Measuring Voltages Reliably and Meter/O-scope Noise/RF Pickup

[PsychedelicBreakfast]

I've been asked to design a load cell amplifier. The load cell has a sensitivity of 2mV/V and at the recommended 10V excitation I should get a reading of 20mV at the rated load.

I've used an instrumentation amplifier to amplify the differential voltage (which is usually around 2mV or so as we are not using the load on it's rated load yet). The problem I keep running into is that I notice there is a lot of noise at the output and the input - even if I aggressively put filters at the inputs and outputs. My guess is that my o-scope's probes (Rigol DS1052) are picking up RFI from the nearby radio tower (1.3 MHz). If I disconnect everything from the o-scope, the noise floor is a few hundred uV. If I connect the probes, the noise rises to a few mV.

Even more peculiar is the fact that if I have a probe connected to the output of my in. amp and I try to measure something with any of my multimeters the o-scope shows a very strong 50 Hz signal. In other words, touching the in-amp with any sort of wire/cable is inducing a 50 Hz signal. Here my guess is that the wires are picking up 50 Hz mains frequency. The meter's display also fluctuates a bit (a few mV here and there) when connected to the output of the in-amp so I can't even be sure of my readings. This has caused me a lot of headache as it seems that my gain varies with input voltage and does not seem constant at all - I suspect this is because of noisy measurements.

So I'm just looking for advice on figuring out how to measure these low voltage signals i.e. how do I know, with reasonable certainty, what I see on my o-scope is real? How do I know that the meter's fluctuation is because of 50 Hz pickup and not because the output is actually fluctuating?

[Pippy]

Well, if the scope has a software filter setting you could maybe set that to low-pass filter mode and something like 100kHz or so (or much less) to attenuate the 1.3MHz.

Multimeter leads act as VERY good antennas, they will certainly mess up any analogue/RF circuit probing. Any RF/EMI picked up by your multimeter leads (antennas) will be fed directly into your circuit.
You could maybe rap one leg of a high value resistor around the multimeter red lead (black lead to circuit ground) and probe your circuit using the other leg of the resistor. A value of say 100k might be OK, but you'll need to take that into account when wanting an accurate voltage reading. If your trying to measure AC voltage though it might not work as holding a multimeter lead will induce a large 50/60Hz signal into it.

[AndyC_772]

Could you describe the technique you're using to connect the scope probe to your signal?

With low level signals like this, it's all about the layout. A photo of your circuit showing the scope probe connected could be very revealing.

Also, if your circuit isn't built on a PCB with a solid ground plane, then you're not going to get far.