Teardown: Kathrein GSM900 base station sector antenna

[wkb]

different frequency? maybe transmit and receive frequencies ?
the size is so different, it doesnt make sense.
best step forward is to put the antenna's individually on a network analyzer.

Well, the short dipoles had symmetrical coax lengths to the junction box. I imagine
the H and V dipoles are for some kind of polarisation diversity setup(?).  That I can understand.
But the long dipoles with different cabling. no. ENOCLUE

I don't own a network analyser. And (uhm..) I did a rather thorough job in dismantling the thing.
Kept only two long dipoles with cable still attached.

There is quite a bit more to this beast than I had imagined

[PE1RKI]

if you need a little antenna for 23cm then you can also use the 8 antenna.
about 15 years ago i made a little guide for this little antenna http://pe1rki.com/quad.html
very crusty pics  , i have very little free time to update my website, i'm aware it has tons of mistakes.
i made one for 13cm as well and my 500mw tv picture was received 40km's away.
tx and rx were 25meter and 35meter above ground level though. 

[wkb]

if you need a little antenna for 23cm then you can also use the 8 antenna.
about 15 years ago i made a little guide for this little antenna http://pe1rki.com/quad.html
very crusty pics 
i made one for 13cm as well and my 500mw tv picture was received 40km's away.
tx and rx were 25meter and 35meter above ground level though. 

Ah yes, I once built one like that for WiFi. Worked like yours: very nicely indeed!

The whole idea is to play with this GSM thing "because I can". Very practical in use... ah, maybe not.

[PE1RKI]

you can truly play with it if you can measure the results.
without equipment it is always walking in the dark.

[wkb]

you can truly play with it if you can measure the results.
without equipment it is always walking in the dark.

I plan to use a SA with TG and. RLB. Should give me an idea or?

[PE1RKI]

yes that works. i never used that myself though.
with no reflected signal from the antenna you at least know it radiates, but you dont know in what direction yet. 

[wkb]

yes that works. i never used that myself though.
with no reflected signal from the antenna you at least know it radiates, but you dont know in what direction yet. 

At least I know the direction won't be through the alu base plate   

Measured the dipoles in the meantime: 14 and 16cm span respectively.  16cm I can match to the 900MHz
GSM band. The Kathrein label says  870-960MHz.   But the 14cm ??

And there is a little problem: 1/4 lambda in coax with VF 0,71 is really short on 23cm. Too short to bridge the distance between 2 dipoles using the original base plate. 

[PE1RKI]

you can add 50 ohm coax in between the antenna and 75 ohm pieces.

i dont get those lengths either.
maybe it has to do with the distance above the alu baseplate.

[wkb]

you can add 50 ohm coax in between the antenna and 75 ohm pieces.

i dont get those lengths either.
maybe it has to do with the distance above the alu baseplate.

Both dipole types are approx 5.5cm above the baseplate. The 16cm version slightly protrudes over the alu reinforcement(? for stiffness?) 'rib'.

Originally it uses 75R soldered onto the dipoles. Does that matter?

[PE1RKI]

yes that may mean that the dipoles are not 50 ohm.
then our combining method will not work.

[wkb]

yes that may mean that the dipoles are not 50 ohm.
then our combining method will not work.

Ok.  There must be a reason for the interesting mix of 50,75 and 100R coax after all.