don't want to compare. I want to measure. The 2 - 4 Megaohm resistance should translate to 0 - 3.3 V, which I measure at an analog pin. Right now I'm testing it on an Arduino, so I aim at 0 - 5 V, but the final project is on a Cypress PSoC 62S2, which deals with 3.3 V.
And I ordered some MCP6042 for this project. Should work better than the 741.
