AD9375 matching circuit


We are using AD9375 transceiver in 2 different designs, with corresponding matching circuits in frequecy range 300-1000MHz and 3300-3800MHz. (I might add that the only difference between our two different designs is in the transceiver matching circuit)
However, the output gain of the transceiver varies drastically between 2 boards, in which in the 3300-3800 frequency range, we experience almost 8dB less gain in the output (after matching circuit).
I would appreciate if you provide us with a performance test on the working frequency range of the transceiver including associated matching circuits or any kind of additive data to see if this reduction in gain is intrinsic or we are missing something.

Thank you in advance,

  • 0
    •  Analog Employees 
    on Feb 1, 2021 3:21 PM

    Please refer to the perofmance summary plot from datasheet figure 141, page 41 and also the "3.5 GHz BAND" page 37 onwards.

    Please refer to the section "Selected Balun and Component Values" from userguide  UG-992 for matching circuit component values for AD9375 EVB.

  • Dear Pvalavan,

    Thank you for the indications.

    Regarding the Balun and component values, our board configuration is exactly as recommended by analog device in userguide.

    But regarding power graphs in datasheet, there are some vague points that I want to address and ask your guidance.

    Figure 23, 40, 141 and 158, are demonstrating output power in case of 800Mhz and 3,5 GHz. I was wondering are the test condition, the same in both cases (800MHz and 3,6 GHz)? ( same digital power input power to AD9375,  same AD9375 total TX attenuation including also RF attenuation of any external RF network)? 

    For instance, figure 158 claims that the TRX output cannot arrive to more than 4dBm of output, but in figure 141, the output power is around 6dBm, is it the 2dB digital backoff responsible for that or the test condition is different? In case of digital backoff being responsible for the difference between output power is both cases, I assume then the measurement of fig.141 was performed with max power, but comparing fig.23 and 40 discourage this conclusion.

    As indicated before, what I am looking for is a comparison between output power in both cases, initially from output port of the TRX and then including also the matching circuits.

    Thank you in advance.



  • 0
    •  Analog Employees 
    on Feb 18, 2021 12:10 PM in reply to sarahmd

    This graph is with 0dB attenuation and no digital backoff 

    This is the Tx output power variation with RF attenuation and no digital backoff.

    This graph is with 0dB attenuation and no digital backoff 

    This is with 2db digital backoff and hence the difference in output power. 

  • Hello,

    Thank you, they are still some vague point:

    -  The power of digital signal input in all the cases is the same?

    - The figure 40 is a single tone or a modulated 20MHz signal? The same question in case of figure 158, if so what is the indication of 100MHz for? and in the case that it is a 100MHz RF modulated signal what does the 3505MHz CW desired signal indicate?

    - What is the reason behind the difference in the output power (3dB almost) in case of figure 23 and 40? and why the same difference does not exist in case of figure 141 and 158 (in which considering 2dB digital backoff they are almost the same)

    Thank you in advance,


  • 0
    •  Analog Employees 
    on Feb 24, 2021 9:46 AM in reply to sarahmd

    Are you using custom board or eval board? Since you are getting a difference of 8dB, so it is not because of difference in testing conditions. If it is a custom board, then check the matching at that frequency and share the S11 plot with us.