How to eliminate 50 Hz noise from amplifier's output

My goal is to amplify signals with a frequency from 100Hz above. My circuit is looking as the picture below. With an oscilloscope, I am trying to measure the output of the amplifier. If the input of the circuit is not connected at all, I can see a small 50Hz signal. Touching the input gives me a huge 50Hz signal which comes from the AC power supply from around the room. But I expected them to be eliminated due to the capacitor added.

I know that the oscillator's probe behaves like an antenna. Maybe it induces the noise into the feedback of the amplifier? But then, I don't understand why touching the input increases this noise.

• The human body may act as a antenna too. Commented Nov 20, 2014 at 15:25
• Why did you expect the capacitor to eliminate 50Hz AC? Your 1uF capacitor only has a reactance of 3.18k ohms at that frequency.. Commented Nov 20, 2014 at 15:27
• What I understand in the article about high pass filter on wikipedia, this circuit is an active high pass filter filtering AC signal below 100Hz. en.wikipedia.org/wiki/High-pass_filter Commented Nov 20, 2014 at 15:40

Your filter is a very mild one (single pole) and won't have any dramatic effect on the mains-frequency pickup. Also, trying to get a gain of 1000 in one stage is a bit demanding on the op-amp. Here is what the response of your circuit (built with an OPA177) looks like with a 1mV input. The gain is 1000 at 300Hz, and above that the op-amp pole knocks it down.

If you really want to seriously knock down response at 50Hz (rather than reducing it as much as possible, which should be your first thought) without affecting the 100Hz signals much you'll need a much higher order filter.

Ignoring the op-amp rolloff, below is the shape of the response of the filter:

• but even with a filter of first order, i must have seen some improvement which i didn't. Commented Nov 20, 2014 at 15:42
• @YannickWald You can't know if you've seen improvement with a first order filter unless you've looked at it without that filter in place. I suspect it is a bit lower amplitude than it would be without the filter. Commented Nov 20, 2014 at 15:49
• What Scott said- you can't do an accurate A/B test with a finger. Commented Nov 20, 2014 at 16:08

It's not the oscilloscope probe behaving like an antenna -- it's YOU. You are capacitively coupled to noise in the environment. The cap prevents DC signals from reaching the amplifier, not AC signals.

If you don't want to see the 50Hz noise, hook the input up to the signal you're trying to amplify, and don'd touch the input pins.