How close is too close?

Having finished a couple of oscilloscope measurements, I wonder if anyone has run into a similar issue before. Specifically, I have a current transformer (CR Magnetics 8348-2500-N) and a small transformer (2000:1) mounted fairly close on a PCB. The intended use is power measurements. Both devices feed a differential signal into a MCP3911, a analog front end meant for power measurements.

Today's oscilloscope readings on the analog pins in question were a bit worrying, as the pair of pins dedicated to current showed a sinusoidal differential voltage even with no load attached (i.e. probe of oscilloscope to CH1+, and GND of probe to CH1-). When comparing the output of one leg of the CT relative to AGND instead of the other leg, only noise was apparent.

I am wondering if my CT is picking up a signal from the voltage transformer, which is less than 1/2" away. The CT is at a right angle to the transformer, i.e. the axis that the current carrying wire is aligned with is parallel to the axis from low-to-high on the voltage transformer.

I presume that greater distance and rotating the CT by 90* would help? What minimum distance would you use between a 0.08VA transformer and a sensitive magnetic sensor like the CR Magnetics CT? I'd like to stick with the voltage transformer, if I can, to help with isolation from high voltages.

I think that is indeed too close. Turning it 90 degrees might not help a lot, but I would use a distance of at least 1 or 2 inches (I'm just guessing).

The transformers are inductive and capacitive coupled to the mains. So you will always get some noise with both transformers.

Using the voltage transformer to measure voltage without current still causes the magnetic field. Some cheap transformers are so cheap that the iron core gets into saturation. That would increase the current from the mains and the transformer would get warm.