Load cell interference

[Braeden Hamson]

I built an instrumentation amplifier following this circuit exactly.
https://www.allaboutcircuits.com/textbook/semiconductors/chpt-8/the-instrumentation-amplifier/
I used a load cell out of a wii balance board. And one output I have two LEDs in series to monitor the output voltage of the whole circuit. When I apply pressure to the load cell one way and another the led's become darker and brighter. This is good, it's what I was looking for. However, the problem is that if I bring my hand near the load cell without touching it the lights become brighter. I connected a jumper lead to the load cell body and connected it to ground, which didn't help. Interestingly when I detached the lead from the load cell and it was still connected to the breadboard ground and raised the wire up the output voltage of the circuit goes up. So I'm certainly getting RF interference. My question is how I can I minimize this effect or normalize it?

 

[(*steve*)]

Shielding of your current may help, also reducing the gain at high frequencies may help. Try a capacitor across the resistor which sets the gain of the instrumentation amplifier. And don't leave the inputs disconnected.

I'm not 100% sure that a capacitor across the gain setting resistor wont cause instability, so beware of that too.

Edit: that won't work. An inductor in series with the gain setting resistor might though (because increasing resistance lowers the gain)

 

[Braeden Hamson]

No unused inputs, I'm tempted to get that tattooed on my arm. Maybe a dip 8 with lines extending out with ground symbols on the end, could look cool.

I'll try putting a capacitor on the gain resistor. I'm also using one of those board mount trimmer pots, if that means anything. Should I use a high farad or low farad capacitor? (not like super cap big but you know)

 

[(*steve*)]

I'll try putting a capacitor on the gain resistor

DONT do that. Read my edit.

What is the value of your current gain setting resistor? If you can find a 1000uH choke, try adding that in series. If your gain is critical, remember that the choke will have a DC resistance.

 

[Braeden Hamson]

I didn't try the cap so no worries. I'll try the choke, and the gain value isn't critical yet. It will be eventually as id like a 0-5 v range. I'm using 1k resistors for the whole circuit. And a trimmer in parallel with a 400 odd ohm resistor. What would the capacitor have done?

 

[(*steve*)]

A low value gain resistor gives you a high gain, so a capacitor in parallel with the gain resistor would both increase gain for noise, but also potentially lead to instability.

 

[Braeden Hamson]

I was on the magic Google machine and I found out that this specific problem of load cell noise is a rather big issue. Some used caps between signal wires and so on. One study (white paper? Are such publications still called studies in engineering?) Managed to reduce the signal to noise ratio by 27 db. But I only took a cursory look at that. I'll need to read it in full.