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: External pattern file format; Multiple channels

Part Number: TSW14J56EVM

Hello,

I'm using the TSW14J56EVM connected to a custom board with 2 DAC37J84 DACs.  I've been able to communicate with all 8 channels, but I would like to be able to load custom waveforms into each for system prototyping.

First, what is the format that the FPGA is expecting from the HSDC GUI?  I'll create the waveform in MatLab, but not sure what the data should contain - I'll be writing a complex waveform.

Second, is it possible to write all 8 channels from the same file?  If so, what should that data look like?  If not, can I write to more than one? 

Thanks,


Justin

  • Hi Justin,

    Please see attached for the format of a 4 channel pattern file sent through HSDC Pro /TSW14J56EVM. I believe that what you are asking is achievable, but this may require two files. I will research this some more.

    DAC38J84 Test_368.64MSPS_100MHzSignal.csv.zip

    How are you confirming communications with the DACs and TSW14J56EVM? Have you successfully sent a pattern to any of the DAC channels? What LMFS settings are you using?

    Here is a TI reference design that was done using two DAC3XJ8X and two TSW14J56EVM s. It's not exactly the same as what you have described, but it may be helpful for your design.

    http://www.ti.com/tool/TIDA-00996/

    Best Regards,

    Dan

  • Thanks, Dan.

    I've seen the *.tsw test files available (and have sent them to my DACs), my question was more to what those numbers represent for complex data. Would it be "Real, Imag" four times for the four channels, for a total of 8 columns?

    I've successfully sent frequencies from the HSDC GUI to each of the 8 channels and verified the frequency with an analyzer. I BELIEVE I've sent one of the *.tsw test files, as well... and a *.csv, as well.

    I'm using an LMFS setting of 4421, which actually brings me to another question... how to communicate with all 8 channels using one INI file? Thus far, I've used an INI file for lane 0-3 and a second file for lane 4-7 using the following lines in the ini:
    "Lane Mapping=lane0:0,lane1:1,lane2:2,lane3:3"
    or
    "Lane Mapping=lane0:4,lane1:5,lane2:6,lane3:7"

    What variables would I need to change in the INI to address all 8 lanes?

    Thanks for the ref design. We'll be linking 6 of these modules together and time alignment is one of the issues on the table. It'll help a ton.
  • Hi Justin,

    I believe that this can be accomplished, but I have not personally seen this in action. I think an issue that may arise (at least with using the TSW14J56EVM) is that both of the DACs will need to be synchronized simultaneously by the FPGA (using the SYNC signal).

    As far as the ini file goes, I think you are on the right track, but I am unable to confirm. I believe you will need to use an ini file that utilizes 8 lanes (doubling most of your current ini file setting in order to support all lanes).

    Best Regards,

    Dan
  • Hi Dan,

    I have the Sync lines for each DAC connected to the FPGA - the second connection was made to LA_33_P/N_A (pin G36/37 on the FMC) so that either IC can request a sync.  I'll be modifying the TSW source to utilize some additional control lines and that's one of the changes on my list.  I'd like to connect the second Sync line to TX_ALT_SYNC_P/N, but I don't think I can get the lines out without tearing up my comm lines.  There will likely be some timing differences, as the pins are on different banks, but I'm hoping I can compensate for that within the FPGA.  If you think that will be problematic, please let me know and I can try a bit harder to gain access to the ALT_SYNC pins - I found some clock issues on Friday and am in the process of moving some lines around.

    For the INI file, are there any guidelines to the variables and what they mean?  I didn't want to just start changing things and end up breaking it, but I would like to have a better understanding of which variables need to be changed to utilize the full 8 channels.  If I have to use two files to load the waveforms, that would be doable, but changing INI files wouldn't be feasible.

    Thanks for your help

    Justin

  • Justin,

    Dan is on vacation this week. Yes, it is possible to use a single TSW14J56EVM and synchronize two DAC37J84 devices, each operating in 4 lane mode (LMF 442 in this case). The SYNC signals from the two DACs need to be ANDed as a single SYNC signal to achieve synchronization, which I think you are doing inside your FPGA.  If one of the DAC devices sends a SYNC request, the JESD link will go down for both, and it will be re-established the link for both synchronously.

     

    To test the TSW14J56EVM with two DAC37J84 DAC's with a single ini file, you can take the 4 lane ini file and modify the # of lanes, lane mapping, number of channels and format pattern to support 8 lanes. Attached is an example ini file with the changes made for you. With the attached ini file, the firmware gets configured for 8 lane mode and works as a single JESD link with 8 lanes, even though there are two DAC devices connected.  You should then be able to use a single pattern file with data for all 8 channels, real & complex data, using a total of 8 columns (data format similar to LMF 442). I have also attached a document that explains the parameters of the ini files.

     

    Regards,

     

    Jim

    https://e2e.ti.com/cfs-file/__key/communityserver-discussions-components-files/73/DAC3XJ84_5F00_2X_5F00_LMF_5F00_442.iniTSW14J56 DAC INI File.docx