Alek Petkovic wrote:
>
> Looks good Dave. Big is always beautiful. Handles on a tube do it to me
> every time. Hi hi.
>
> I am particularly interested in the directional coupler you mentioned.
>
> I have some 500W power meters ex mobile phone base stations which have
> directional couplers designed for the 850 -900 MHz region. I'm of course
> more interested in 2m and 70cm and I'm not too clued up on the ins and outs
> (pardon the pun) of directional couplers.
I'll dig the directional coupler out. The first thing I did was try to
expand my finite difference software at
http://www.david-kirkby.co.uk/ham/finite.html
to handle directional couplers. As it stands, the code can find the
impedance of a rectangular inner inside a rectangular outer, with any
offset between them - a case for which there is no analytical solution.
In theory, the code can be altered to find the odd and even mode
impedance's of a directional coupler, from which you can calculate the
performance. This would be really useful, but despite a lot of effort, I
could never get it to work. For a while at least, I have given up with
that.
In principle, it is not that difficult to find the impedance,
directivity and coupling factor of a coupler (all as a funcition of
frequency) consisting of two rectangular lines in a rectangular box. All
one needs to calculate is the odd and even mode impedance's and the rest
is easy. However, each time I tried this with my finite difference code,
so the answers I got for Z-odd and Z-even disagreed with some published
values of odd and even mode impedance's in the professional literature.
In the end, I built a coupler using my results, and not surprisingly it
worked badly. If anyone would like to continue this, or have a try, let
me know. A paper describing the theory of how to find Zo of a simple
transmission line was published in QEX and can be found on my web site.
This just needs expanding to find Z-odd and Z-even of a pair of coupled
lines. I can't see why I could not get it to work, but I could not. I've
never had so much hassle with a bit of scientific code.
The coupler I made for the 2 m amp consists of two 7-16 connectors at
each end of a piece of rectangular cross section square tube. Inside
that, between the two centres of the 7-16 connectors, is a round rod, of
the diameter such that the lot forms a 50 Ohm transmission line. The
7-16 connectors had threaded centres, so I tapped the inner at both
ends, so it could all be screwed together. The square tube I used was
thick enough to allow it to be taped so the 7-16 connectors could be
attached.
Two PCB tracks are inserted on inside of the square tube - one at each
opposite side. These are glued to the side walls. I adjusted the widths
of the tracks by trial and error with a spectrum analyser, to make the
coupler as directional as possible.
SMA connectors are put on the ends of the two PCBs. At one end, a 50 Ohm
load is put, at the other the RF output is taken.
The biggest problem was getting the directivity. Getting the track
widths just right was very difficult. I would have probably been better
to etch tracks of various widths, but what I did was to cut the tracks
on a band saw. Hence the copper and dielectric are of the same width.
They are not as precise as one could etch.
I don't have my own digital camera and the coupler is in a box with a
load of bits I removed from work when I left (on Friday). I'll try to
dig it out to find the dimensions of everything.
I must add I've not checked the coupler at high power. I see no reason
the inner to outer should break down, as the spacing is very large. I'm
less sure about the effect of the RF on the PCB. However, the E-field
close to the edge of the square cross section, where the PCBs are, is
very low.
I'll sort the coupler out and put some dimensions on the web later.
David Kirkby (G8WRB).
>
> Hope you have/had a fun Christmas. We had a hot one here. 40C here in
> Perth, Western Australia at lunch time today!!!
>
> 73 de Alek,
> VK6APK
>
> At 09:32 PM 12/24/2000 +0000, you wrote:
>
> >Hi,
> > I'm in the process of constructing a 2m amplifier using the YC156 or
> >3CX5000A7 tubes - the tubes are not identical, but are similar. See
> >http://www.david-kirkby.co.uk/ham/yc156.html
> >
> >The driver, which uses a conduction cooled version of the 4CX250B is
> >detailed at
> >http://www.david-kirkby.co.uk/ham/y799.html
> >
> >If anyone has any comments, I'd be interested. The designs are not
> >complete and working, so don't bother copying just yet. However, it
> >might be of interest to some.
> >
> >--
> >Dr. David Kirkby Ph.D,
> >email: drkirkby@ntlwold.com (formally davek@medphys.ucl.ac.uk)
> >web page: http://www.david-kirkby.co.uk
> >Amateur radio callsign: G8WRB
> >
> >--
> >FAQ on WWW: http://www.contesting.com/FAQ/amps
> >Submissions: amps@contesting.com
> >Administrative requests: amps-REQUEST@contesting.com
> >Problems: owner-amps@contesting.com
--
Dr. David Kirkby Ph.D,
email: drkirkby@ntlwold.com (formally davek@medphys.ucl.ac.uk)
web page: http://www.david-kirkby.co.uk
Amateur radio callsign: G8WRB
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|