To average noise in TES sofware as it is possible to do in IIO Oscilloscope, (for sensitivity and dynamic range measure)

Hello.

I have ZC706 (rev 1.2) board with adrv9002 board.

We need to measure the characteristics of the receiver in ADRV9002. In order to more accurately measure the signal-to-noise ratio on the spectrogram, we need to average the noise track.

I tried to run an image for ZC706 from this topic - ez.analog.com/.../zc706-adrv9002-iio-oscilloscope-cannot-find-zynq-zc706-adv7511-adrv9002 to start the iio oscilloscope, but without success.
Questions:

1. Is it possible to average noise on RX spectrum in TES software as it is possible to do in IIO Oscilloscope software (We using TES 0.18.1)?

2. Is there an instruction on how to properly measure the sensitivity and dynamic range of the receiver ADRV9002?

  • Hello AntonBro,

    I'll do my best to address each question you have:

    1. Within TES, no it is not possible. However you can save Rx data into files (we support multiple file formats for the sake of supporting different analysis tools) and perform a post-analysis on your Rx data that way. This is the method we most strongly recommend for performing noise analysis on the Rx signal chain.
    2. I will have to look into this further for you. It likely won't be the most simple set of measurements, we do not make the ADC code values visible to the user. The user can configure the entire signal chain, start to finish, and has access to all of the calibration, gain / attenuation, and DSP that this chip offers, however some elements of the signal chain are masked for simplicity. One of those is the raw ADC output. We do provide specifications in the Datasheet on these metrics, so if you haven't already I do recommend you read that over on https://analog.com/ADRV9002 However for performing your own measurements, I will have to ask the team and get back to you.

    Best Regards,
    Oisín.

  • Good day, Oisin.

    Thank you for answer.

    1. Ok, I get it.

    2. The purpose of our measurements is to obtain reference values so that in the future, when developing a device on the ADRV9002, we could start from these values and decide on the success of our design. So, if this instructions is exist and it is possible to post it here (and if it makes any practical sense at all), I would be happy to consider it as a guide to action.

    Thanks.

  • Hello AntonBro,

    After consulting with my team, here is the run down on our method for computing Dynamic Range:

    The Dynamic Range test measures the SNR degradation when high level blockers present for DMR profile. 2 cases are considered. In the first case, the blocker is 1MHz away from the input signal, and 90dBc higher than the input signal. The input signal level is at -92dBm. In the second case, 2 blockers are 50kHz spaced from the signal and from each other and 68dBc higher than the input signal. In this test, external LO should be employed.

    Case 1: 

    1. In TES, configure the profile for LTE and set carrier frequency at 400MHz with external LO. (The carrier frequency is set lower due to the equipment limitation of the frequency range for external LO.)  
    2. Feed an Rx input signal at 400.500MHz. Adjust the input level to make sure the actual input level is -92dBm. Observe the Rx output “Fund Amplitude”, which represents the output signal level. Save the output data and use VisualAnalog tool (or another suitable noise calculation tool, such as MATLAB) to calculate the in-band 8KHz noise power. Calculate the SNR1. 
    3. Remove the signal and apply a blocker at Rx input at 401.500MHz at 90dBc. Reduce the Rx gain by 15dB to avoid saturation. Dump the output data and use VisualAnalog tool (or another suitable noise calculation tool, such as MATLAB) to calculate the in-band 8KHz noise power. Calculate the SNR2. The degradation is calculated as SNR1-SNR2. 

    Case 2: 

    1. In TES, configure the profile for LTE and set carrier frequency at 400MHz with external LO. (The carrier frequency is set lower due to the equipment limitation of the frequency range for external LO.)  
    2. Feed an Rx input signal at 400.500MHz. Adjust the input level to make sure the actual input level is -92dBm. Observe the Rx output “Fund Amplitude”, which represents the output signal level. Save the output data and use VisualAnalog tool (or another suitable noise calculation tool, such as MATLAB) to calculate the in-band 8KHz noise power. Calculate the SNR1. 
    3. Remove the signal and apply 2 blockers at Rx input at 400.550MHz and 400.560MHz at 68dBc. Save the output data and use VisualAnalog tool (or another suitable noise calculation tool, such as MATLAB) to calculate the in-band 8KHz noise power. Calculate the SNR2. The degradation is calculated as SNR1-SNR2. 

    Let me know if this guide helps you out at all!

    Best Regards,
    Oisín.

  • Hello, Oisín.

    Thanks a lot.

    Do I understand correctly, that in the first case the dynamic range for blocking is obtained, in the second case the dynamic range for intermodulation components?

  • Hello AntonBro,

    I'm afraid I've never done a deep dive into interpreting the results of this measurement, however there are a host of documents available online that should go through this measurement technique and what its results mean. Apologies for not being able to provide more context.

    Best Regards,
    Oisín.