AnsweredAssumed Answered

AD9361 Tx filter calculations

Question asked by dhgeon on Nov 29, 2017
Latest reply on Dec 6, 2017 by dhgeon

Regarding the AD9361 Tx third-order Butterworth BaseBand (BB) Low Pass Filter (LPF):

  • UG-570's "BASEBAND Tx ANALOG FILTER CALIBRATION" allows for the derivation of the following calculation for maximum precision of nominal (theoretical) in-situ (actual) BB LPF 3dB cutoff frequency:
    • BBBW_DESIRED,MHz // this is desired BaseBand BandWidth, which is "half the complex bandwidth"
    • BBPLF_3dB_DESIRED,MHz = 1.6 * BBBW_DESIRED,MHz // "normally calibrated to 1.6x the BBBW"
    • BBLPF_3dB_ACTUAL,MHz = BBPLL_ACTUAL,MHZ * ln(2) / (2 * pi * ceiling(BBPLL_ACTUAL,MHz * ln(2) / [2 * pi * BBPLF_3dB_DESIRED,MHz] )


Regarding the AD9361 Tx single-pole secondary LPF:

  • this filter (which appears to be a single-pole RC filter with R values from UG-671 Table 37.) has no such formula (missing element appears to be the C values?)
    • other forum posts refer to:
      • [AD9361 RF Bandwidth ] an older AD9361 Calibration Guide which was supposedly merged into UG-570, but UG570 doesn't specify any secondary LPF formulas
      • [AD9361 tx secondary filter resistor and capacitor ] refers to:
        • The MATLAB Filter Design Wizard, which provides formulas for analog filter response (butter(...)), but they do not provide any insight into in-situ values for secondary LPF via AD9361 registers 0x0D0, 0x0D1, 0x0D2
        • The No-OS ad9361_tx_bb_second_filter_calib() function, which calculates register 0x0D2 values but we still don't know what capacitor values they correspond to


Can you 1) confirm the validity of the AD9361 Tx BB LPF 3dB cutoff frequency formula and 2) provide a similar formula (or enough information to derive it) for AD9361 Tx secondary LPF?