• State Resolved
  • Locked Locked
  • Replies 18 replies
  • Answers 4 answers
  • Subscribers 10 subscribers
  • Views 663 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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:

AFE7950EVM: About HSDC Pro parameter change

AFE7950EVM: About HSDC Pro parameter change

Kazuto Odajima
Prodigy 80 points
Part Number: AFE7950EVM

How does the evaluation of the AFE950EVM change the value of the Digital Step Attenuator?
For TX, is it the Scaling Factor?
What should RX do?

  • 0
    TI__Guru* 91947 points

    Kazuto,

    Someone will get back to you soon regarding your question.

    Regards,

    Jim

  • +1
    TI__Genius 11980 points

    Hi Kazuto,

    To control the AFE's DSA, digital step attenuator, you can use the following commands in Latte.

    AFE.DSA.setRxDsa(Channel,Value)

    AFE.DSA.setTxDsa(Channel,Value)

    Regards,

    David Chaparro

  • 0 in reply to David Chaparro
    Prodigy 80 points

    Hi David,

    I was able to set the DSA using the method you taught me, and the TX was generally good, but the RX settings and measurement results were very different.

    what is the reason?

    The frequency is 1090MHz and the input level is 4dBm.

    DSA changed the range from 0 to 25dB in 1dB steps.

    Regards,

    Kazuto Odajima.

  • 0 in reply to Kazuto Odajima
    Prodigy 80 points

    Accuracy was also bad at 9.5 GHz.

  • 0 in reply to Kazuto Odajima
    TI__Genius 11980 points

    Hi Kazuto,

    The issue that you are facing is that the DSAs are not calibrated. To calibrate the DSAs you can use functions in Latte and I am working on creating a guide for this. I should be able to share this guide with you by Thursday. 

    Regards,

    David Chaparro 

  • 0 in reply to David Chaparro
    TI__Genius 11980 points

    Hi Kazuto,

    The attached PPT has the steps that can be followed for the DSA calibration.

    DSA_Calibration.pptx

    Regards,

    David Chaparro 

  • 0 in reply to David Chaparro
    Prodigy 80 points

    Hi David,

    Thank you for the PPT.

    I tried to calibrate using the internal TX method according to the PPT example, but the Log display is a little different.

    After this, I measured the accuracy of DSA again, but the result was still bad.

    What is the cause?

    Regards,

    Kazuto Odajima.

  • 0 in reply to Kazuto Odajima
    TI__Genius 11980 points

    Hi Kazuto,

    Is it still off by the same amount as before the calibration? I will run a sweep on my side to verify the DSA at 1GHz.

    Regards,

    David Chaparro

  • 0 in reply to David Chaparro
    Prodigy 80 points

    Hi David,

    Yes, they are off by the same amount.

    Does it have something to do with the lack of the last three lines in the log compared to the guide?

    also I calibrated the TX DSA according to the guide, but I got an error on TX2 and TX4.

    Regards,

    Kazuto Odajima.

  • 0 in reply to Kazuto Odajima
    TI__Genius 11980 points

    Hi Kazuto,

    Just to confirm after doing the calibration are you programming the AFE again or doing any type of reset? This calibration must be done after every AFE bringup. 

    Also, the error that you see is not expected. I will try and recreate this on my side. 

    Regards,

    David Chaparro

  • 0 in reply to David Chaparro
    Prodigy 80 points

    Hi David,

    After configuring the DSA, I am programming with Latte to change the sampling frequency and NCO frequency.

    I would like to resolve this issue within this week.

    Thank you very much.

    Regards,

    Kazuto Odajima

  • 0 in reply to Kazuto Odajima
    TI__Genius 11980 points

    Hi Kazuto,

    As mentioned in my previous reply the issue is that the DSA should be calibrated after the AFE is programmed. If you calibrate the DSA first then re-program the AFE then the DSA is no longer calibrated and needs to be re-calibrated. If you are attempting to test multiple configurations then what you have to do is calibrate the DSA after configuring the AFE each time. 

    One thing that may be useful for this case is that when calibrating the DSA Latte exports the DSA calibration data to a document that can be reloaded to the AFE later. This means that after calibrating the DSA once for each of your configurations you can save the calibration data. This way after programming the AFE into your specific configurations you can load the DSA calibration data instead of having to calibrate the DSA again. 

    By default the AFE79xx GUI stores the DSA calibration packet at the following location: C:\Users\<User Name>\Documents\Texas Instruments\Afe79xxLatte\lib. After DSA calibration the file names should be updated in order to avoid overriding the previous calibration data. 

    To read the packet and load it you can use the following commands in Latte:

    byteList=AFE.convertHexFileToByteList("C:\\Users\\<>\\Documents\\Texas Instruments\\Afe79xxLatte\\lib\\RxDSACalPkt.hex")
AFE.applyRxDsaCalibPacketFromByteList(byteList)

    Regards,

    David Chaparro

  • 0 in reply to David Chaparro
    Prodigy 80 points

    Hi David,

    Alright, I'll try.

    Is it okay to change the sampling rate and scaling factor when calibrating with a different configuration?

    Also, Why did TX2 and TX4 fail during TX calibration?

    I didn't do anything special except change the cables.

    Regards,

    Kazuto Odajima.

  • 0 in reply to Kazuto Odajima
    TI__Genius 11980 points

    Hi Kazuto,

    Are you attempting to use the AFE in multiple configurations? Is this why you are updating the sampling rate and decimation/interpolation? The DSA should be calibrated for each different configuration that you will be using. Each configuration can have different sampling rates and decimation/interpolation factors. Can you share the procedure you are following when you program the AFE and calibrate the DSA?

    I am not sure on the exact reason you see the error on TxB and TxD. I would suggest trying a different cable for the loopback connection and see if the issues occurs again.

    One thing that may be of interest to you is that in the AFE79xx secure folder we have two documents that cover the Rx and Tx DSA Calibration Procedure. They are both in the Application Notes section of the AFE79xx secure folder.

    Regards,

    David Chaparro

  • 0 in reply to David Chaparro
    Prodigy 80 points

    Hi David,

    Following the guide I received in the previous reply, I configured with the configuration of "S1_OnboardClk_RX_250M_TX_FB_500M" and loaded the packet, and the RX configuration was successful.

    However, when I change the sampling rate and NCO frequency, I get an error during configuration even with the same procedure.

    Are there any restrictions?

    The error contents are 'INPUT SIG ERR', 'DGPC SIG WITH HIGH SPUR'. Looking at the waveform, there are no large spurs around the tone.

    Specifically, based on "S3_OnboardClk_RX_TX_500M_FB_disabled", 'setupParams.fpgaRefClk' was changed to 368.64 and each NCO was changed to 9000.

    I checked the secure folder document you introduced, but I could not solve the problem.

    Regards,

    Kazuto Odajima.

  • 0 in reply to Kazuto Odajima
    TI__Genius 11980 points

    Hi Kazuto,

    I am facing a similar issue and I am debugging this. The issue may be that with the NCO frequency of 9GHz there is a nyquist boundary in-band in the ADC. The ADC is sampling at 2949.12MHz so there is a nyquist boundary at 8847.37MHz. If you try the DSA calibration with the NCO set to a different frequency, like 9.5GHz, do you get the same error?

    Regards,

    David Chaparro

  • 0 in reply to David Chaparro
    Prodigy 80 points

    Hi David,

    When I set the NCO to 9.5GHz without changing the sampling frequency, I was able to calibrate.

    I also tried several other NCO patterns, but there were times when calibration was possible and times when it was not possible, so as you said, Nyquist may be related.

    Regards,

    Kazuto Odajima.

  • 0 in reply to Kazuto Odajima
    TI__Genius 11980 points

    Hi Kazuto,

    For the times that it did not calibrate are you close to a Nqyuist boundary. If so then what you will have to do is frequency plan such that the Nyquist boundary is not within the band.

    Regards,

    David Chaparro