FCC/CE RF Certification Fail

Hi,

We are using the Nordic Semi nRF51822-QFAA-T to do BLE project, but when we test CE/FCC RF test(Conducted / Radiated), there has a lot of noise around 2.4G.

BTW, I used Nordic Studio DTM to test it.

Nordic Studio DTM setting

Test result shows below:

image description

backup image

schematic

PCB layout

Could anyone help me please?

Elliott

  • When doing the measurements, could you turn on "peak hold" in the spectrum analyzer? What you are seeing is most likely transient noise from the PA being turned on/off.

    Update 17.06.16: The PA is automatically turned on when the radio is transmitting, and off when it is not transmitting anymore. It cannot be controlled by a user. During the ramp-up of the PA there will be some transients. However, they will only be there for a very short amount of time, and before modulation has been turned on. By using "peak hold", you can see the envelope over time. The spectrum will look a little wider than what it actually is because of the transients. However, since the transients occur before modulation has been turned on, it is ok. Reducing the resolution bandwidth in the spectrum analyzer (if possible) will make the transients less visible.

    Baluns and nRF51822:

    BAL-NRF01D3 was made for silicon revision 1 of nRF51822, and does not work very well with later revisions because it doesn't damp RX LO sufficiently. It is possible that the output power is lower as well. For the latest versions of nRF51822 QFN there are the following options (see nRF51822 webpage --> development tools and software):

    • Johanson Technology 2450BM14E0003

    • ST BALF-NRF01E3

    • ST BALF-NRF01D3 was the recommended balun for silicon revision 2 of nRF51822. Whether or not it can be used with the latest version of nRF51822 depends on which version of nRF51822 you are going to use. When using silicon revision 3, BALF-NRF01D3 can only be used with nRF51822 QFN 256 kB flash and 16 kB RAM. With the other versions, it will give lower output power and and smaller margins for passing FCC/ETSI.

    Changing balun will most likely not change the transients, they are there due to the PA ramp-up, but they should not affect the FCC certification. Are you facing any problems with FCC certification due to the transients?

    Note: I would recommend you to change to one of the recommended baluns.

  • Hi Kristin

    I have to test it in RF lab, but the signal waveform would be the same when I turn on the peak hold/Max Hold, right?

    How can I control the PA turn on/off?

    Thanks,

    Elliott

  • The PA is automatically turned on when the radio is transmitting, and off when it is not transmitting anymore. It cannot be controlled by a user.

    During the ramp-up of the PA there will be some transients. However, they will only be there for a very short amount of time, and before modulation has been turned on. By using "peak hold", you can see the envelope over time. The spectrum will look a little wider than what it actually is because of the transients. However, since the transients occur before modulation has been turned on, it is ok. Reducing the resolution bandwidth in the spectrum analyzer (if possible) will make the transients less visible.

  • Hi Kristin

    I found a new part of match balun, BALF-NRF01D3 ( not BAL-NRF01D3), do you know which part is suitable for nRF51822 ? There have a lots difference in RF performance, for example, BALF-NRF01D3 Return loss = 17dB, but BAL-NRF01D3 Return loss=10dB.

    Could I solve this problem by changing match balun?

    Elliott

  • Hi Elliot,

    I have updated my answer regarding the baluns, please have a look at it.

    Kristin

Related