• State Not Answered
  • Replies 11 replies
  • Subscribers 65 subscribers
  • Views 2381 views
  • Users 0 members are here
Attachments (2) Download All
Nordic Case Info
  • Case ID: 296019
Options

nRF52840 matching network for pin 31(ANT), matching circuit needed for regulatory requirements, EMC

bjornshnrf

Hi,

We are close to going into a certification process but before we start we need to be sure we are not missing out on your recommendations with regards to EMC performance.


Is the radio matcing network needed for other purposes than matching towards 50ohm? Is it needed for regulatory approval? 


Can you share the impedance at pin 31 (ant port) of the nRF52840?



Thanks,

Regards

Bjørn

Parents
  • 0

    Hi, 

    Thank you for the answer and clarifications.

    I did a simple test with a simple circuit and VNA. I made two circuits - both equal to your two alternative matching circuits. I will only give feedback on the 3 component pi network.

    When assuming 50ohm looking back into the nRF52840 through the matching network (as you listed in the post) - we can do the measurement from the other side of the matching network to get an idea about the optimum load for the radio and ANT port. This impedance is  the optimum load impedance. 

    I then simulated this by making a source with this impedance in serie with the matching circuit and observed optimum transfer of power (low S11) through the matching circuit to the 50ohm matched antenna. 

    Anyways, good feedback on the need for this network for regulatory certification and that the ANT impedance is not 50 ohm.

    Thanks

    Regards

  • 0 in reply to bjornshnrf

    And note that we tune the matching network for the radio, not with the impedance but by adjusting the components values to optimize the output power and get low enough harmonics. So when tuning the radio we connect it a signal analyzer. 

    Best regards,
    Kaja

  • 0 in reply to Kaja

    Hi,

    Can you please specify what spectral components you improved by introducing the matching network? 

    It would be very beneficial to have both the frequencies and the power levels of these components.

    Thanks,

    Regards

    Bjørn 

  • 0 in reply to bjornshnrf

    Hi Bjørn,

    here is a description of the two first components in the matching network and the track grounding the parallel capacitor. 


    You can also have a look at our RF performance test guide.


    Best regards,
    Kaja

  • 0 in reply to bjornshnrf

    Hi,

    Thank you for your answer.

    Can you tell what specific spurious you was challenged on that triggered the design of current recommended nRF matching network?

    The suggestion you gave hint that you needed a notch at 6GHz. 

    I can read up om the standard to check if our design is above limits at various frequencies - I will be able to find any spurious with the spectrum analyzer and the GTEM cell i use for desktop RF measurement (for those below roughly 6GHz).

    Regards

    Bjørn

  • 0 in reply to bjornshnrf
    bjornshnrf said:
    Can you tell what specific spurious you was challenged on that triggered the design of current recommended nRF matching network?

    2nd and 3rd harmonic emissions, as listed in the description below: resonant freq. between 2nd and 3rd harmonic. 

    Best regards,
    Kaja 

  • 0 in reply to Kaja

    Hi,

    We measured the unit in a GTEM cell that should be suitable for up to 8GHz. With 0dBm output power, we could not find any 2nd harmonic and a weak 3rd harmonic with level below -70dBm. (added 13/10: this is per GTEM port and needs to be adjusted . reading up on the EN 610000-4-20 standard for the 250 design, one should ideally measure EMC with 3 vectors - I only used one and this gives not accurate results! )

    We made the layout prior to the final recommendation on the matching network and did assume 50ohm into ANT (back in the days). The potted antenna has 50ohm at 2.45GHz. I believe we could gain some dB on power out with the recommended 0dBm output power matching network. 

    Regards

    Bjørn 

Reply
  • 0 in reply to Kaja

    Hi,

    We measured the unit in a GTEM cell that should be suitable for up to 8GHz. With 0dBm output power, we could not find any 2nd harmonic and a weak 3rd harmonic with level below -70dBm. (added 13/10: this is per GTEM port and needs to be adjusted . reading up on the EN 610000-4-20 standard for the 250 design, one should ideally measure EMC with 3 vectors - I only used one and this gives not accurate results! )

    We made the layout prior to the final recommendation on the matching network and did assume 50ohm into ANT (back in the days). The potted antenna has 50ohm at 2.45GHz. I believe we could gain some dB on power out with the recommended 0dBm output power matching network. 

    Regards

    Bjørn 

Children
Related