Noise between vcc and ground on BGA2869

[Uknown4Ever]

I designed a GPS L1 band receiver circuit with simple amplifications using BGA2869 and LNA (ADL5523 ) , the circuit consist of sma connecter -> LNA(ADL5523 at L1)-> stage1 (BGA2869-5V) -> passive BPF ->  stage2 (BGA2869-5V).

all the components design using the recommend PCB on the datasheet. the issue is when I tested the board using normal sweep from 1.3 GHz to 1.7 GHz (L1 range) I was able to amplify the signal but I was able detect a single stone noise of -50dbm and another noise sweeps through the L1 ranges on the test point between stage1 and 2 (BGA2869-5V) vcc and ground on spectrum spectrum analyzer.

then i removed the amplification second stage (BGA2869-5V), the single tone noise  was removed but i am still stuck with sweep noise.

so do u have any idea to solve this problem i just want to remove the noise between vcc and ground.

thanks in advanced.

Ahmed

[LPP]

Probably, the noise is caused by parasitic feedback from output to input. A chain of three LNA amplifiers provides total gain more than 84dB. The isolation between output and input must be more than 94..100dB to avoid unstable operation. Turning off the last LNA, reduces the loop gain.

Common reasons of instability are poor supply decoupling and poor layout (proximity of the signal and power traces).

Good practice is to feed each stage of the amplifier through 3-terminal capacitor in series with ferrite bead (Murata NFM + BLM). Capacitors should be located close to the LNAs power pins.

Best Regards,
Pavel
NXP CAS

[Uknown4Ever]

I designed a PCB to test a single BGA2869 with its own power supply input and output, the circuit design is based on the proposed circuit on the datasheet. 
I tested the circuit using a network analyzer on three SMA test points, BGA_Inp_SMA, BGA_Out_SMA and VCC. 
network analyzer ports are set to Port1=BGA_Inp_SMA , Port2 = BGA_Out_SMA and the output is :

the output does not make no sense.
I would really appreciate the help.