Problem with ADRF6720 - at warming up to 40 degrees (IR thermometer measured, VPTAT = 1.3V), the output signal is lost. Оn the spectrum analyzer is like a self-excitation of the PLL - a narrow spectral components is scattered in the noise of 0.5 - 1 second. Disappears as the parasitic carrier wave and a modulated signal.
After switch-off and cooling down chip, the signal is restored.
ADRF used in custom topology, six-layer PCB, with good heatsink. At beginning used PLL filter nominals (and all other) from datasheet recommendadions, now i recalculate PLL filter in ADIsimPLL, and no changes in breakdown.
ADRF configured for 2080 MHz carrier, used 5 MHz test modulating IQ signals 0.2Vpp. Supply 3.3V from personal LDO, with nominal voltage and current consumption.