I've attached a strain gauge to a bending arm of a robot, and it varies by about 1.5 ohms from 356 to 357.5 ohms. I place this in a Wheatstone bridge.
The strain gauge is R3. R1, R2, and R4 are all 378.6 ohms (measured by ohmeter) and I added a resistance in series with R3 to make it range from 380 ohms to 381.5.
I put Vo+ and Vo- as the inputs to a differential amplifier.
Where R1 ~= R2 ~= 1.2 Kohm and R3 ~= R4 ~= 680 Kohm which gives a gain of 567. Vo ranges from about .003 to .006 volts and Vout ranges from around 1.7 Volts to 3.6 Volts. I want my reading to be as stable as possible for a fixed position of the strain gauge but it ranges by about 20 mV peak to peak when the strain gauge is fixed in place. And it is not a stable oscillation; it will just slowly move around that range, changing by about 2 mV per second.
I am measuring this with a voltmeter, and I have also hooked it up to an Arduino. My op amp is an LM741 with +- 8.2 Volts power supply. My voltage source is an Arduino, applying 5 V across the Wheatstone bridge.
I realize that any tiny change in the resistance or voltage source would be amplified and could cause my problem, but I'm not sure how to isolate and fix it. My ohmmeter is only specific to .1 ohms and my voltmeter is only specific to .001 V, so it's hard to tell what is going wrong.
What I can think to check is: how consistent my voltage source is, how consistent my resistance is, and how consistent my op amp is. What is the best way to do those three things given my testing equipment? Is there something else that might be introducing error that I haven't thought of?