• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 553 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

ADS54J64EVM: ADS54J64EVM+TSW14J56EVM issues

Amy Luo
Expert 6895 points
Part Number: ADS54J64EVM
Other Parts Discussed in Thread: ADS54J64, ADS58J64EVM, ADS58J64

Hi team,

One of our customer's issues, I'm forwarding it below, could you please provide some troubleshooting suggestions

We now use the ADS54J64EVM+TSW14J56EVM two boards for sampling waveforms. Several questions have been raised, and I sincerely hope to receive a response.

1. We are now using the HSDCPro Automation DLL in the HSDC. Using the Get_Capture_Data_32bits function in Automation DLL C Example, the waveform digital data is obtained. And we want to go back to the analog signal we're sampling, how do we handle the digital signal? Is it possible to process the digital signal on a PC to get the original wave?

2.  We have Channel8 (8 channels) in the HSDC. The test found that the data from the four acquisition channels of the ADS was mapped to the eight channels of the HSDC. AIN, BIN, CIN, DIN for Chanel 1/2, Chanel 3/4,Chanel 5/6, Chanel 7/8 respectively; for example, the signal from AIN, corresponding to Chanel 1and Chanel 2, so what is the difference between Chanel 1 and Channel2? We can only observe that the magnitude of the two data is very different.

3. We see the process of the data in the configuration software of ADS. Is Channel 1, 2 in Question 2 above the I signal and Q signal for the diagram below?

On the ADS54J64EVM chip, "ADS54J64 Quad-Channel, 14-Bit, 1-GSPS, 2x Oversight, Analog-to-Digital Converter"Sbas841 document, See the description of DDC Mode0 with a little description below.

Is this modulation process for sampling data QPSK, or can you provide more detailed explanation?Is there a specific demodulation method for the data obtained through the Get_Capture_Data_32bits function? How do we restore the original signal of the sampled data

4. From the picture below, we can see that the sampling frequency of our adc is Fs = 1 GSPS. According to Nyquist sampling theorem, can we sample the signal up to 500 MHz?At the end of the picture we get Fs = 250MSPS at JESD204B, what is the maximum frequency of the sampled signal that can be recovered?

Best Regards,

Amy Luo

  • 0
    TI__Mastermind 23920 points

    Hi Amy,

    Is the customer using an ADS54J64 or an ADS58J64? The question mentions ADS54J64 but the software shows for ADS58J64EVM? Please clarify as these are different devices and we need to know which is used before being able to provide any support.

    Regards, Chase

  • 0 in reply to Chase W
    TI__Expert 6895 points

    Hi Chase,

    Thank you for your response. I have confirmed to the customer that he is using ADS54J64+TSW14J56.

  • 0 in reply to Amy Luo
    TI__Mastermind 23920 points

    Hi Amy,

    Thanks for checking for us. Regarding the original post, kindly find my comments below:

    1. To recover a signal from a decimated data converter, the NCO value must be known to digitally mix the sampled signal back up to the NCO value if this is what the customer is seeking to do.
    2. The converter is setup for a complex output, and therefore each I and Q will have its own channel in HSDC Pro. HSDC Pro Channel 1/2 refers to Channel A's I/Q data, HSDC Pro Channel 3/4 refers to Channel B's I/Q data, etc.
    3.  Ignoring the ADS58J64 software since it is incorrect, and only referring to the ADS54J64 datasheet image for Mode 0: Decimate-by-4 With IQ Outputs and fS / 4 Mixer , the data converter outputs a decimated-by-4 signal, meaning the converter samples at 1 GSPS, tosses away every 2 samples, down-converts this data by mixing with the LO frequency (in mode 0, this is set to fs/4), then decimates again to reduce output data rate by another factor of 2. The resulting data transmitted by the ADC is only a small window of the nyquist zone due to digital down-conversion performed by the ADC in this particular mode. The Get_Capture_32bits is just a function from our HSDC Pro Automation DLL to save the data which is present in HSDC Pro at the time of the function call. It has nothing to do with the type of data that the ADC is outputting.
    4. We cannot look at the picture below since this is a for a different device GUI. Please refer to figure 54 above, this shows the mixer as fs/4. Given that the mode is decimating-by-2, the integrated lowpass filter passband is only around 80% useful before filter rolloff becomes an issue (see Figure 50 rolling of after 200MHz (250MHz * 80%), as a result, the ADC will only down-converter signals which are within +- 200MHz of fs/4. After mixing, the converter decimates-by-2 again, resulting in another output data reduction by 2 and thus the bandwidth will drop down to +- 80MHz (this is because +- 200MHz /2 * 80% = +- 80MHz). This is all using the device in operating mode 0, where the mixer LO frequency is non-negotiable at +fs/4 or -fs/4. To gain flexibility in this LO frequency, using device operating mode 1 will enable the NCO as the mixer LO frequency. The NCO can be set to any 16-bit frequency representation up to the sample rate, fs (1GHz) using the equation below from section 7.4.1.1 Numerically Controlled Oscillators (NCOs) and Mixers

    Using the NCO instead of the fs/4 will enable a ~160MHz data window to pass at any NCO frequency. The analog input bandwidth for this device is specified at 1GHz for a 3dB cutoff, so setting the NCO to any value corresponding to a frequency between 80MHz and 920MHz should be fine.

    Regards, Chase