• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Subscribers 53 subscribers
  • Views 1084 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.

CC1310: Front end RF design

aurelienr
Intellectual 370 points
Part Number: CC1310
Other Parts Discussed in Thread: CC1190

Hello,
I'm making a design with the CC1310 and I'm not sure of the RF front end stage. I hesitate between all-inclusive balun solution or full discrete solution.
The SWRA524A doc shows the results of the balun use : it seems that there is a sensitivity reduction of ~2-3dB (-108dBm @ 50kbps) compared to the device specs (-110dBm).
Do you confirm the characterisation published in datasheet has been performed using the discrete front end solution implemented in the ref design ?
I've downloaded the ref design guides, but I'm not sure I can make my boards with the same stackup. In your ref design you have 0.175mm width between layer 1 and 2 or 3 and 4. In my case I can have 0.39mm (that's already the stackup I use for other boards).
May that difference force me to change the values of the discrete element network ? Should I use a coefficient to determine the width of the inter-components tracks (I'm not speaking of the final 50 Ohms track to antenna) or is it negligible ?

What about single ended design ? I did not see any example. What may be the improvements from it ?
I just need my board to be compliant with both 868MHz and 915MHz bands.

Aurelien

  • 0
    TI__Guru**** 317180 points
    - The datasheet numbers are measured using the full discrete solution. As you see from the numbers using the IPC gives some dB less link budget compared to using the discrete solution.
    - Single ended: Tx: 3dB less output power, Rx: 1 dB less sensitivity.
    - I have used 0.4 mm for the CC1310 - CC1190 ref design without seeing any effect but I haven't measured directly on the CC1310 side. If you look at the reference design we provide we also have a 2 layer design with slightly different component values. Using the same values as the 0.175 mm version will probably give you slightly different numbers than the ref design but not significant change.
  • 0 in reply to TER
    Intellectual 370 points
    Hello,
    I noticed that the IPC report manual indicates that optimal perfs could be obtained using wirewound inductors, but your ref design uses multilayer inductors (LQG15HS family). Does that imply that using wirewound inductors may increase perfs announced in datasheet ?

    Another thing : the DC/DC inductor is specified in the BOM as Ir=1.1A, but the part selected (TDK MLZ2012N6R8LT000) have only Isat=0.11A and Ir=0.5A. What are general recommendations about this inductor parameters ? I'm not sure you have documented that (operating frequency, ripple current, serie resistance). Of course I will probably use your reference, but I just ask that for clarity...

    Aurelien
  • 0 in reply to aurelienr
    TI__Guru**** 317180 points
    The datasheet for the MLZ2012N6R8LT000 part gives the correct information. I suspect that the company that makes the library we use has done a typo in the description field.

    Wire bond inductors gives somewhat better performance than multilayer inductors due to higher Q but they also increase the radiation of harmonics due to the physical design.
  • 0 in reply to TER
    Intellectual 370 points

    You speak about harmonics radiated through the antenna or directly by the inductors ?
    In my case we will have a small EMC shield covering the CC1310 and all the front end RF.

  • 0 in reply to aurelienr
    TI__Guru**** 317180 points
    From the inductors. If you use a shield you can use LQW inductors without issues.
  • 0 in reply to TER
    Intellectual 370 points

    Thank you.
    Could you please clarify something : the characterization values in datasheet are based on the 4 layers board with multilayer inductors. For 50kbps, the sensitivity is given at -110dBm typ.

    In the IPC report, the sensitivity is about -108dBm. But the conclusion says that using an IPC device is similar to using discrete design with multilayer inductors. So why values are so different ?

    Aurelien

  • 0 in reply to TER
    Intellectual 370 points
    Another thing : I don't find any reference in the web relative to the Murata IPC reference (not even on their website). What's wrong ?
  • 0 in reply to aurelienr
    TI__Guru**** 317180 points
    Murata is slow updating their web page. In addition it could be difficult to find the part on the page. If you look at this thread: e2e.ti.com/.../1942804 it contains Murata contact information.