Was probing around QCx's two I2C line (ATmega328P pins 27, 28) to see how the Si5351A is configured.
These two lines are pulled up to Vdd line that powers Si5351A by R3(1k) and R4(1k), about +3.5V. That's fine, within the limits of Si5351A spec sheet,
which suggests "don't exceed +3.63V".
Han's design also uses SDA line to drive the RS pin of the LCD display, which may be violating Si5351A' voltage limit....
Have attached a 'scope display that shows SDA(blue) and SCL(yel) traces. SDA starts at left edge already @ 5.0v,
then the ATmega328 writes to many registers of Si5351A, with logic levels within Si5351A specs,
then pops back up to 5.0V afterwards - at the 'scope display right edge. It doesn't stay @ +5.0V forever, but mostly
idles @ 0V, popping up to 5.0V for short periods (4.5ms).
I'm concerned that Si5351A chip may not enjoy being over-voltaged to 5.0V on its SDA line, when it shouldn't exceed 3.63V.