Finished my DC40-based receiver last week and getting some strange results. Running QRP Labs VFO at 7.050 MHz but hearing strong audio breakthrough from about 7.080-7.090 range when tuned to CW portion. Diode ring mixer appears wired correctly - used 1N4148s with matched forward voltage drops within 5mV.

Audio output shows clean signal on 'scope but breakthrough is definitely there. Anyone experienced similar with direct conversion designs? Wondering if I need better band-pass filtering ahead of mixer or if this is oscillator isolation problem.

Using Manhattan construction on double-sided PCB material, copper pads every 0.1 inches. Supply is clean 12V with proper decoupling throughout.

Check your VFO drive level into the mixer first - too much LO can cause overload symptoms exactly like you describe. Should be around +7dBm. Also verify your I-Q balance if you're doing phasing - unequal amplitudes in the signal paths create unwanted sideband breakthrough.

This sounds like insufficient front-end filtering. The Band Pass Filter should reject everything above and below your target band. A 40m BPF with sharp skirts is essential for direct conversion work - the mixer sees everything your antenna picks up.

DC receivers are sensitive to LO stability issues. If your local oscillator isn't locked properly to the signal you want, you'll get frequency offset problems that sound like what you're describing. Consider adding a PLL or using a more stable reference.