Combining two RF systems

[Tudor]

Hi.

I have two separate working systems as you can see in the pic. I need to use them into the same case. That mean the antennas (directional type) will be pretty close (like no more than 1 meter distance). Obviously, there will be a lot of RF interferences between antennas, which will make the systems working improperly.
The only differences between that two systems is that they are using different channels (ARFCN) , but on the same frequency range (890-960 MHz). This mean different physical radio carriers and channels used for transmission and reception.

What should I do? I need to use a combiner? If so, where I need to put it, regarding the RF chain? Please note that I don't mind to use two antennas. The only think I need is to avoid RF interferences between antennas.

Thanks a lot for helping me.

 

[mmckenna]

A receive multicoupler after your bandpass filter. The BPF should be wide enough to not be an issue.

You'll need a two port transmit combiner tuned to your specific frequencies.

This isn't any different than a 900MHz trunked system, same equipment would apply.

 

[kayn1n32008]

Wirelessly posted (Mozilla/5.0 (BlackBerry; U; BlackBerry 9780; en-US) AppleWebKit/534.8+ (KHTML, like Gecko) Version/6.0.0.600 Mobile Safari/534.8+)

If the transmit and receive frequencies are fixed, and you do not need to point the RF in two different directions then a receive multi-coupler and a transmit combiner would eliminate the second antenna. If you need to point the RF in different directions then corner reflectors may also help with RF isolation between the two systems. Do remember that the panel antennas are fairly direction on their own.

 

[ramal121]

What is the freq spread between the two transmits and two receives?

Usually a receive multicoupler and a transmit combiner each have their own separate antennas. Switching two duplex devices to a combining system doesn't eliminate the second antenna and doesn't buy you anything. This does not mean a single antenna is out of the question, it just takes a little forethought.

Lets say the freq spread between receives and between transmits is close, say a couple of hundred kilohertz. You could use a single duplexer (assume a tight band pass) off of a single antenna. Receive no big deal, just feed the duplexer to a multicoupler then split to your receivers via amps and band pass filters (although I don't think they're really needed). Transmit at a tight spacing is a little trickier. Passive transmit combiners have a minimum separation between freqs. If you can't meet the required separation then you'll have to consider a hybrid directional combiner. They work OK but are not very efficient as part of your transmit power goes to heating the other radios dummy load.

Now on the second hand, if you have a good freq spread, you can use a multicoupler and a passive combiner but a regular duplexer will not be broad enough in the pass skirts to be of any good. You'll need to have custom window filters built and hooked to a common output instead of a duplexer.

If you are going to spend the $$$ to assemble a scheme, all the popular manufacturers will engineer something that works. If you are going to DYI, well good luck and god speed.

 

[prcguy]

I would recommend placing the LNA between the BPF and radio, otherwise it could get some transmit noise and cause problems depending on how good the diplexer is.
prcguy

 

[zz0468]

As is usual when a student engineer comes here asking for solutions to a homework problem, you fail to provide enough information to solve the problem. Are seperate antennas necessary? Are the two sets of frequencies paired similarly, i.e. both tx high and both rx low? Are these devices frequency agile?

Here's my thoughts.

1. Preamps ALWAYS after any filters.
2. Ferrite isolators and a filter on both transmitter outputs.
3. Combine filtered and isolated tx outputs to one transmission line.
4. Receiver bandpass filter to a single preamp, to a splitter feeding each receiver.

At this point, you've saved the cost of a preamp. You could...

1. Run separate antennas, one for tx, one for rx. Or
2. Duplex with a bandpass duplexer and save yourself an antenna, or depending ong powere level, use a circulator branchong network.

 

[zz0468]

After re-reading my reply, I can say with great confidence that I clearly wasn't all there last night! I hate when I do that. Other than acknowledging that some of the missing data I grumbled about is actually there, I would add that you could hybrid combine the transmitters then apply the combined transmitters to the duplexer, and multicouple the receive side.

 

[prcguy]

I agree, that is the standard way to combine two transmit systems that are nearly on the same frequencies but with some transmit to receive frequency separation.
prcguy

After re-reading my reply, I can say with great confidence that I clearly wasn't all there last night! I hate when I do that. Other than acknowledging that some of the missing data I grumbled about is actually there, I would add that you could hybrid combine the transmitters then apply the combined transmitters to the duplexer, and multicouple the receive side.

 

[Tudor]

Re:

Hi guys.

Thank you very much for your replies.
I know that the title is a bit confusing. Instead of "combining" I should use "working close to each other". My apologies.

Pleas let me be more specific:

I am pretty happy by using two separate antennas, one Rx/Tx antenna for every system. This is not a problem at all.

I am concerned about RF interferences that may occur: using that antennas pretty close (less than one meter), I can encounter some troubles or even damages on Rx side, due to crosstalk. The Tx power is 4 to 10 W.
In fact, that two Rx/Tx antennas are working on the same frequencies (890-960 MHz), but on separate channels (ARFCN's) on Tx side. The channel spacing on Tx is 200 kHz (at least). The Rx-Tx spacing is 45 - 50 MHz, and by using that duplexer and BPF, there is a good separation between them. I am using only good quality components: Kathrein duplexer (commercial grade), Kathrein BPF (same commercial grade), one Minicircuits LNA.

So, do you think I can use this two antennas close to each other (less than one meter)?

@ zz0468: Thanks for all info you gave me. I am using one system at the time, with that BPF before LNA, exactly as I described on the pic. And its' working. What if I'll put the preamp (LNA) after BPF? More gain or accuracy?

Thank you very much and Happy Holidays to all of you.

 

[zz0468]

I am concerned about RF interferences that may occur: using that antennas pretty close (less than one meter), I can encounter some troubles or even damages on Rx side, due to crosstalk. The Tx power is 4 to 10 W.

If you have enough isolation to duplex TX/RX on one antenna, then you're probably just fine with an identical unit nearby. The adjacent units are transmitting close in frequency, and receiving close in frequency, with identical frequency separation. I don't think you'll have any problem putting the two antennas close together.

@ zz0468: Thanks for all info you gave me. I am using one system at the time, with that BPF before LNA, exactly as I described on the pic. And its' working. What if I'll put the preamp (LNA) after BPF? More gain or accuracy?

It's pretty important to put the LNA after all the filtering. It will protect it from overload, and the overall noise performance of the system is better with more filtering in front of the LNA. The filtering from the duplexer itself is usually not adequate much outside it's operational frequency range.

The cost is that the insertion loss of the filter adds directly to the overall system noise figure, so you want that LNA to have a very low noise figure to start with. It's a small price to pay.