Post Go back to editing

Pluto rx adc different reference level problem

Hello

I measured today twice using pluto rx function in python.

Once the transmitter was active before the beginning of rx and the second time it only started after receiving.

On both times the noise level was different. It was like the pluto adjust by itself the adc reference level According to the first sample.

On both time the hardware gain was manually controlled and the same.

Do you know why it happens?

Thanks,

Etamar

Top Replies

    •  Analog Employees 
    May 17, 2021 in reply to et.l +1 verified
    1. If i calibrate the devie once, and afterwards save all the gain table used, only to charge them in later use - will it the measurement results be repeatable? 

    I cannot guarantee this. You will…

  • This is likely from calibrations that are performed when you set the sample rate, update the gain modes, or tune the LO. Note that with pyadi-iio every time you write to the sample_rate property it will recalibrate the system, even if the rate is not changing.

    -Travis

  • Thank you,

    Where can I read about what happens in this calibration?

    Is it a process done in the pyadi level, pluto c libraries, or in the ad9361?

    Thanks,

    Etamar

  • Within the kernel driver. You can turn them off through python with:

    sdr = adi.Pluto()

    sdr._ctrl.attrs['calib_mode'].value = 'manual'

    based on: https://wiki.analog.com/resources/tools-software/linux-drivers/iio-transceiver/ad9361?s%5B%5D=fastlock#calibration_mode_controls

    However, the ADC codes will not be exactly the same and it will also depend on the source signals power.

    Pluto is not an instrument.

    -Travis

  • thank you!

    two last questions:

    1. If i calibrate the devie once, and afterwards save all the gain table used, only to charge them in later use - will it the measurement results be repeatable? 

    2. I expect to follow both the ad9361 refernce manual, and libiio documents, i expect it to take me a month, when my final goal is to have an "home made" spectrum anlyzer, pre-calibrated at specific frequencies (in several pre-determined refernce level). is it the right way doing so? (without another, proper instrument)

  • 1. If i calibrate the devie once, and afterwards save all the gain table used, only to charge them in later use - will it the measurement results be repeatable? 

    I cannot guarantee this. You will need to verify how well this would work for your application.

    2. I expect to follow both the ad9361 refernce manual, and libiio documents, i expect it to take me a month, when my final goal is to have an "home made" spectrum anlyzer, pre-calibrated at specific frequencies (in several pre-determined refernce level). is it the right way doing so? (without another, proper instrument)

    It really depends on what level of performance you expect and how well you can actually calibrate it based on configuration and input signals. I would highly recommend having a reference instrument for comparison.

    The AD936x family are highly dynamic parts and operation is highly dependent on calibration at a particular temperature, frequency, rate, input signal...  . Depending on your requirements of accuracy it may work or not.

    -Travis