+ Post New Thread
Results 1 to 15 of 15
  1. #1
    Full Member level 2
    Points: 1,745, Level: 9

    Join Date
    Oct 2013
    Posts
    142
    Helped
    4 / 4
    Points
    1,745
    Level
    9

    Multistage LNA design queries

    I would like to design multistage LNA to improve gain. I used inductor between one stage to next stage since inductor will bypass dc components. What best circuit or component can be used to connect one stage to next stage and next stage to further next stage.
    Once I used parallel inductor between one stage to next stage and fabricated LNA , I couldn't get results at all. What is the best way to design two or three stage LNA.
    Below attachments I have designed Single stage LNA, later extended same single stage LNA design to second and third stage. Please give suggestion on what best circuit or component can be connected one stage to next stage LNA design.

  2. #2
    Super Moderator
    Points: 54,847, Level: 57

    Join Date
    Apr 2011
    Location
    Minneapolis, Minnesota, USA
    Posts
    13,361
    Helped
    2654 / 2654
    Points
    54,847
    Level
    57

    Re: Multistage LNA design queries

    One method is to alternate N and P type. It makes for convenient biasing. Gains are multiplied. Similar to a sziklai pair.



  3. #3
    Full Member level 2
    Points: 1,745, Level: 9

    Join Date
    Oct 2013
    Posts
    142
    Helped
    4 / 4
    Points
    1,745
    Level
    9

    Re: Multistage LNA design queries

    Quote Originally Posted by BradtheRad View Post
    One method is to alternate N and P type. It makes for convenient biasing. Gains are multiplied. Similar to a sziklai pair.
    I couldn't get what you suggested, please can you tell me in detail



    •   AltAdvertisement

        
       

  4. #4
    Advanced Member level 5
    Points: 15,526, Level: 30

    Join Date
    Apr 2014
    Posts
    2,424
    Helped
    967 / 967
    Points
    15,526
    Level
    30

    Re: Multistage LNA design queries

    Quote Originally Posted by pusparaga View Post
    I used inductor between one stage to next stage since inductor will bypass dc components.
    It is not clear what you are trying to do. A single series C will be enough to for DC separation, no need for inductors unless you want to build additional high pass response.

    Your component values look a bit unfinished. Before you design multiple stages, you better design a single stage with correct values.



    •   AltAdvertisement

        
       

  5. #5
    Super Moderator
    Points: 54,847, Level: 57

    Join Date
    Apr 2011
    Location
    Minneapolis, Minnesota, USA
    Posts
    13,361
    Helped
    2654 / 2654
    Points
    54,847
    Level
    57

    Re: Multistage LNA design queries

    Quote Originally Posted by pusparaga View Post
    I couldn't get what you suggested, please can you tell me in detail
    Here's the general idea.
    Three stages, the second transistor is PNP.
    Incoming signal= 1mV amplitude.
    Overall gain= 2200.
    Output is AC riding DC component, 1.5 to 3.7 VDC

    Click image for larger version. 

Name:	3 stages amplifier NPN PNP NPN incoming signal 1mV sine gain 2200.png 
Views:	6 
Size:	36.7 KB 
ID:	158634



    •   AltAdvertisement

        
       

  6. #6
    Advanced Member level 5
    Points: 15,526, Level: 30

    Join Date
    Apr 2014
    Posts
    2,424
    Helped
    967 / 967
    Points
    15,526
    Level
    30

    Re: Multistage LNA design queries

    Have you seen this PNP design approach for LNA in the GHz range? I haven't seen much use of PNP, but might be wrong.


    1 members found this post helpful.

  7. #7
    Super Moderator
    Points: 54,847, Level: 57

    Join Date
    Apr 2011
    Location
    Minneapolis, Minnesota, USA
    Posts
    13,361
    Helped
    2654 / 2654
    Points
    54,847
    Level
    57

    Re: Multistage LNA design queries

    Quote Originally Posted by volker@muehlhaus View Post
    Have you seen this PNP design approach for LNA in the GHz range? I haven't seen much use of PNP, but might be wrong.
    Certainly my knowledge and experience don't compare to your own regarding rf electronics. I saw the OP's schematic building a series of N type common emitter amplifiers, whereas I'd seen a handy tip about alternating N & P devices. I don't know for sure it works at GHz frequencies.

    I understand microwave N type are more readily available, and easier to fabricate on chip. And normally first choice because they perform better than P type, etc.



  8. #8
    Advanced Member level 5
    Points: 34,950, Level: 45

    Join Date
    Oct 2001
    Posts
    4,993
    Helped
    1493 / 1493
    Points
    34,950
    Level
    45

    Re: Multistage LNA design queries

    Quote Originally Posted by pusparaga View Post
    I would like to design multistage LNA to improve gain. I used inductor between one stage to next stage since inductor will bypass dc components. What best circuit or component can be used to connect one stage to next stage and next stage to further next stage.
    Once I used parallel inductor between one stage to next stage and fabricated LNA , I couldn't get results at all. What is the best way to design two or three stage LNA.
    Below attachments I have designed Single stage LNA, later extended same single stage LNA design to second and third stage. Please give suggestion on what best circuit or component can be connected one stage to next stage LNA design.
    I see that in your circuits all the caps are 1pF and all the inductors are 1nH.
    You have a surplus of these components and don't know what to do with them? Or you didn't find yet how to change their values in ADS simulator?


    1 members found this post helpful.

  9. #9
    Full Member level 2
    Points: 1,745, Level: 9

    Join Date
    Oct 2013
    Posts
    142
    Helped
    4 / 4
    Points
    1,745
    Level
    9

    Re: Multistage LNA design queries

    Quote Originally Posted by volker@muehlhaus View Post
    It is not clear what you are trying to do. A single series C will be enough to for DC separation, no need for inductors unless you want to build additional high pass response.

    .
    I have attached below single stage LNA, second stage LNA, and third stage LNA.
    In stage second LNA design just I connected end of the first stage to the beginning of the second stage LNA without connecting any additional parallel inductor to by pass DC components or one more series capacitor to block any DC component.
    In third stage LNA design I connected end of the second stage to the beginning of the third stage LNA without connecting any additional parallel inductor to by pass DC components or one more series capacitor to block any DC component.

    My question is do we need to connect any addition impedance matching circuit from 1st stage to 2nd stage or any component ( series C or parallel L) to transform maximum power from 1 st stage to 2nd stage and do we need to connect any additional impedance matching circuit from 2nd stage to 3rd stage or any component (series C or parallel L) to transform maximum power from 2nd to 3rd stage.

    As you said single series C is enough between 1 stage to 2nd stage and 2nd stage to 3 rd stage. Please clarify my doubts



  10. #10
    Advanced Member level 5
    Points: 34,950, Level: 45

    Join Date
    Oct 2001
    Posts
    4,993
    Helped
    1493 / 1493
    Points
    34,950
    Level
    45

    Re: Multistage LNA design queries

    Most of the time, using two capacitors in series in a high-frequency design circuit, is a bad idea because you will get parasitics of one capacitor times 2.
    Interstage impedance matching is a very delicate (and important) phase when designing multi-stage RF amplifiers.
    The attached file may give you more information in this regards.


    1 members found this post helpful.

  11. #11
    Advanced Member level 5
    Points: 21,139, Level: 35
    Achievements:
    7 years registered

    Join Date
    Jan 2011
    Posts
    3,432
    Helped
    1231 / 1231
    Points
    21,139
    Level
    35

    Re: Multistage LNA design queries

    The concept of a multi-stage LNA is somewhat dubious, IMO. Its critical properties are going to be dominated by the first stage. More stages should just add gain without impacting NF.

    I strongly think you should focus on building a robust single-stage LNA (unconditional stability, if possible).


    1 members found this post helpful.

  12. #12
    Full Member level 2
    Points: 1,745, Level: 9

    Join Date
    Oct 2013
    Posts
    142
    Helped
    4 / 4
    Points
    1,745
    Level
    9

    Re: Multistage LNA design queries

    Quote Originally Posted by mtwieg View Post
    The concept of a multi-stage LNA is somewhat dubious, IMO. I strongly think you should focus on building a robust single-stage LNA (unconditional stability, if possible).
    But some applications, we must require large gain as alone single stage will not give more gain. We should design multistage LNA.



  13. #13
    Super Moderator
    Points: 54,847, Level: 57

    Join Date
    Apr 2011
    Location
    Minneapolis, Minnesota, USA
    Posts
    13,361
    Helped
    2654 / 2654
    Points
    54,847
    Level
    57

    Re: Multistage LNA design queries

    Another strategy is to use a darlington transistor. It's the same as placing 2 or more transistors in darlington configuration. Gains are multiplied in one stage.


    1 members found this post helpful.

  14. #14
    Advanced Member level 5
    Points: 21,139, Level: 35
    Achievements:
    7 years registered

    Join Date
    Jan 2011
    Posts
    3,432
    Helped
    1231 / 1231
    Points
    21,139
    Level
    35

    Re: Multistage LNA design queries

    Quote Originally Posted by pusparaga View Post
    But some applications, we must require large gain as alone single stage will not give more gain. We should design multistage LNA.
    Yes sometimes a single LNA stage doesn't have sufficient gain. But normally this is done by cascading a single stage LNA with a more general-purpose gain block stage, rather than cascading multiple LNAs. Or, if there's a frequency conversion in the signal chain, putting the extra gain after the mixer (an IF amplifier).


    1 members found this post helpful.

    •   AltAdvertisement

        
       

  15. #15
    Advanced Member level 5
    Points: 34,950, Level: 45

    Join Date
    Oct 2001
    Posts
    4,993
    Helped
    1493 / 1493
    Points
    34,950
    Level
    45

    Re: Multistage LNA design queries

    The old C-Band satellite LNBs from late 80'ties, early 90'ties, used at least 3 (or even 4) cascaded GaAsFET LNAs.
    Today the very high frequency devices are using the same, many cascaded LNAs.
    If you have access to the book in the link below you will see many examples of multistage LNAs:
    https://www.springer.com/gp/book/9783319690193


    1 members found this post helpful.

--[[ ]]--