HMC547ALP3E RFC connected to HMC641ALP4E RFC with no AC coupling cap, weird insertion loss behavior

I have a fan-out circuit suing a HMC547 (to select src or cap) connected via RFC to a HMC641 to mux out to 4 lines

The board has AC coupling caps on all the RF inputs but the RFCs are connected together DC coupled (oversight)

I am getting weird issues, on 90% of the circuits it works fine but on the last 10, we see weird insertion losses (that can change)

If I put a meter on the RFC to check a diode for instance, I can change the behavior.

I'm going to respin the boards, but do you think this DC coupling may be the issue I am dealing with?

Parents
  • Hi BrendanT,

    I see you have DC blocking caps on the RF lines connecting this block to the previous and next block.
    The RF ports are DC coupled to 0V and these connect to the drain/source of the RF transistors. As long as your control signals are clean, and the both devices are operated with the same VDD voltage level, the DC blocking capacitor between the RF ports of the two ICs may be omitted.
    If you are going to respin the board, the DC blocking caps are a good idea to include. On a side note, have you considered using our newer SOI switches, ADRF5020/ADRF026 or ADRF5044/ADRF5045? These have superior performance in terms of insertion loss, switching speed, etc.

    About the issue you are seeing - there could be something else going on. What kind of a weird insertion loss behavior do you see? Are the currents OK when the insertion loss changes?

    Regards,
    HK.

Reply
  • Hi BrendanT,

    I see you have DC blocking caps on the RF lines connecting this block to the previous and next block.
    The RF ports are DC coupled to 0V and these connect to the drain/source of the RF transistors. As long as your control signals are clean, and the both devices are operated with the same VDD voltage level, the DC blocking capacitor between the RF ports of the two ICs may be omitted.
    If you are going to respin the board, the DC blocking caps are a good idea to include. On a side note, have you considered using our newer SOI switches, ADRF5020/ADRF026 or ADRF5044/ADRF5045? These have superior performance in terms of insertion loss, switching speed, etc.

    About the issue you are seeing - there could be something else going on. What kind of a weird insertion loss behavior do you see? Are the currents OK when the insertion loss changes?

    Regards,
    HK.

Children
No Data