+ Post New Thread
Results 1 to 9 of 9
  1. #1
    Full Member level 1
    Points: 652, Level: 5

    Join Date
    Jun 2018
    Posts
    110
    Helped
    10 / 10
    Points
    652
    Level
    5

    X-Band Multiple Channel Config

    Hello all,

    I am working on designing a X-band radar for motion detection. Problem I am facing now is that I will have multiple Tx/Rx channels(connected to diferent antennas) running(pulsed radar) and I need to synchronize or lock them to each other somehow. I have some ideas but not sure which one would work.

    1-) A way to synchronize different VCO's to each other. I checked out some PLL's, however in their datasheet RefinA and RefinB refers to complementary signals and not two distinct signals.
    Can anyone suggest a way to do it?

    2-) Using a switch to alternate between Antennas; however, due to high power nature of PA, I'm not sure what would be the effect of thermal memory on that.

    3-) Using a N-way Divider at the output of VCO, then using the divided signals as Tx_signal in different channels.

    Ideas are appreciated on the matter.

    Kind Regards,

    ktr

    •   AltAdvertisment

        
       

  2. #2
    Advanced Member level 5
    Points: 33,431, Level: 44

    Join Date
    Oct 2001
    Posts
    4,781
    Helped
    1412 / 1412
    Points
    33,431
    Level
    44

    Re: X-Band Multiple Channel Config

    I think you don't have to synchronize the RF signals coming from the VCOs, but you can synchronize the pulses.
    Synchronizing the pulses should be an easy job, using a microcontroller or even simple digital logic ICs.
    Now depends what power you have, but you can get high-power (100W or even more) PIN diodes which can be used as switches up to 12GHz.
    http://cdn.macom.com//datasheets/MADP-011029.pdf



  3. #3
    Full Member level 1
    Points: 652, Level: 5

    Join Date
    Jun 2018
    Posts
    110
    Helped
    10 / 10
    Points
    652
    Level
    5

    Re: X-Band Multiple Channel Config

    My problem is the VCO I have dont have enough power by itself to support multiple mixers.
    I think I will use 2 PLL's using same XCO, then those VCO's will be locked to the same source, hence I can use N-way divider before Tx_PA and give the aforementioned Rx channel mixers same LO.

    @vfone

    Quote Originally Posted by vfone View Post
    Now depends what power you have, but you can get high-power (100W or even more) PIN diodes which can be used as switches up to 12GHz.
    http://cdn.macom.com//datasheets/MADP-011029.pdf
    Do you think using these diodes is a bettter way of switching compared to controlling gate voltages of PA's? Tx channels will not transmit at the same time.



  4. #4
    Advanced Member level 5
    Points: 33,431, Level: 44

    Join Date
    Oct 2001
    Posts
    4,781
    Helped
    1412 / 1412
    Points
    33,431
    Level
    44

    Re: X-Band Multiple Channel Config

    Switching the PAs is better choice than using PIN diode switches, which have limited isolation vs frequency.
    Have to be sure that settling time of the PAs during their switching is good for your application. Usually, higher the power, higher is the settling time.


    1 members found this post helpful.

    •   AltAdvertisment

        
       

  5. #5
    Member level 4
    Points: 641, Level: 5

    Join Date
    Oct 2017
    Posts
    74
    Helped
    13 / 13
    Points
    641
    Level
    5

    Re: X-Band Multiple Channel Config

    For 10GHz radar I would try 5GHz VCO + relatively cheap 6GHz switches + 5ghz doubling mixer. I have seen such approach used to switch between 10 or more antennas from a single VCO.


    1 members found this post helpful.

  6. #6
    Full Member level 1
    Points: 652, Level: 5

    Join Date
    Jun 2018
    Posts
    110
    Helped
    10 / 10
    Points
    652
    Level
    5

    Re: X-Band Multiple Channel Config

    Quote Originally Posted by vfone View Post
    Switching the PAs is better choice than using PIN diode switches, which have limited isolation vs frequency.
    Have to be sure that settling time of the PAs during their switching is good for your application. Usually, higher the power, higher is the settling time.
    I found a switch which I thought good for me. https://www.analog.com/en/products/adrf5027.html this one(SPDT)Click image for larger version. 

Name:	ctrl_voltage.PNG 
Views:	2 
Size:	14.5 KB 
ID:	151134

    I am thinking of giving a control voltage with duty cycle proportionate with PA's settling time + Tx's switching time(how long my pulse is). Then, I can control 1 PA would work at any given time, and I would have all receiving antennas work (MIMO) if I put switch. Another Switch will turn on/off PA's gate voltage, a low frequency one because gate voltages are DC.

    So, it makes 2 switches, one on the RF line, one at the gate voltage control.

    Is that a correct approach?

    - - - Updated - - -

    Quote Originally Posted by Georgy.Moshkin View Post
    For 10GHz radar I would try 5GHz VCO + relatively cheap 6GHz switches + 5ghz doubling mixer. I have seen such approach used to switch between 10 or more antennas from a single VCO.
    Doubling mixer? Can you tell me about it? First time I hear such a thing.
    Last edited by ktr; 14th February 2019 at 08:11. Reason: wrong wording



  7. #7
    Member level 4
    Points: 641, Level: 5

    Join Date
    Oct 2017
    Posts
    74
    Helped
    13 / 13
    Points
    641
    Level
    5

    Re: X-Band Multiple Channel Config

    Quote Originally Posted by ktr View Post
    Doubling mixer? Can you tell me about it? First time I hear such a thing.
    Check out this patent: US20090015465
    As a result, the power of the RF signal is distributed equally and in phase between the second terminal 12c and the fourth terminal 12e. These arrangements allow an IF signal to be generated at the frequency fIF (=fRF−2*fLO) at the third terminal 12d and output through the IF filter circuit 22 during reception, as well as allowing a second harmonic of the LO signal to be generated (at 2*fLO) at the third terminal 12d and output as a transmission signal during transmission.
    ....
    a primary object of the present invention to provide a relatively simple mixer circuit capable of generating an even harmonic of a received local oscillator signal and further generating an intermediate frequency signal from a received radio frequency signal having a frequency close to the frequency of the even harmonic of the local oscillator signal, and capable of allowing concurrent use of the generated even harmonic of the local oscillator signal and the generated intermediate frequency signal. It is a secondary object of the present invention to provide a radar transceiver using the above simple mixer circuit.
    It uses LO, transmits LO*2 and receives RF which must be relatively close to LO*2.
    I have seen similar doubling mixer used for 10GHz FMCW scanning antenna array with 6ghz switches, no separate transmit antenna. Each antenna had separate rat-race doubling type mixer, which produced 10GHz from 5GHz. Very interesting design.
    Last edited by Georgy.Moshkin; 14th February 2019 at 09:36.



  8. #8
    Advanced Member level 5
    Points: 33,431, Level: 44

    Join Date
    Oct 2001
    Posts
    4,781
    Helped
    1412 / 1412
    Points
    33,431
    Level
    44

    Re: X-Band Multiple Channel Config

    Initially you mentioned: "high power nature of PA".
    The switch from Analog Devices supports only +24dBm of power (250mW).



    •   AltAdvertisment

        
       

  9. #9
    Full Member level 1
    Points: 652, Level: 5

    Join Date
    Jun 2018
    Posts
    110
    Helped
    10 / 10
    Points
    652
    Level
    5

    Re: X-Band Multiple Channel Config

    Quote Originally Posted by vfone View Post
    Initially you mentioned: "high power nature of PA".
    The switch from Analog Devices supports only +24dBm of power (250mW).
    Yes, at first I thought I would put switch after PA, then I realized that switch can not handle 1-2 W power, so I decided to use it before PA, where line power is around 10mW. I will use a secondary switch to alternate between gate voltages of PA's, so the one that wont be used wont dissipate unnecessary energy.

    Click image for larger version. 

Name:	edaboard.png 
Views:	4 
Size:	13.6 KB 
ID:	151138



--[[ ]]--