Problem with HMC253

[electronicmonk]

I am using the HMC253QS24 non-reflective SP8T switches for a project. While testing the switch with an input signal at the RFC pin varying from 1-10MHz, 20-100 mVpp from a USRP Tx, it was observed that the signal passes to all of the 8 RF outputs regardless of the control pins selection.
I have tested the output with an oscilloscope as well as on the USRP Rx, with the same result in both the cases.
I have tested, both, with and without the 100pf capacitors(as present in the evaluation board shown in the data sheet) at the input and outputs of the HMC253 but the results remained the same.
Multiple HMC253 IC's were also tested to no avail.

Thank you.

[jmsigler]

You should post schematic, its pretty hard to say what the problem could be without it.

[electronicmonk]

The schematic is attached.
This is exactly identical to the one in the datasheet.

Thank you.

[CJay]

More than the schematic, the PCB layout is critical to achieve good performance from this or any other RF switch, are you using the evaluation board, a copy of the evaluation board or something designed with 50R impedance 'lines' because if you're not, it's going to be awful.

You get a minimum 32dB isolation from DC to 1GHz so I *think* there will be some signal leakage no matter what you try and achieve, what sort of levels of signal are you seeing on the outputs?

[electronicmonk]

The problem was solved by using capacitors of higher capacitance on the input and output's,
We used 1nF capacitors which resulted in considerably better results .
The solution was suggested at the analog devices forum.
link :
https://ez.analog.com/thread/97697#comment-312139


Thank you.