I've been working on a direct conversion receiver for 40 meters based on the classic SA602 mixer design. Circuit works great on the bench with a signal generator, but when I connect my Si5351 VFO module, I'm getting significant noise and spurious responses. The noise seems to be coming from both the MAX7219 display and somewhere else in the Si5351 circuit. Has anyone encountered similar issues with Si5351-based VFOs in DC receiver applications? I'm using the Etherkit library with an Arduino Nano. Any suggestions for RF filtering or isolation would be appreciated.

Classic problem with digital VFOs in sensitive receivers. Try adding some ferrite beads on the power and I2C lines, and consider switching away from the MAX7219 display - those things are notorious noise generators. I had good luck with a simple LCD display instead.

Are you using proper decoupling capacitors on the Si5351 power pins? Also, make sure you have adequate shielding between the VFO and receiver sections. Sometimes a simple metal partition helps immensely.

I built a similar setup and found that using the Si5351's quadrature outputs directly rather than dividing from a single clock reduced some of the noise issues. Worth trying if your mixer can handle quadrature LO injection.