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.

AFE7799: JESD204B Link Fail

Part Number: AFE7799
Other Parts Discussed in Thread: AFE7769

Hi expert,

Customer used AFE7799 and when they did 100 times reboot test, there is chance to see JESD204B link failed. Some questions and debug process would like to consult with you,

  1. How to check the AFE7799 JESD 204B RX status by SPI command?
  2. How to reset the AFE7799 JESD 204B TX by SPI command?
  3. How to reset the AFE7799 JESD 204B RX by SPI command?
  4. is it a right approach and how to use Tx signal loopback to Rx/Fb then to let FPGA check the Rx/Fb path signal quality? 

Regards,

Allan

  • 1. getJesdRxAlarms, getJesdRxLinkStatus are the CAPIs that customer can use to check the JESD Rx status.  In addition, to verify the signal quality in the serdes physical layer, customer can use the serdes eye readout commands in the link below to read the eye margin.  If the signal quality is good, the readout values should be > 600.  For example, in the screenshot below, the readout values for register 4661 and 4660 are 0x57 and 0x60 in hex respectively.  Together, the eye margin value becomes 0x5760. The eye margin value is written in 12 bits, so the first #, 5, is ignored.  Converting 760 into decimal becomes 1888.  1888 is greater than 600, this means the signal quality is good in this particular case.   

    serdes eye readout commands:

    /cfs-file/__key/communityserver-discussions-components-files/220/serdes-eye-read-commands

    2. clearJesdTxAlarms CAPI can be used to reset the AFE7769 JESD204B Tx  

    3. clearJesdRxAlarms CAPI can be used to reset the AFE7769 JESD204B Rx  

    4. There can be multiple reasons why the link failure happened, it could be because of the FPGA is not sending the signal, or AFE somehow not receiving the signal or something happened in the serdes physical layer, which is the layer responsible for sending/receiving data from FPGA to AFE or vice versa.  The fact that customer did 100 times reboot tests and there is a chance the link failed indicates something must have changed when the link failure happened.  Without knowing more information about the link failure or test setup, it is difficult for TI to pinpoint what exactly caused the failure to happen; more investigations are needed before we can make a conclusion on the root cause of this problem.

  • Kevin,

    Customer applied below API into their RF init procedure:

    getJesdRxAlarms

    getJesdRxLinkStatus

     

    Referred to the C file, it seems lack of definition of ftdi_writeReg, ftdi_readReg when compile. 

    Could you please provide suggestions?

    Thanks,

    Allan

  • How does alphanetwork implement other AFE7769 CAPIs in its radio? For the two CAPIs, getJesdRxAlarms, getJesdRxLinkStatus, you set the parameter inside the function to 0 and the function should work without any issues.  I already verified this on our AFE7769 EVM.  To move forward with this issue, i have forwarded the screenshot to the software engineer responsible maintaining the AFE7769 CAPIs, I will let you guys know once i hear back from him.

  • Hi Kevin and experts,

    I simply refer to AFE77xxKeyCFunctions_v1p12.c as only file we have, and put getJesdRxAlarms, getJesdRxLinkStatus and all their related function to our RF Initialization process.

    I need files or complete code for missing definition of ftdi_writeReg and ftdi_readReg to finish code compilation.

    Thanks

  • ok, i will work with our engineer who is responsible for maintaining the AFE7769 API and get back with Alpha next week.