[OZAPRS] Cavity info

[Ray Wells]

Hi Tony,


Tony King wrote:
> Hi Ray and thanks for the reply,
>
> The APRS Digi will be running on a separate antenna which will be about 
> 12 meters away from the antennas for the repeater.
>   
Is that vertically or horizontally?
> As is usual for these repeater sites the hut is full of other services 
> and even though the intermod software indicates no significant problems 
> with interference from or to other services there is so much gear 
> including the local FM station on the same tower I thought it would be 
> wise to try to avoid any desense problems in advance.
>   
Even a single cavity is a worthwhile addition on a busy site.
> The 2m repeater is running on two separate antennas 5m vertical 
> separation with the receive having two band pass and one notch cavity  
> (on the tx frequency). The Tx has a single band pass on the tx frequency.
>   
You may find the existing cavities will provide some protection to/from 
the aprs system. I would erect the new antenna and do some isolation 
measurements ....

1. Grab a test rx and measure its sensitivity.
2. feed the sig gen into the aprs antenna  and connect the rx in place 
of the repeater rx, and again measure sensitivity. The difference is the 
isolation.
3. feed the sig gen into the aprs antenna and connect the test rx in 
place of the repeater tx, and again measure sensitivity. Then you'll 
have some meaningful isolation figures. Then you'll know what additional 
measure are needed, if any.
> Ok on the silver plating etc. and temperature problems.  The repeater 
> hut is air conditioned and hopefully the fairly constant temperature 
> will mean that I don't need an Invar or similar material  compensating 
> rod for the centre conductor, it's just that I seem to recall seeing 
> articles on "home building" cavities way back in the early days of QST 
> magazine and even one in A.R. and always preferring to build rather than 
> buy reckoned it is worth a try
>   
If you have the facilities to manufacture, why not.
> A google search didn't find any references and I don't' have any of the 
> older articles.
>
> I have the machine tools to do the job and I had also thought that if 
> design formulas were available I could adjust the width and length as 
> required so that I could make the whole outer from one piece of 
> aluminium bar or tube and even turn a very fine internal thread (like a 
> camera lens thread) on the other for fine adjustment.
>   
I'd source some large diameter tubing with a reasonable wall thickness, 
for stability. The inner could be solid if that was more convenient.

The RFS cavities have a course adjustment that is locked in place in a 
threaded boss. The boss is screwed in or out for fine tuning. The TCA 
cavities have a long threaded rod that attaches to the adjustable end 
section of the inner conductor. Lots of ways to skin a horse.
> Another reason both for building and designing my own is that I also 
> need half a dozen or so in my shack. With the APRS,  IRLP, packet BBS, 
> and several radios on 2m all running 24/7 I suffer an awful lot of 
> desense and cross mod.
>   
With such close frequency spacings you may never eliminate desense but 
every bit helps. Close frequency spacing is a pain to work with.
> It would also be nice if I could reduce the number of antennas hanging 
> off my shack as well. It looks like a frightened Echidna at the moment.
>   
You could run all receivers from a common antenna, with a bit of filter 
protection from the tx's. Use a gasfet preamp and a splitter after the 
filters. We used to do that on our UHF sites with up to a dozen UHF 
repeater sharing two antennas.

Transmitters can be combined using a variety of methods but the least 
lossy is probably the circulator followed by a cavity for each tx. The 
outputs of the cavites are combined as a star point for the antenna 
connection. Fot tx only 60dB isolation is considered ok.

I've scanned 11 pages of that book as pdf files. The files have ended up 
a little on the large side so I've created an account for you on my 
Linux system so you can ftp them from me.
ip address is 115.70.139.149
login vk3api (case sensitive)
password ynot (case sensitive)

Yell if you have any dramas. You can do ftp transfers with Internet 
Explorer or Mozilla.

Cheers ... Ray