+ Post New Thread
Results 1 to 11 of 11
  1. #1
    Junior Member level 2
    Points: 200, Level: 2

    Join Date
    Apr 2018
    Posts
    23
    Helped
    0 / 0
    Points
    200
    Level
    2

    Single balanced mixer

    I am designing a mixer RF,LO 20-25 GHz.. IF 1-3 GHz using ads.
    I want to simulate the P1db compression point and the gain flatness for a constant IF.. Which frequency must i use? (20? 22.5? 25? for RF 1? 1.5? 3? for IF)
    For the ideal schematic the simulation results was the same for all the range but i replace the ideal components and the simulation results are different for each frequency

  2. #2
    Advanced Member level 1
    Points: 3,253, Level: 13

    Join Date
    Nov 2013
    Posts
    483
    Helped
    104 / 104
    Points
    3,253
    Level
    13

    Re: Single balanced mixer

    I would simulate at 20 22.5 and 25, the sides are important as the centre I think. If you simulate with Periodic Steady State (PSS) analysis or transient then with higher IF the simulation should done earilier. I would use lower IF than the output IF filter's max. cut-off freq, but the highest as possible.
    "Try SCE to AUX." /John Aaron/



    •   AltAdvertisment

        
       

  3. #3
    Advanced Member level 5
    Points: 28,020, Level: 40
    BigBoss's Avatar
    Join Date
    Nov 2001
    Location
    Turkey
    Posts
    4,050
    Helped
    1216 / 1216
    Points
    28,020
    Level
    40

    Re: Single balanced mixer

    You'd better to sweep RF and LO signals while one of them is fixed in order to see the real response of the Mixer.
    I mean, while RF is swept (LO fixed) you will get Gain vs. RF.Then, you fix the RF, sweep LO, you'll get Gain vs. LO etc..
    So, every combination will give you an insight about the Mixer.



  4. #4
    Junior Member level 2
    Points: 200, Level: 2

    Join Date
    Apr 2018
    Posts
    23
    Helped
    0 / 0
    Points
    200
    Level
    2

    Re: Single balanced mixer

    Big boss i do both simulations that you say but i need a constant value of IF in order to do both of them....Which value of IF should i use? And for P1db simulation which if freq and rf freq should i use?



    •   AltAdvertisment

        
       

  5. #5
    Advanced Member level 5
    Points: 28,020, Level: 40
    BigBoss's Avatar
    Join Date
    Nov 2001
    Location
    Turkey
    Posts
    4,050
    Helped
    1216 / 1216
    Points
    28,020
    Level
    40

    Re: Single balanced mixer

    Quote Originally Posted by Kooky613 View Post
    Big boss i do both simulations that you say but i need a constant value of IF in order to do both of them....Which value of IF should i use? And for P1db simulation which if freq and rf freq should i use?
    Try different IF Frequencies such as 1GHz, 2GHz etc.You can define HB Frequencies depending on RF and LO frequencies like
    freq_IF=freq_LO-freq_RF or freq_IF=freq_LO+freq_RF...
    In additional to this, P1dB Compression Point is not a proper metric for Mixers.IIP3 is generally used to characterize the non-linearity of the Mixers.

    Or simply use Mixer Design Guide of ADS.There are many templates for Mixers.
    Last edited by BigBoss; 22nd September 2018 at 14:18.



    •   AltAdvertisment

        
       

  6. #6
    Advanced Member level 1
    Points: 3,253, Level: 13

    Join Date
    Nov 2013
    Posts
    483
    Helped
    104 / 104
    Points
    3,253
    Level
    13

    Re: Single balanced mixer

    Point of IF is that it is fixed because of the output IF filter. For P1db choose IF to the centre of your IF filter's frequency. If you design a sliding-IF receiver, than choose the min/max IF and simulate with both values.
    The RF frequency should vary for P1db from the minimum to maximum RF values, step is arbitrary, but at least simulate with the lowest/highest RF and with the centre value.
    "Try SCE to AUX." /John Aaron/



  7. #7
    Advanced Member level 5
    Points: 33,205, Level: 44

    Join Date
    Oct 2001
    Posts
    4,747
    Helped
    1401 / 1401
    Points
    33,205
    Level
    44

    Re: Single balanced mixer

    Quote Originally Posted by Kooky613 View Post
    but i replace the ideal components and the simulation results are different for each frequency
    In this situation I recommend to spend some time designing (and tuning) the LO, IF, and RF ports matching networks. This should be done for a wide range of frequencies.



  8. #8
    Junior Member level 2
    Points: 200, Level: 2

    Join Date
    Apr 2018
    Posts
    23
    Helped
    0 / 0
    Points
    200
    Level
    2

    Re: Single balanced mixer

    vfone the s parameters simulations show less than -10 db at all the range of three ports but i use harmonic balance simulation for the gain and from 20 to 23.5 Ghz i see the gain flatness i had in the ideal schematic...but from 23.5 to 25 Ghz the gain shows a sharp drop and i can not find the reason



    •   AltAdvertisment

        
       

  9. #9
    Advanced Member level 5
    Points: 28,020, Level: 40
    BigBoss's Avatar
    Join Date
    Nov 2001
    Location
    Turkey
    Posts
    4,050
    Helped
    1216 / 1216
    Points
    28,020
    Level
    40

    Re: Single balanced mixer

    Quote Originally Posted by Kooky613 View Post
    vfone the s parameters simulations show less than -10 db at all the range of three ports but i use harmonic balance simulation for the gain and from 20 to 23.5 Ghz i see the gain flatness i had in the ideal schematic...but from 23.5 to 25 Ghz the gain shows a sharp drop and i can not find the reason
    S-parameters do NOT characterize the Mixers.



  10. #10
    Junior Member level 2
    Points: 200, Level: 2

    Join Date
    Apr 2018
    Posts
    23
    Helped
    0 / 0
    Points
    200
    Level
    2

    Re: Single balanced mixer

    I use it only for matching



  11. #11
    Advanced Member level 5
    Points: 33,205, Level: 44

    Join Date
    Oct 2001
    Posts
    4,747
    Helped
    1401 / 1401
    Points
    33,205
    Level
    44

    Re: Single balanced mixer

    Impedance matching for mixer ports, it is a complex task.
    Is not enough to provide good impedance match for fundamental of RF, LO, IF, or image frequency, but also should provide short or good load impedance for harmonics of their frequencies.
    You have to spend some time reading the theory:
    http://www.rfcafe.com/references/art...tems_part2.pdf



--[[ ]]--