We use two ADF4350 at our custom board.
Both synthesizers have the same behavior
We are using RFoutA+ only, while the complementary output RFoutA- is terminated with 50 ohm resistor to GND via 100pF capacitor (see ADF4350_Schematics.png)
I've made a couple of changes (comparing to the schematics attached):
1) Replaced biasing inductors (L30 & L31) by 50 ohm resistors
2) Removed terminating resistors (R73 & R74) from an unused B output. I've cut the traced leading from the chip to these resistors
It's important to notify, that I'm using some custom made probe for test purposes.
Now, I remove the DC blocking capacitor C125, and connect an RF probe at RF1outA_P (C125 pad).
I measure 6dB less than the programmed level. For example when I program it to -4dBm, I measure -10dBm. May it happen due to using single ended output? It can explain the 6dB attenuation (since the output voltage level is 0.5)
The second problem is output derating with frequency increase. For example I measure -15dBm at 3GHz. Is it acceptable?