• State
  • Locked Locked
  • Replies 15 replies
  • Subscribers 54 subscribers
  • Views 3053 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.

Frequency deviation RF-PHY failure

Ara Warr
Prodigy 70 points
Other Parts Discussed in Thread: CC2541, CC2540

I am having a problem during certification of our product with Bluetooth module CC2541 and CC2540. We conducted LE RF-PHY conformance certification in test house, and all the modules came back with % of maximum frequency deviation <185KHz less than 99.9% (Result show below). According to test house this is a fail.

  We are closely following the sensor tag reference design. The firmware used is the HostTestApp from TI, which was part of BLE stack v1.40. Crystal is 10ppm stable and test temperature is 20 degC. Could you point to any reason for the failure.

    TX Freq |  Payload  | 

   (MHz)    |  Pattern   |  Test Criteria                               |  Value |  Unit  |  Limit      |Result |  Test Status              

--------------------------------------------------------------------------------------------------------------------

   2480    |  10101010 |  Max Freq Dev(Δƒ2) < 185 kHz  |  99.76  |  %    |  >= 99.9 |  FAIL   |  Already failed

The ChipID and CHVER are above

The power supply stability is captured below

AC coupled - 20mV/div

DC coupled 500mV/div

The schematic can be made available on request (not public). The schematic and layout very closely follow keyfob reference design.

We have thousands of units waiting for your answer. Our production is  really stalled because of the issue.

  • 0
    TI__Expert 6995 points

    Hi Ara,

    This parameter has been marginal on some designs. We've got some knobs to turn in the RF registers to improve this. I'll send you a patched BLE stack library for CC2541 that should resolve it.

    Peder

  • 0
    TI__Guru 57865 points

    Hi Ara,

     Could you please try attached Hex file (I assume you are using UART).

    Best Regards

    Joakimhttp://e2e.ti.com/cfs-file.ashx/__key/communityserver-discussions-components-files/538/6404.HostTestCC2541_5F00_fix.hex

  • 0 in reply to J Lindh
    Prodigy 70 points

    HI Joakim,

    I am using UART0. But I have to modify couple of GPIO states in initialization to get the unit to work. It would be much useful if you could send the library stack.

    We dont have ability to do a CDF/PDF of deviation less than 185kHZ for GFSK, which makes it difficult to do the test you asked in house. So these would have to be sent to test house to test. It is kind of expensive to keep trying. 

    Could you say what parameter was changed? If you made the deviation higher, I could try with older and newer firmwares with modulation analyser for a wider deviation before sending to the testhouse for confirmation.

    Cheers

  • 0 in reply to Ara Warr
    TI__Guru 57865 points

    Hi Ara,

    Please see attached files. It's a minor update in the Link Layer (register setting). Please try the library files and provide me with feedback.

    Best Regards

    Joakim

    5353.Libaries_fix.zip

  • 0 in reply to J Lindh
    Prodigy 70 points

    Hi Joakim,

    Could you pls. send me CC2540_BLE.lib as well. We are using CC2540 and CC2541. So we need both the libraries.

    Best Regards,

    Aravind

  • 0 in reply to Ara Warr
    Prodigy 190 points

    Hello Ara,

    did the patch help fix your issue? I am encountering similar problem in certfication process as well. We have a frequency deviation of 68KHz.

    It would be greatly appreciated if you could follow up on your post~

  • 0 in reply to Steve Chen1
    Prodigy 70 points

    Hi Steve,

    We haven't finished testing yet. But you are seeming to have a different issue that we do. We have the proper near 250KHz deviation most of the time(99.86%). Rest of the 0.14% it fails to be above 185 KHZ, which makes us fail. 

  • 0 in reply to Ara Warr
    TI__Expert 6995 points

    Hi Ara,

    The CC2540 has a different radio and is not marginal on this parameter. Have you seen failures on that device as well?

    Peder

  • 0 in reply to Peder
    Prodigy 70 points

    Hi Peder,

    Yes we were seeing failure in both CC2540 and CC2541

  • 0 in reply to Steve Chen1
    TI__Guru** 106705 points

    Steve Chen1 said:
    did the patch help fix your issue? I am encountering similar problem in certfication process as well. We have a frequency deviation of 68KHz.

    Hi Steve,


    Sounds like you have to adjust the load capacitors of the crystal. If the offset is positive you should add some capacitance to reduce the frequency.

    68kHz is anyway less than 30 ppm, so you should still be within spec.

    Cheers,

    Fredrik

  • 0 in reply to Ara Warr
    TI__Expert 6995 points

    Hi Ara,

    I've checked again and the settings used by CC2540 on the 1.4.0 should be very robust. Please pass me the full test report, schematics and layout offline and we'll see if we can spot something.

    Peder

  • 0 in reply to Peder
    Prodigy 70 points

    Hi Peder,

    I have the TestReport and rest of the package you asked me ready. How do I pass that to you?

    Best Regards,

    Aravind

  • 0 in reply to Ara Warr
    TI__Expert 6995 points

    Hi Ara,

    Thanks, I got it from TI FAE. We'll reveiw it and get back to you.

    Peder

  • 0 in reply to Fredrik K
    Prodigy 190 points

    Hi Fredrik,

    Thank you very much. I've changed the capacitance and it did reduce the deviation frequency to around 50KHz. It is ok for lab testing now. However, I do need help with increasing transmitted RF power.

    I've compare the transmitted power of my module against the TI USB dongle module. The maximum TX power output of my module is 0.41dB, where the TI USB dongle is up to 4dB (using same FW for both module with TX power adjusted to 4dB). I've followed closely to the reference design and made the PCB board with same layer stack (Antenna layer, GND, VCC, Bottom layer) with 1.6mm thickness, FR4 substrate.

    I am not a RF expert so I need some help in checking the PCB layout. Thank you very much!

    Cheers,

    Steve 

  • 0 in reply to Steve Chen1
    TI__Guru** 106705 points

    Hi Steve,

    If you post the RF-part of your layout I will take a look at it. Can you also describe in detail how you do the measurements?


    Cheers,

    Fredrik