Ground issue?

I was using my Nano to check some SPI signals on an AVR. DC levels and slow signals read fine, but the SPI clock was giving strange readings. Voltages were showing up in the 15-18 v range (actually 5v and then passing through a voltage divider down to around 3.3v) and had a “mixed in modulation” (the clock pulses moved up and down sort of in a AC sine-wave, but centered on the 15-18v level). Unfortunately, I didn’t get a capture of it.

I noticed that holding the Nano closer to the circuit made the “modulation” disappear and the voltages dropped to the expected 3.3v. After a little tinkering, I realized that making a electrical connection from the backplate to the signal ground had the same effect, which suggests that the Nano can be susceptible to some sort of noise?

Has anybody else experienced this? Any suggestions?