HMC487LP5E RF amplifier drawing excessive current

Hi there,

I was hoping to possibly get some help regarding the HMC487LP5E RF amplifier drawing an excessive amount of current.

For reference: We designed an RF board that combines the ADF5356 RF synthesizer and the HMC487LP5E RF amplifier. It also contains the PS setup for the 2 circuits along with a Transimpedance amplifier based off of OPA858. The Synthesizer circuit and the Amplifier circuits are based of the IC's eval boards (ADF5356 and HMC487LP5E)

We have 2 of the boards already up and running for months without problems but we have had problems with 2 others. 

When connecting our working Custom RF boards to our Power supply, the max current drawn on the boards that worked was approximately 430mA @ 11.9V (input to our power supply)

-->RF synthesizer supply: 6V

-->RF amplifier Vg: -0.74v   Vd: 6.3V

When connecting the ADF5356 eval board along with the HMC487LP5E eval board to the same Power supply, the max current drawn is similarly 410mA @ 11.9V

-->RF synthesizer supply: 6V

-->RF amplifier Vg: -0.74v   Vd: 6.3V

When connecting the Problematic Custom RF boards to our Power supply, the max current drawn is around 1.3A @ 11.9V

-->RF synthesizer supply: 6V

-->RF amplifier Vg: -0.74v   (gate voltage range jumps from 0V..-2V to 0V..-4V)

-->Vd: 6.3V

I realize that according to the datasheet, the HMC487LP5E can draw up to 1.3A, but occasionally when our entire unit is setup, that already high current has experienced a surge for whatever reason resulting in the IC shorting out and the problem is, the RF amplifier is expensive. Seeing how 2/4 boards worked without a hitch logic would dictate that there is an outside source causing the problem. I have uploaded the schematics for our board, in case it helps-but I was wondering if you could recommend some sort of protection circuit, that could potentially be added to our board to try to limit the current to the expected value (the eval board run at about 430mA). Ironically there was/is no difference between the boards that 'worked' and those that drew excessive amounts of current.

I would greatly appreciate any insights on the matter. 

Regards,

Tayeb