Scribd Logo
Upload

Welcome to Scribd!

  • 178 views

    10 Thru 400 MHZ Directional Coupler With Ferrite

    Uploaded by

    agmnm1962
    This document summarizes the creator's attempt to build a directional coupler in a compact enclosure to measure forward and reflected power. Key points: 1) The initial version of the directional coupler worked reasonably well up to 6 MHz but performance degraded at higher frequencies. 2) Modifications were made to address isolation and insertion loss issues but the coupler was still not able to reliably distinguish between forward and reflected power above 6 MHz. 3) Further design improvements would be needed to achieve more than 20 dB of wideband isolation in a simple enclosure like the one used. The coupler design was not able to accurately measure impedances as originally hoped.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd

    10 Thru 400 MHZ Directional Coupler With Ferrite

    Uploaded by

    agmnm1962
    178 views7 pages
    This document summarizes the creator's attempt to build a directional coupler in a compact enclosure to measure forward and reflected power. Key points: 1) The initial version of the directional coupler worked reasonably well up to 6 MHz but performance degraded at higher frequencies. 2) Modifications were made to address isolation and insertion loss issues but the coupler was still not able to reliably distinguish between forward and reflected power above 6 MHz. 3) Further design improvements would be needed to achieve more than 20 dB of wideband isolation in a simple enclosure like the one used. The coupler design was not able to accurately measure impedances as originally hoped.

    Original Description:

    10 THRU 400 MHz DIRECTIONAL COUPLER WITH FERRITE

    Original Title

    10 Thru 400 Mhz Directional Coupler With Ferrite

    Copyright

    © © All Rights Reserved

    Available Formats

    DOC, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document summarizes the creator's attempt to build a directional coupler in a compact enclosure to measure forward and reflected power. Key points: 1) The initial version of the directional coupler worked reasonably well up to 6 MHz but performance degraded at higher frequencies. 2) Modifications were made to address isolation and insertion loss issues but the coupler was still not able to reliably distinguish between forward and reflected power above 6 MHz. 3) Further design improvements would be needed to achieve more than 20 dB of wideband isolation in a simple enclosure like the one used. The coupler design was not able to accurately measure impedances as originally hoped.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd
    Download as doc, pdf, or txt
    178 views7 pages

    10 Thru 400 MHZ Directional Coupler With Ferrite

    Uploaded by

    agmnm1962
    This document summarizes the creator's attempt to build a directional coupler in a compact enclosure to measure forward and reflected power. Key points: 1) The initial version of the directional coupler worked reasonably well up to 6 MHz but performance degraded at higher frequencies. 2) Modifications were made to address isolation and insertion loss issues but the coupler was still not able to reliably distinguish between forward and reflected power above 6 MHz. 3) Further design improvements would be needed to achieve more than 20 dB of wideband isolation in a simple enclosure like the one used. The coupler design was not able to accurately measure impedances as originally hoped.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd
    Download as doc, pdf, or txt
    of 7
    At a glance
    Powered by AI
    The author built a simple directional coupler to sample forward and reflected power but found it only worked reliably below 6MHz. Issues included high insertion loss, poor isolation between ports, and inability to distinguish between forward and reflected power above 6MHz.

    The author notes the directional coupler started to fail above 2MHz and frequencies above were unreliable. The isolation between input and output ports decreased with increasing frequency.

    The author measured insertion loss between ports 1 and 2, forward power indicated at port 4 with port 3 terminated, and reverse power by reversing ports 3 and 4. Charts of the measurements over frequency were created.

    Directional Coupler Vers 0.

    1
    My first attempt to make a directional coupler turned out OK but not brilliant. The goal
    was to be able to sample forward and reflected power using a simple stand-alone box. I
    think I can call this a success through 6M. M it starts to fall apart and fre!uencies
    abo"e M are simply relati"e. #ere is a schematic of "ersion $.%&
    images'(irectional)oupler'(irect)oupl*$%+chematic.,pg
    The transformers are %$ turns of -%. on a T-/$ type -6% toroid. 0y making this a 1
    port coupler2 I can re"erse either the input or output to measure reflected power.
    images'(irectional)oupler'#omebrew)oupler.,pg
    It seemed like a good idea to put some isolation between the input and output sections.
    The nifty #ammond box pro"ided a place for a di"ider so I put it into the circuit. Two
    .'34 holes were punched through some double sided board to allow the passage of the
    wires. #ot glue was used to ensure stuff remained where it was supposed to be. 5fter
    initial measurements2 the copper board was remo"ed because it added about %./d0 of
    insertion loss. The two parallel wires were separated as much as possible because they
    were coupling and damaging the isolation between ports at higher fre!uencies. I
    experimented with a di"ider between the two parallel wires but couldn6t obser"e any
    meaningful results so the di"iders were omitted completely.
    The design is completely re"ersable. The inputs can be outputs. The point is there is
    supposed to be -$d0 of separation between in'out. The two hal"es are completely
    symetrical.
    Inside 20dBCoupler2
    images'(irectional)oupler'Inside-$d0)oupler.,pg
    Initial data was collected to measure the performance. I used a 7a"etek /%$
    generator and a 0oonton 1$ power meter. 5fter calibrating the power meter2 the rest
    of the fixture needed to be documented. 8irst I calibrated the 9:/; and the related
    coupler'adapters I used to get 98 from type < connectors on my instruments to 0<).
    #ere is some raw data. The =)ontrol> column is the measured loss of the
    cables'connectors with only a 0<) barrel connector inserted instead of my new box.
    images'(irectional)oupler'(irect)oupl*$%9aw(ata.,pg
    8irst I measured insertion loss. The secondary output ports 3 and 1 were terminated
    with /$ ohm terminator 0<)s. The cable losses ?control@ were subtracted from the
    measured responses. The table therefore isolates the true insertion loss of the (AT2
    which I ha"e indicated as +%. #ere is a graph of the input port % to output port
    insertion loss o"er fre!uency. 5s you can see2 measuring down to $.%d0 has some
    funny anomalies. I belie"e you can safely attribute the weirdness to cable
    transformations as the roughly 14 of coax can do funny things at these fre!uencies. If
    you smooth out the line from /$ to 3/$ you will see the likely true response of the
    circuit. If you look at the raw data you will see that insertion loss really goes into the
    dumpster abo"e 1$$M#B2 which is why I ,ust ended my graphing of the data at 1$$.
    DirectCouplV01S12Chart
    images'(irectional)oupler'(irect)oupl*$%+%)hart.,pg
    <ext I measured the forward power indicated at port 1 with port 3 terminated using a
    /$ ohm 0<). The isolation between ports should be about %../d0. 5t low fre!uencies
    it works well. It starts to fall apart at about 6M and is still usable as a forward power
    tap point through M.
    DirectCouplV01FWDChart
    images'(irectional)oupler'(irect)oupl*$%87()hart.,pg
    Then I re"ersed port 3 and 1 to measure reflected power. )onsidering the 0<)
    terminator has a pretty flat response2 the re"erse power measured should remain
    minimal. 5bo"e 6M2 the re"erse power is not much different than the forward power.
    This does not bode well for my new gismo6s ability to tell the difference between forward
    and reflected.
    DirectCouplV01REVChart
    images'(irectional)oupler'(irect)oupl*$%9C*)hart.,pg
    5nd finally2 using the data collected2 I prepared a chart of the 8orward minus 9e"erse
    power. I belie"e this to be a measure of how effecti"e the de"ice is at measuring
    Ddirecti"ity6 of the power. 5s you can see2 when terminated into /$ ohms you would
    expect there to be "ery little re"erse power. 5s fre!uency goes up2 the measured
    reflected power increases dramatically2 which tells me that the de"ice is not able to
    determine directi"ity "ery well. I was hopeful to use this circuit to measure impedances
    at some point. I6m not so sure this will work as hoped.
    The transformers were wound on type -6% ferrite. 7ould -13 be betterE +ome other
    fla"or perhapsE Fart of the high fre!uency anomaly likely relates to the transformer
    being self-resonant at some fre!uency. I didn6t attempt to measure what that was as it
    would likely mo"e around o"er temperature !uite a bit. +omebody suggested adding
    turns but this would lowere the +98 and likely create other problems. 5ttempting more
    than $d0 of wideband isolation in a simple enclosure like this is wishful thinking.
    DirectCouplV01FWDREV
    images'(irectional)oupler'(irect)oupl*$%87(9C*.,pg
    5lthough the picture abo"e shows the circuit board used to compartmentaliBe the
    #ammond box2 I pulled it out because it was actually increasing the insertion loss for
    some reason. I also obser"ed that separating the two parallel wires helped reduce
    insertion loss. Initially2 my losses were as high as 1./d0 at %:#B and I was able to pull
    that down to about 3d0. I also noted that holding a screwdri"er across the #ammond
    box with the lid remo"ed changed the insertion loss at certain fre!uencies dramatically.
    5 ground loop needed attention. 5t 1/$2 .$$2 the obser"ation was most noticable with
    a change of o"er %d0 possible using the blade of a screwdri"er. C"entually I pulled the
    weather proof insulation out of the box lid and pro"ided a decent electrical path through
    the lid2 which helped a bit. These measurements were taken after all inter"entions were
    completed.

    You might also like