RF witch 232A isolation

I am working with hmc 232A switch. The PCB used is 2 layer Rogers 4350B 10 mil thick. While testing at 9-10 GHz, I am not getting isolation as projected inthe data sheet. Can we cascade three switches to improve isolation to a higher value something like 90 dB. Do I need to house the pcb in an aluminium housing during testing. Can anyone suggest how to proceed to get rated performance.

R R Rao

Parents
  • Thank you very much for your reply. I am providing  further inputs for your review and feedbak.

    a)        My  PCB design is as follows.

    The PCB used is RO4350B with 10 mil thick and copper thickness of 1 ounce.

      1.  The PCB  transmission line width for 50 ohms characteristic impedance used is  21.35 mils.

      2.  Ground plane was extended on both sides of the transmission line at a distance of 20 mils for each plane.

    b)        HMC 232 ALP4E Evaluation board (sent by you)

      1.  The PCB transmission line width is approximately 16 mils  with  ground spacing of 13 mil. This works out to  58.42  ohm as characteristic impedance of the transmission line ( instead of 50ohms) in  which case the return loss will be much degraded  than my PCB.

      2.  Incidentally for  the HMC 424ALP3E ( DSA)  evaluation board gerber file also fabricated using  16 mil and 13 mil dimensions for transmission line and ground spacing.

      3.  Can you please clarify  how the eval board has better performance.

    c)         Further clarifications:

      1.  Can I use two or more switches (HMC232As) to improve the on/off ratio.

      2.  Can I cascade two or more DSAs HMC 424 devices to get more attenuation like  90 dB or more.

    Will be glad to receive your comments as I am unable to proceed further.      Thanks and with best regards

Reply
  • Thank you very much for your reply. I am providing  further inputs for your review and feedbak.

    a)        My  PCB design is as follows.

    The PCB used is RO4350B with 10 mil thick and copper thickness of 1 ounce.

      1.  The PCB  transmission line width for 50 ohms characteristic impedance used is  21.35 mils.

      2.  Ground plane was extended on both sides of the transmission line at a distance of 20 mils for each plane.

    b)        HMC 232 ALP4E Evaluation board (sent by you)

      1.  The PCB transmission line width is approximately 16 mils  with  ground spacing of 13 mil. This works out to  58.42  ohm as characteristic impedance of the transmission line ( instead of 50ohms) in  which case the return loss will be much degraded  than my PCB.

      2.  Incidentally for  the HMC 424ALP3E ( DSA)  evaluation board gerber file also fabricated using  16 mil and 13 mil dimensions for transmission line and ground spacing.

      3.  Can you please clarify  how the eval board has better performance.

    c)         Further clarifications:

      1.  Can I use two or more switches (HMC232As) to improve the on/off ratio.

      2.  Can I cascade two or more DSAs HMC 424 devices to get more attenuation like  90 dB or more.

    Will be glad to receive your comments as I am unable to proceed further.      Thanks and with best regards

Children
No Data