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.

TSW14J56EVM: Captured data discontinuity

Part Number: TSW14J56EVM

Hi,

I send +94kHz CW, 491.52Msps, 0.5ms length (245760 samples) as below figure 1.

I connect TX and FB directly and capture as below figure 2, capture data length is set to 245760 samples and analysis length is set to 131072.

The spectrum of captured data is different from that of TX data. The data discontinuity is found in the code part of ADC tab, around 50000 sample and 180000 sample.

There is also discontinuity in the saved csv file. From the saved file, data has 65536 sample periodicity.

How can I solve this discontinuity?

figure 1. HSDC Pro DAC tab screen capture.

Figure 2. HSDC Pro ADC tab screen capture.

  • Tsutomu,

    From the DAC screenshot, we see no discontinuities on the waveform.  Are you using a custom INI file which was not included in HSDC Pro installer? If so, there could be some mistakes/typos in the Channel pattern or Bit packing pattern defined in the INI file which can cause this issue. Can you send both ADC and DAC INI files being used for this?

    Regards,

    Jim

  • Hi Jim,

    Not DAC but ADC screenshot. I marked the discontinuities with red circles.

    I'm using custom DAC/ADC INI files. afe77xx_dac_adc_ini.zip

    Would you kindly check them?

    BR,

    T. Hibino

  • Hi Hibino,

    This TSW14J56EVM transceiver firmware (TSW14J56REVD_BRAM_ADC_DAC_DDR_128K_XCVR_FIRMWARE) uses on-board DDR memory for DAC samples and FPGA internal BRAM memory for ADC capture data. On-board DDR supports a large memory depth. So, 245760  DAC samples that you are using is supported. However FPGA internal BRAM has a limited memory depth. So you cannot capture more than 8192 samples. If capture depth is increased beyond 8192, samples will just be repeated and you will see discontinuities like this.

    Regards,

    Vijay

  • Hi Vijay,

    I attached ADC capture data in CSV format. From this file, the discontinuities are found at 49152, 114688, and 180224. Sampled data from 1 to 65536 is completely the same as from 65537 to 131072.

    My understandings of ADC capture for long data are:

    • The maximum capture length per channel is 65536, instead of 8192 (your answer).
    • Sampled data from  65537 to 131072 is saved to capture memory address 1-65536. Old data is overwritten.
    • Sampled data from 131073 to 196608 is saved to capture memory address 1-65536 again. Old data is overwritten.
    • Sampled data from 196609 to 245760 is saved to capture memory address 1-49152. This causes discontinuity at 49152 samples.
    • When reading captured data, these 65536 samples are repeatedly used. So, discontinuity occurs at 49152+65536*n.

     Is there any way to capture long data?

    tsw14j56_adc_245760sa.zip

  • Hi Tsutomu,

    Your understanding is correct. Maximum sample depth is dependent on number of channels. If there are more channels, capture depth per channel is lesser. I said 8192 assuming a higher channel mode. In your case, you can double capture depth by modifying the ini file for single channel. 

    But we can't increase capture depth beyond that in transceiver mode. 

    Regards,

    Vijay

  • Hi Vijay,

    Thank you for your response. I'll try to reduce the channel number.