• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 12 subscribers
  • Views 368 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.

Original question:

AFE7950: Can you tell me about calibration?

AFE7920: RX DSA Calibration Issues

Frederik Steiger
Prodigy 80 points

Part Number: AFE7920

Hello 

We are using your newest C-library (CAFE 2p9) to calibrate the RX DSA. To be more precise, we are using its function "doRxDsaCalib(..)". As guide line we are following the procedure as descripted in the document "AFE79xx_RxDSACalibrationProcedure_Jun32019.pdf". Unfortunately, the execution failed and responded with Error_Code_1 = 8 and Error_Code_2 = either 1 or 2.

As a result of a deep dive into your python library we were able to decode these values as PHASE_ERR::DGPC_LARGE_PHASE_ERR_WRT_REF_RC (Error_Code_2=1)  and PHASE_ERR::DGPC_LARGE_PHASE_ERR_WRT_PREV_RC (Error_Code_2=2) 

Can you provide more information about this specific issue and also how to solve it?

Thanks and best regards

Frederik

 

  • 0
    TI__Intellectual 1150 points

    Hi Frederik, 

    Please ensure the instrument providing the AFE clock source is synced with 10MHz. If these are not in sync, that could be causing a phase error. Please let me know if you continue seeing unexpected behaviors.

    Additionally, the input tone should be at the NCO frequency + 11.52MHz for all RX channels being calibrated. The input tone should be around -7 dBFS on all RX channels being calibrated. Please check the capture power on all RX channels and set the signal generator power level such that the power on RXA is around -7 dBFS. I recommend using the capture power rather than the signal generator power, because the matching network or other elements in RX signal path can affect capture power. If you are on your own radio, you will still need the 10 MHz sync from the signal generator to the clocking chip on your radio. If you are on the AFE7920 EVM, you should feed the sync into the LMK_CLKIN pin.

    Thanks!

    Mirana

  • 0 in reply to Mirana Emigh
    Prodigy 80 points

    Hi Mirana

    Thanks for your input. I am wondering, why this important information is not documented in these application notes: "AFE79xx_RxDSACalibrationProcedure_Jun32019". This sheet does not mention the necessity of the 10MHz clock synchronization for the calibration process. 

    You mentioned this frequency offset of 11.52MHz. Right now it is hard to understand where this number is coming from. Would you mind to explain me, how this frequency offset is calculated?

    Since our HW-design does not contain an extra port for the LMK_CLKIN signal, we would like to use the AFE-chip's own TX pin for the calibration. 

    As before, the calibration failed, but his time with another error: COEFF_RANGE_ERR::DGPC_CORR_COEFF_TOO_LARGE?

    Can you provide more information about this specific issue and also how to solve it?

    Thanks in advance and best regards

    Frederik

  • 0
    Prodigy 80 points

    Part Number: AFE7920

    Hello 

    As already described in here: https://e2e.ti.com/support/rf-microwave-group/rf-microwave/f/rf-microwave-forum/1614828/afe7920-rx-dsa-calibration-issues

    We still face issues while using the c-library function "doRxDsaCalib(..)" for the calibration.

    We are using the AFE-chip's own TX pin for the calibration. We generate with the DAC an output signal of around 0dBm (constant test pattern) by using the c-function "dacJesdConstantTestPatternValue(...)" and feed it back into the RX channel. Considering the path attenuation of our HW design, the level is around -15dBm at the RX input of the AFE chip. 

    Before we tried the same procedure on the eval board and the calibration was successful, but with our own HW the calibration failed with the error: COEFF_RANGE_ERR::DGPC_CORR_COEFF_TOO_LARGE? 

    Can you provide more information about this specific issue and also how to solve it?

    Thans and regards

    Frederik

  • 0 in reply to Frederik Steiger
    TI__Guru* 81185 points

    Hi Frederik,

    I believe Mirana is reaching out to you offline to discuss further. Please let me know how this method would work. Thank you

    -Kang

  • 0 in reply to Kang Hsia
    TI__Intellectual 1150 points

    Hi Frederik, 

    We discussed via email, so I'm closing this thread. 

    Thank you!

    Mirana