Post Go back to editing

HMC313 : Test data in the datasheet and Evaluation board


Was the test data of HMC313 datasheet measured by HMC313 Eva board which is shown page 8 in the datasheet? If yes, were LCR parts used the table in the page 8 of the datasheet?

As the Eva. board of HMC313, the pad of J1 and J2 are larger than other Eva. board, e.g. HMC311ST89. Could you advise me the purpose of the large size pads?

And, should we need to do the matching on a PCB for using HMC313? I received some inquiries from our customer.

Best regards,


in case of Eva. HMC313 board;

in case of HMC311ST89 Eva. board

  • Hi,

    Yes, Datahsheet measurements are taken with the EVB given on the DS. `List of Materials for EVB` table has the component values on the EVB.

    Purpose of the pads on the PCB, where you marked with red, is to compensate the effects of the RF connector used. Main objective is to have a controlled and continuos 50 ohm from.

    You do not need them on your own PCB, you may need to change their dimensions if you will use different PCB stackup of RF connector. Which is the case for HMC311ST89 EVB.



  • Hi

    Thank you for your answer.

    I'like to confirm R1 value just in case,

    Electrical perfomance was tested at Vcc=5V according to the datasheet. In this case, R1 is used 0 ohm accoring to the table on page 7 in the datasheet, right?

    Our customer couldn't get good perfomance. So they worry about the LCR value on the Eva board and necessity of a maching circuit for their board. Should we use matching circuit when we get good perfomance?

    Thanks & Regards,


  • Hi,

    Yes 0 ohm is suitable for 5 Vs.

    What is the freq range, and how much degradation is observed on which parameters?



  • Hi Kagan,

    Could you please look at the attcement. Vcc=5V, Rbias=0 ohm, RF Choke =22mH,  C3=2.2uF + 1000pF and C1 & C2=0.1uF. Do you have any ideas about the cause of this poor frequency response?

    Best regards,


  • Hi AkiraO,

    The heavy roll-off vs. frequency shown in the plot at right below is likely caused by the RF choke on the EVB.

    That choke is Coilcraft part # 0805CS-220XJLB, with S-parameters that can be found here:

    The S-parameters (plotted at left) show its SRF to occur at 3.5 GHz...the exact frequency at which the gain starts to roll-off on the plot at right.

    The following is noted below the Electrical Specifications table of the HMC313 data sheet:

    I looked back through our records and found that the HMC313 was indeed validated and characterized on a very generic SOT26 board (not the one shown on the data sheet) that did not have an on-board RF choke. Thus the use of the broadband bias tee.

    Considering that we were able to validate the part using such a generic PCB with an external broadband bias tee, I would bet that the part is not very prone to oscillation and thus the 22 nH value is not that important. On your own board you could perhaps use a choke of larger inductance in a smaller package, such as the one used on our HMC8410LP2FE EVB. We use that choke across the range of 0.01-10 GHz, though you could select differently based upon your frequencies of interest.

    FYI, we have an app note regarding selection of broadband biasing of amplifiers. It’s not the best, but it provides some basic conceptual instruction plus reference to some L & C manufacturers for component selection:



  • Hi,

    Thank you for your reply and informing the application note. They were very helpful.

    Thanks & Regards,