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: TSW14J56EVM

Part Number: TSW14J56EVM

I got the following questions from another team, and we need to know if the ADS52J90 ADC in combination with the TSW14J56 eval board and either the DATACONVERTERPRO-SW or ADCPRO is a good fit?

  1. Measure continuously for a few minutes and calculating the average value. This must be able to handle high frequencies (up to half the sampling frequency) without aliasing.
  2. Measure continuously for a few minutes and calculating the highest peak measured at different low-pass filtered frequencies. E.g. 1 MHz, 100 kHz, 1 kHz. The low-pass filter can be implemented using a moving average over a number of samples.
  3. Do a one-shot measurement for a short period of time (1-2 seconds), saving every sample. Then showing an oscilloscope view of the captured waveform with the ability to zoom in and out.
  4. Same as 3 but exporting all samples for external processing.
  5. Is it possible to automate use case 1 and 2 from a Windows or Linux environment? We are using NI TestStand for this type of automation but we are flexible. For example: a set of command line programs, a DLL API, Python bindings, PowerShell scripts etc.
  6. Does the typical usage involve signal processing on the FPGA? Or is the FPGA only for capturing samples to the evaluation board DRAM and later processing on a PC using DATACONVERTERPRO-SW?
  7. Do you have any Linux software for this solution? Either a full GUI or enough low-level software in order to do use case 1 above on a headless Linux system.

 

Best Regards,

Bhaaeddin Alhomsi

  • Hi Bhaaeddin,

    Thanks for reaching us with your questions.

    Let me start with a brief introduction about the memory interface. The ADC and FPGA (TSW14J56) communication interface is fast but the FPGA and PC interface is slow, hence the TSW has 32Gb of memory to store the samples which then is transferred to the PC.

    For ADS52J90, the max sampling rate is 62.5Msps, which means that if I want to continuously capture the data and store all the samples, what will be the maximum duration after which I fill my memory depth. In 1 second I will fill

    62.5Msps x 16 Channel x 16 bit = 16 Gb

    So I will be able to continuously measure for 2s. After that it will be discrete measurement. 

    1. Measure continuously for a few minutes and calculating the average value. This must be able to handle high frequencies (up to half the sampling frequency) without aliasing

    2. Measure continuously for a few minutes and calculating the highest peak measured at different low-pass filtered frequencies. E.g. 1 MHz, 100 kHz, 1 kHz. The low-pass filter can be implemented using a moving average over a number of samples.

    Ans (1-2): Firstly, can you clarify what you mean by continously. I am assuming it means that you want to have all the samples. If that is the case then as I explained earlier, continuous measurement in minutes will not be supported.

    3. Do a one-shot measurement for a short period of time (1-2 seconds), saving every sample. Then showing an oscilloscope view of the captured waveform with the ability to zoom in and out.

    4. Same as 3 but exporting all samples for external processing.

    Ans (3-4): This should be possible. However exporting the samples, you have to do it manually.

    5. Is it possible to automate use case 1 and 2 from a Windows or Linux environment? We are using NI TestStand for this type of automation but we are flexible. For example: a set of command line programs, a DLL API, Python bindings, PowerShell scripts etc.

    6. Does the typical usage involve signal processing on the FPGA? Or is the FPGA only for capturing samples to the evaluation board DRAM and later processing on a PC using DATACONVERTERPRO-SW?

    7. Do you have any Linux software for this solution? Either a full GUI or enough low-level software in order to do use case 1 above on a headless Linux system.

    Ans (5,6,7) We rarely get queries regarding the automation. I will have to consult members of my team on this one. Let me get back on this as soon as possible.

    Thanks & regards,

    Abhishek

  • Hi Bhaaeddin,

    I checked with my team regarding your questions. Below if their feeback/suggestion

    5. Is it possible to automate use case 1 and 2 from a Windows or Linux environment? We are using NI TestStand for this type of automation but we are flexible. For example: a set of command line programs, a DLL API, Python bindings, PowerShell scripts etc.

    TI: We do not have any experience in the Linux environment. However we have released an automation example on MATLAB. You can find the relevant information about the same here: https://www.ti.com/tool/DATACONVERTERPRO-SW#downloads

    6. Does the typical usage involve signal processing on the FPGA? Or is the FPGA only for capturing samples to the evaluation board DRAM and later processing on a PC using DATACONVERTERPRO-SW?

    TI: We do not do any signal processing on the FPGA. We only use FPGA for capturing the samples. Also, we have used third party's support for making relevant software since we do not have much expertise in that domain.

    7. Do you have any Linux software for this solution? Either a full GUI or enough low-level software in order to do use case 1 above on a headless Linux system.

    TI: No, we don't have Linux based software solution.

  • Thanks, Abhishek for your reply,

    Can we reduce the sampling rate to 1MSPS? then the captured time will increase.

    because for some test cases the low sampling rate is enough for us.

    Best Regards,

    bhaaeddin

  • Hi Bhaaeddin,

    The minimum sampling rate of ADS52J90 device is 5MSPS.

    I am afraid you wont be able to run at 1MSPS.

    Thanks & regards,

    Abhishek

  • Hi Abhishek,
    In HSDC Pro, under the capture option, there is an option for continuous capture. what does that mean? can we use continuous capture for channels?

    Best Regards,

    Bhaaeddin

  • Hi Bhaaeddin,

    Yes, the continuous capture option will allow to capture continuously for channels. It is just  an extension of single capture, so the data will not be continuous between each captures.

    It is discrete in the sense that it creates new trigger event after capturing the said no of samples.

    Whenever data capture happens, the data will be written to files in below location-

    'C:\Users\Public\Documents\Texas Instruments\High Speed Data Converter Pro\HSDCPro Data'. You can find all channel data in ADCTemp0.bin to ADCTemp15.bin files present in this location.

    Though triggers are automatically created, the new capture will overwrite the files created in the previous capture. So at any time you'll only have latest captured data.

    Thanks & regards,

    Abhishek