Post Go back to editing

AD9631 AUXDAC effects



In the manual, it is stated that the DAC output voltage is set in two register 0x18 and 0x1A.


In my experiment, I only change the value in register 0x18. It will cause the drift of output carrier frequency, say a two-three hurdred Hz.

My carrier frequency is 2.35GHz and the reference clock is 19.2MHz.


Based on the manual, the settings of DAC output voltage should not affect the drift of carrier frequecy.


Could you help to explain such strange effects?



  • Which hardware platform are you using for evaluation?

    Is the drift transient, i.e. does the frequency go back to its original value?

    Are you operating in TDD or FDD mode?

    Is the evaluation platform and your reference clock stabilized in temperature?

  • It works in TDD mode. I think the reference clock is stabilized in temperature.

    These effects are repeatable.

    My question is that is this effect expected when only changing the value in register 0x18?

  • My suggestion would be to have all the required blocks enabled and turned on before doing a synthesizer configuration/calibration.

    FDD VCO calibration is more accurate than TDD VCO calibration and in this case can be used to rule out temperature drift as a cause. You can use the part in FDD mode and do the same AUXDAC experiment like you did in TDD mode to see if that is what is causing your issue.

  • My previous problem is solved. Many thanks!

    For another question, I set the RF transceiver in manual gain mode. I injected a single tone signal with constant power into the transceiver and then I read back the RSSI register 0x1A7. I got a value with that receiver power.

    With the same setting, I only change the input power but I read back the RSSI register 0x1A7. It give me the same value.

    Please advise.

  • AD9361 Gain Control and RSSI User Guide document goes into detail on how and when RSSI alogrithm restarts and how the accumulator is cleared. You will find Table 21 useful.

  • AD9361 Gain Control and RSSI User Guide document states that

    " Equating the RSSI read-back value to an absolute power level (e.g., in dBm) requires a system factory or bench calibration..... "

    Actually, my design requirement is that when I read back the RSSI value and receive path gain in MGC mode. I can determine the exact received input power in the front-end.

    How do I perform such calibration ?

  • The procedure is explained in the Factory Calibration, Rx RSSI section of AD9361 Calibrations User Guide document (you will find AD9361 Register Map document also useful).

  • I read both the Factor Calibration n the Register Map document. But it doesn't give any details about converting RSSI value to absolute Rx input power.

    The setting for my experiment are that: MGC mode, Full Gain table, FDD mode, RSSI mode is 011.

    The observation for my experiment is that, register 0x1a7 and 0x1a9 are only changed when the gain value index is change. When the Rx input power increase or decrease, the RSSI value doesn't change.

    Please advise.


    The AD9361 measures RSSI by measuring the power level in dB and compensating for the receive path gain. The various options available support both TDD and FDD applications. Note that the RSSI value is not in absolute units. RSSI value is in 0.5dB resolution when 8 bits are read or 0.25dB resolution when 9 bits are read. Equating the RSSI read-back value to an absolute power level (e.g., in dBm) requires a system factory or bench calibration. To calibrate the RSSI word to an absolute reference, inject a signal into the antenna port of the completed system and read the RSSI word. From this test, generate a correction factor that equates the RSI word to the injected signal level at the antenna port.


    RSSI Mode Select Table in AD9361 Gain control and RSSI User Guide shows different events for which RSSI measurement is restarted. You may find mode 100 useful in your application. The behavior you are seeing is correct for the mode you are in (mode 011)



  • Thanks for the information.

    One more question is that the range of the registers of 0x1A7 and 0x1A9 is 0 to -128dB with 0.5dB resolution.

    I am confused with two different calculation about the above register.

    For example, if the value of 0x1A7 is 0x0A,

    Calculation 1)

    10 * 0.5 - 128 = - 123 dB.

    Calculation 2)

    - (10* 0.5) = -5 dB

    Which calculation is right?