Extra Virgin Can 70 - cavity filter for 70cm :)

Vodafone build new GSM site here at the roof in my job. There are three antenna systems for 800 and 900 MHz and this killed my experiments with LoRa at 433 and 868 MHz. I have found very nice cavity filter for 868 MHz at Aliexpress which works great but I need something also for 433 MHz. 

I found 5 l can in my stock with drilled holes and N connectors. Marek OK2DL drilled it many years ago when I ordered these cans for 50 Ohm oil dummy loads :) I have decided to try it. 3D printed part helped with mechanical stability and works also as a nut holder. Tunning screw is metal, what is a bad idea but I do not have better now... The result is not so bad! I focused to very low Insertion Loss where the result is lower stopband and also higher BW. There, the IL is around 0,7 dB and 3 dB BW 9MHz. By the moving the internal loops from the tunning screw you are able to obtain better stopband but higher IL.

RX ANT Switching box, part 5 - Assembling and measuring

   I have already finished new RX antenna switch for OL7M. You can see parts and measurements in old articles - here. This box is modular one, so you can build configuration as you need. The main idea is to allow the users to use one or independant two RX per band. It means that you can share every antenna input up to two independant RX per band. With no destroying effect. (For 2 RX you lost abt 3,5dB)

   Each antenna input goes to splitter board. There are bypass relays - if you use just single RX on the band, you can go around the splitter. For two RX, signal goes to splitter. Splitter must be isolated by LC band pass filter. If you do not use it, than you destroy impedance for all rest bands.

   Than the signal goes to the high isolation antenna switch. If you use it together with the splitter, than there must be the load resistors : 50 or 75 ohms. These resistors load unused ports of the splitters. There are two relays in series because of higher isolation. Isolation here is more than 90dB on 40m ( I am not able to measure it ).

   And finally signal goes to the single band PCB with: BPF + notch filters + ATT + HIGH IP3 preamp + simple BPF and some protections... more

First ideas ...

Maybe it should be bigger :D

All parts inside!

DONE :)

And some results:

160m Single A output

160m Output A - 2 RX outputs

160m Output - 2 RX Preamp ON

80m Single A output:

80m Single A + ATT ( 6dB )

80m Single A + preamp

80m Output B - 2 RX

80m Output B - 2 RX + preamp

40m Single A output

40m Single A output + preamp

40m output B - 2 RX + preamp

Common-mode current choke for the RX cables

   I am rebuilding all the RX low band antennas switching and filtering at OL7M. As you can find in a lot of literature - LBDX etc - it is really a good idea to do maximum for suppression of common-mode signals induced on the outside of the feed line:

source: ON4UN's LBDX

   Than I tried to use our ferrite cores. You can find it there: FT-25M533-CTY. So I used a good quality coax cable RG316 and wind it to 4 pieces of that core. As you can see on the pictures:

 There are the cores in original plastic cover - see the measurements

 Example of use in Bi-directiona beverage feeder - see www.RemoteQTH.com

Common-mode attenuation for 4 cores and 5 turns of RG316

 Common-mode attenuation for 4 cores and 8 turns of RG316

 Common-mode attenuation for 4 cores in the plastic cover and 7 turns of RG316

   And the results? These cores are really suitable for the common-mode chokes! And 4 cores and 5 to 8 turns can by use in really wide frequency range: 160 to 6m. The 8 turns is great on 160m to 20m and fine up to 6m. Less turns improve parameters on higher frequencies.

   If you are interested in that, you can order the cores on the RemoteQTH.com web page.

UPDATE:

Kit or assembled CMCC are available at QRO.cz