• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 16 subscribers
  • Views 1579 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BLE Chip CC2540 : Radiation Emission Failure, 2nd Harmonics 4.8GHz @ RX Mode

WL Chan
Prodigy 40 points
Other Parts Discussed in Thread: CC2540

Hello everyone. We am facing a severe problem here.

There is radiation emission at 4.8GHz found on CC2540 RX mode when proceed the compliance testing for EN300 328.

is there any expert can share us the theory behind on having significant 4.8GHz emission on RX mode, which is the second harmonics of the fundamental?

We do compare our boards to TI CC2540_MiniDK (Keyfob) board and we found out

CC2540_MiniDK --> Passed with ~6dB margin under the EN300 328 limit

Our boards --> Pass with 1~2dB margin under the EN300 328 limit only., which is considered marginal pass result. This is definitely not our/customer preferable test outcome.

Regarding the test setup:

All DUT boards we tested are only the PCBA level and we operate them with 2 x AA battery, configured the DUT via TI SmartRF Studio 7 (ver 1.7.1).

We hope to get some advices & hints from professional point of views. Thanks.

  • 0
    TI__Genius 12545 points

    WL Chan,

    We have seen this issue in some customer boards. It usually a function of the layout of the board. It is essential that you follow proper layout guidelines to get the best possible results. The fact that you pass, shows that you are close.

    The radiation emission during RX is the LO leakage. This is usually kept in check by having solid ground plane under the CC2540 and RF path and also use plenty of vias to form a solid ground plane between the different ground planes. Also, the decoupling caps, the placement of which greatly helps reduce these effects. Is your board a 2 layer board or a 4 layer board?

    The app note below is a good resource that talks about layout reviewing and techniques to get best results of your RF pcb

    http://www.ti.com/litv/pdf/swra367

     

  • 0 in reply to Chatto
    Prodigy 40 points

    Thanks Chatto, we do refer to the application note in the design. Perhaps our layout is still not yet optimum.

    We design 4-layers boards with the de-coupling caps value & placement are followed closely to TI CC2540_MiniDK.

    Yes, we do have solid ground. Our PCB stack up is signal/GND -> Solid GND -> Solid VDD -> Signal/GND. And there are lots of stitching via along the RF path, chip BALUN area & RF matching network. We are not sure how many via are needed for the best optimization. As many as possible?

    Is the radiation emission is from the antenna path via antenna or from the board itself?

    We sniffed them with near field probe and found the significant emission is around the pin 24 to pin 30 (RF_N, RF_P & AVDD1/2/3/4).

    Any improvement need to be conducted on the crystal circuitry as well?

    Thanks for great help.

  • 0 in reply to WL Chan
    TI__Genius 12545 points

    WL Chan,

    As I said previously, your reference design is close and by the description you gave me, you have adhered to the CC2540 reference design.

    The radiation you see is from the board itself. The observation you've made is correct. The supplies AVDD1/2/3/4 supply the different modules inside in the radio which is operating at RF frequencies and so it would have emissions around those pins.

    Do you have the different layers stitched with vias? I'm talking around the board that is not close to the RF path. We've seen that small unstitched ground planes at RF frequencies become radiators.  SO it's nice to have as many vias as possible. I don't have a number for you.

    For the crystal circuitry, use the same rules. Solid ground plane under the lines from the crystal to the chip. Also, no signal lines/supply cutting under these lines. Keep the crystal as close to the chip as possible and on the same layer as the chip, if possible. 

  • 0 in reply to Chatto
    Prodigy 40 points

    Hi Chatto,

    We do have GND via on other layer/portion not close to RF path.

    Do you mind if we send the board layout to your email address for review? Thanks.

  • 0 in reply to Chatto
    Prodigy 40 points

    Hi Chatto, We do have GND via on other layer/portion not close to RF path. Do you mind if we send the board layout to your email address for review? Thanks.

  • 0 in reply to WL Chan
    TI__Genius 12545 points

    WL Chan,

    It is important to have ground vias through out ground plane. This enables the reduction of ground planes acting as antenna at RF frequencies.

    I already have your board and am sending you more comments on them