I did a little work on the Frankenstein R2 phasing receiver. I noticed that I had a lot more noise and hum on 30 meters and above than I did on 40 and below. I don't think it was common mode hum -- switching to a battery supply didn't help much. But when I took a look at what happened to RF signals between the antenna connector and the input to the DC receiver mixers, I noticed that the signal level seemed to vary quite a bit with frequency.
There were only two circuits in there: a 1.7 MHz high pass filter (to knock down AM broadcast interference) followed by a simple bifilar toroidal transformer signal splitter. My guess is that the 1.7 MHz high pass filter's response was being messed up by the bifilar toroid transformer that I was using as a signal splitter. There might have been some unplanned-for resonances between them. This might have had the effect of knocking down the higher freq signals, making any noise in the receiver (probably from the digital VFO) more apparent. Also, I noticed that I had this toroid too close to the digital VFO box and to the DC power plug for the Arduino/Si5351 combo -- that might have been sending some noise into the DC receiver input.
To make a long story short, I took out the toroidal signal splitter and went with a resistive splitter like the one above. This seems to have helped quite a bit. I know it adds some additional loss -- about 3 db over the toroidal transformer, right?
Another possibility: While rummaging around I found a little 3-terminal TV signal splitter. It is marked "5-900 MHz" but I'm guessing it would be fine down to 1.8 MHz. Any thoughts on trying that?