[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

Re: Power divider for X-yagis

Hi All,

Building power combiners out of transmission lines and paralleling antennas 
has a lot of limitations. If you assume that both antennas are exactly 50 
ohms, then a piece of coax of 35.4 ohm impedance will rotate the impedance 
to 50 ohms. This type of splitter will have exactly 3.01 dB of split loss on 
each port ONLY if the two antennas are exactly 50 ohms. If for example, your 
antennas are matched to 1:1.2 VSWR this means that one antenna could be a 
41.7 ohms and the other at 60 ohms. In this case, one antenna will be at 
-3.9 dB and the other antenna will be at -2.3 dB. At an VSWR of 1:1.4 (still 
considered good) the split between one antenna and the other will differ by 
almost 3 dB. As you can see, the circularity is going down hill quick. It is 
much much better to use another type of splitter that offers isolation 
between the antennas.

These numbers above are pretty much the same if you use a pair of 70.1 ohm 
lines to convert each antenna impedance to 100 ohms and then parallel the 
transmission lines. In order to become less sensitive to antenna match, you 
need to add a 100 ohm resistor between the antenna legs (in the case of the 
70.1 ohm line example). That will make the split -3.01 dB even with 1.4 

One last thing to keep in mind is reactive terminations. If your antenna is 
not matched to look purelly resistive, any reactance (even with relativelly 
low VSWRs) will make your splitter not be 0 degrees.

Pieter, N4IP

>From: "i8cvs" <domenico.i8cvs@tin.it>
>To: <pa1jav@AMSAT.Org>, "AMSAT-BB" <amsat-bb@AMSAT.Org>
>Subject: Re: [amsat-bb] Power divider for X-yagis
>Date: Fri, 26 May 2006 16:03:13 +0200
>MIME-Version: 1.0
>Received: from amsat.org ([]) by bay0-mc9-f3.bay0.hotmail.com 
>with Microsoft SMTPSVC(6.0.3790.1830); Fri, 26 May 2006 07:23:08 -0700
>Received: from amsat.org (localhost.ucsd.edu [])by amsat.org 
>(8.13.6/8.13.1) with ESMTP id k4QE4Cas026908;Fri, 26 May 2006 07:04:13 
>-0700 (PDT)
>Received: (from majordom@localhost)by amsat.org (8.13.6/8.12.3/Submit) id 
>k4QE4CWP026907;Fri, 26 May 2006 07:04:12 -0700 (PDT)
>Received: from mailbox4.ucsd.edu (mailbox4.ucsd.edu []) by  
>amsat.org (8.13.6/8.13.1) with ESMTP id k4QE3Ngk026870; Fri, 26 May  2006 
>07:03:24 -0700 (PDT)
>Received: from vsmtp2.tin.it (vsmtp2.tin.it []) by  
>mailbox4.ucsd.edu (8.13.6/8.13.5) with ESMTP id k4QE3JCb002236; Fri,  26 
>May 2006 07:03:20 -0700 (PDT)
>Received: from B3O7F1 ( by vsmtp2.tin.it ( id  
>4476ADF30005C9B1; Fri, 26 May 2006 16:04:59 +0200
>X-Message-Info: LsUYwwHHNt0+mmY1v7oVPPjYWJ7uWOmlGkCUtST0VHI=
>References: <000001c680a1$93f14370$0800a8c0@Kronos>
>X-Mailer: Microsoft Outlook Express 5.50.4522.1200
>X-Greylisting: NO DELAY (Relay+Sender autoqualified); processed by  
>UCSD_GL-v2.1 on mailbox4.ucsd.edu; Fri, 26 May 2006 07:03:21 -0700  (PDT)
>X-Spam-Level: Level X-Spamscanner: mailbox4.ucsd.edu  (v1.6 Aug  4 2005 
>15:27:38, 0.0/5.0  3.0.4)
>X-MailScanner: PASSED (v1.2.8 824 k4QE3JCb002236 mailbox4.ucsd.edu)
>Precedence: bulk
>Return-Path: owner-AMSAT-BB@AMSAT.Org
>X-OriginalArrivalTime: 26 May 2006 14:23:09.0031 (UTC) 
>----- Original Message -----
>From: "Joop & Tineke Verdoes" <joop.verdoes@nameplanet.com>
>To: <amsat-bb@amsat.org>
>Sent: Friday, May 26, 2006 10:51 AM
>Subject: [amsat-bb] Power divider for X-yagis
> > Hello to you all.
> >
> > When buidling a phasing harness for (switchable) circularly polarised
> > crossed-Yagis one needs a device to split the line from the TRX to the
> > two legs of the yagi.
> > In the ARRL antenna handbook (and others) I see quart lambda 36 ohm
> > transformers in the figures.
> > In my current state of confusion I have the following questions.
>Hi Joop, PA1JAV
>Assuming that the impedance of each leg of the yagi is 50 ohm than two
>legs in parallel make the resulting impedance of the antenna equal to
>25 ohm
>To match 25 ohm impedance of the antenna to 50 ohm impedance of a
>transmission line we can use the property of impedance transformation
>given by a 1/4 electrical wavelengh long transmission line wich
>characteristic impedance is Zo = SQR ( 25 x 50 ) = 35.4 ohm  or about
>36 ohm in this case.
> > 1- this 36 ohm transformer is apparently a square pipe with a round rod
> > to which three N sockets connect. Where do I find "the right" formula
> > for calculating the characteristic impendance for square pipe with a
> > round rod in it. I coulnd't find it in any of the ARRL or RSGB books I
> > have. (but I might have gone blind for just this :-)) On the net I found
> > different formulas.
> >
>If you look at the Microwave Handbook Volume 2 of the RSGB edited
>by M.W.Dixon, G3PRF page 13.11 you will find that for a square tube
>with air as dielectric and equal sides with inside dimension (D) and a
>round inner conductor of outside diameter ( d ) the characteristic 
>Zo can be computed using the following formula where (D) and (d) are
>expressed in millimeters.
>                               1.178 D
>Zo= 138 log      ( ----------- )
>                       10         d
> > 2- this 36 ohm transformer correctly transforms the two parallel
> > impedances of the 50ohm lines to the legs of the yagi (so 25 ohms
> > combined) to the 50 ohms of the TRX. But.....seen from the yagi there is
> > a mis-match, impacting the received signal.
>The impedance transformation is made at sacrifice of the SWR but only
>inside the 1/4 wavelenght long matching line wich is 35.4 / 25 = 1.41
>as seen with the antenna as a source while receiving and 50 / 35.4 = 1.41
>as seen by the antenna as a load while transmitting.
>In both situation the loss of power inside the above matching line due to
>a SWR = 1.41 is in the order of  35 dB wich is insignificant.
> > Wouldn't it be better to use a Wilkinson power divider/combiner, from
> > either (70.7 ohm) square pipe or coax? There isn't much difference in
> > the work to make one or the other, so I might as well do this the best
> > way.
> >
>Assuming that the impedance of each leg of the yagi is 50 ohm than a  75 
>1/4 electrical wavelenght long matching line made with RG-11 or RG-59
>coax cable connected in series to one leg will make an impedance
>transformation of  75^2 / 50 = 112.5 ohm as seen at the other end.
>If you do the same for the other leg and you put two 112.5 ohm in parallel
>to a T adapter than you have a load impedance of 56 ohm to be connected to
>any 50 ohm feed line with a theoretical SWR of only 1.12
>This metod is convenient up to 144 MHz  and is more economical than using
>a 36 ohm maching line made with square tubing.
>Above 144 MHz the 36 ohm maching line made with square tubing is preferred
>for several reason.
> > I really would appreciate some help here.
> >
> > Thanks
> >
> > Joop, PA1JAV
> >
>Best 73" de
>i8CVS Domenico
>Sent via amsat-bb@amsat.org. Opinions expressed are those of the author.
>Not an AMSAT member? Join now to support the amateur satellite program!
>To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org
Sent via amsat-bb@amsat.org. Opinions expressed are those of the author.
Not an AMSAT member? Join now to support the amateur satellite program!
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org