Can measure DC as well. Wow!!I think most all of us see that 330k and 1meg resistors reduce the line voltage only slightly. So much of the line voltage could make it to the amplifier. So the output of the amp is a " square wave signal". The input common mode range of the amp can not take that kind of signal. Reduce the 1 meg to 5k to get the input voltages down. (still other problems but …)
Here is a IC that isolates from the power line, and hands you 0.4V/amp up to 5A. It is a good place to start looking for safe ways to measure power line current.
View attachment 189675
The signal you want to see is getting smaller, with the 0.1 ohm resistor, and the voltage you substract is hundreds of volts. The gain is already so high the output of the amp is a square wave. The input to the amp is over driving the amp.your circuit just might work. …. give your amp a gain of 10
I agree with that.bipolar supply
On retrospect, I believe that you are correct. Even though the difference between the inputs is small, the individual inputs are way too high for your common variety, everyday opamp, with peak voltages around 165V. I'm now surprised that the opamp didn't die immediately. The OP is going to have to use a specialized opamp that is designed for these voltages, or implement a discrete version. It's possible that a resistive voltage divide prior to the opamp may work, say twin divide by 10.The signal you want to see is getting smaller, with the 0.1 ohm resistor, and the voltage you substract is hundreds of volts. The gain is already so high the output of the amp is a square wave. The input to the amp is over driving the amp.