I am experiencing repeated device failures (3 so far) shortly after turn-on. Schematic and related layout are embedded below. We attempted to adhere close to the schematic from the datasheet (400 MHz operation). There is a full ground plane (not shown) on one layer beneath the part.
Is there a defined startup sequence? I am currently driving the input regardless of supply switch state (about +12dBm). Or does the bias need to be settled prior to applying the input drive?
I do not see any overvoltage or undervoltage spikes on the supply rail, so that does not seem to be the cause.
The device is on for no longer than 600mS, off for 200mS (basically CW morse code). The entire test sequence is approximately 12 seconds.
The 5V supply is protected by a 700mA eFuse.
Each device has operated for a few cycles before failing. After failing, each measures about 5 ohms to the ground tab.
I haven't been able to look for oscillations because the device fails rapidly. My current plan is to replace the device and bring the supply up slowly so that I can actually make measurements without damaging the device.
I have also measured the impedance looking into the output network and it appears to be close to the expected impedance (at 400 MHz). Is it possible to damage the device with a bad VSWR? The RF switch state is set to the load, but it will ultimately drive an antenna, so I'd like to know the device limit or if we need to do something to limit the VSWR seen by the amplifier.
