Representing captured data in sine wave form

Hi Syed

Wavevision applies some formatting/parsing to the raw data transferred from the board. I think that is probably the issue. I'll track down the details of how the data is handled for that board and post back here.

Best regards,

Jim B

Thanks for your help with this Jim. Any updates on this ?

Jim, 

I believe I've narrowed down the issue and it looks like the data that I'm getting from the board is invalid. I was able to reproduce the issue with WaveVision as well and got similar waveforms to the ones that I was getting in Matlab. I was able to resolve the issue with WaveVision by using the Sync Req option before capturing the data but I'm not sure what action WaveVision performs when using that option so that I can implement that functionality in Matlab as well. Can you shed some light on it ? Thanks!

Hi Syed

The Sync Req item in the script sends some configuration and control information to the FPGA.

The relevant code from the script (spio5.5_xc5vsx95t_LM97600.dut.cpp) is:

/*
    -------------------------------------------------------------------------
    BEGIN: lm97600_Sync_GTP_Link

    Example call: ret = lm97600_Sync_GTP_Link(data);
    -------------------------------------------------------------------------
*/
Interface_Ret lm97600_Sync_GTP_Link(Adc08s5000_Data* data) {
    Interface_Ret ret; 
    WvWord temp_read;
 
    script_debug_out_pre("lm97600_Sync_GTP_Link");
    script_debug_out_endl();

    ret = lm97600_Check_GTP_Link(data);
    if (ret == INTERFACE_RET_OK) {
        // Don't perform sync if link is good
        return script_process_return(ret, "lm97600_Sync_GTP_Link: Good link status, don't perform sync");
    }

    script_debug_out_pre("lm97600_Sync_GTP_Link: performing Re-sync\n");
    // setup sync pulse width to a healthy amount: default = 0x80
    data->fpga_adc_sync_width = FPGA_ADC_SYNC_WIDTH_DEFAULT;
    script_request_mm_wr(0, FPGA_ADC_SYNC_WIDTH_ADDR, data->fpga_adc_sync_width);

    // Ensure that the following are set to 0. default=0x00
    //    Bit[7](R/W) - man_sync_req_ur (R/W for now, should be self clearing)
    //    Bit[6](R/W) - man_sync_en_ur Manual Sync Enable
    //    Bit[5](R/W) - static_sync (static sync level)
    //    Bit[4](R/W) - static_sync_en (static sync enable, overrides manual one shot)
    //    Bits [3:0] (R) read only, not affected by write.
    data->fpga_adc_sync_csr = FPGA_ADC_SYNC_CSR_DEFAULT;
    script_request_mm_wr(0, FPGA_ADC_SYNC_CSR_ADDR, data->fpga_adc_sync_csr);

    // Enable man_sync_en_ur and write to FPGA
    script_util_set_bits(
 1,
 FPGA_ADC_SYNC_CSR_MAN_SYNC_EN_UR_P,
 FPGA_ADC_SYNC_CSR_MAN_SYNC_EN_UR_S,
 &data->fpga_adc_sync_csr
        );   
    script_request_mm_wr(0, FPGA_ADC_SYNC_CSR_ADDR, data->fpga_adc_sync_csr);

    // Enable man_sync_req_ur and write to FPGA
    script_util_set_bits(
 1,
 FPGA_ADC_SYNC_CSR_MAN_SYNC_REQ_UR_P,
 FPGA_ADC_SYNC_CSR_MAN_SYNC_REQ_UR_S,
 &data->fpga_adc_sync_csr
        );   
    script_request_mm_wr(0, FPGA_ADC_SYNC_CSR_ADDR, data->fpga_adc_sync_csr);
   
    // Set all back to 0x00 to return to default state.
    data->fpga_adc_sync_csr = FPGA_ADC_SYNC_CSR_DEFAULT;
    script_request_mm_wr(0, FPGA_ADC_SYNC_CSR_ADDR, data->fpga_adc_sync_csr);   

    script_debug_out_pre("lm97600_Sync_GTP_Link: Re-sync done, checking link\n");
   
    ret = lm97600_Check_GTP_Link(data);
    if (ret != INTERFACE_RET_OK) {
        // Sync was unsuccessful
        return script_process_return(ret, "lm97600_Sync_GTP_Link: Re-sync was unsuccessful");
    }

    return script_process_return(
        INTERFACE_RET_OK,
        "lm97600_Sync_GTP_Link: Re-sync performed successfully and Link is okay"
        );
} 
/*
    -------------------------------------------------------------------------
    END: lm97600_Sync_GTP_Link
    -------------------------------------------------------------------------
*/

There is another section that follows that allows checking status as well.

Unfortunately that's about all of the help I am able to provide. The software engineer who wrote the C++ script is no longer at TI.

I hope this information is helpful.

Best regards,

Jim B

Thanks a lot for your help so far Jim! Could you tell me a bit on how WaveVision actually interacts with the C++ script ? I believe this is general for all DUTs and the way Wavevision is setup.