TX/RX switch or circulator info?

[chrisjmeagher]

Hi Everyone,
Sorry if this topic has already been discussed on the forum. I tried searching for "tx/rx switch" and got a lot of unrelated hits and the single "circulator" hit turned out to be a different topic.

I'm working on a specialized antenna to be hooked up to either the SR5 or the XR5 and currently have two antennas, one for TX and one for RX (each have amps). I'm splitting the single line from the XR5 using a circulator.

In order to put the two antennas together, it would be best if I could know when the XR5 is transmitting. That way I can put my own TX/RX switches in and just choose between the LNA or the PA bank on my antenna. I can't really get away with just adding another circulator because the isolation is too low.

Any ideas on how I can get either an electrical signal or a very fast, low-level software signal that the XR5 is transmitting? I'm basically trying to undo the TX/RX combining that most COTS radios do.

- Chris

[WHT]

As I understand it, you have what is called a "duplexer" for your two antennas (a more accurate term than "circulator"). While it is an antenna splitter of sorts, it uses cavities to split the signal to reduce losses of sending the signal into the wrong component, i.e. the transmit frequency is router out to the antenna, not into the receiver.

Lets start with what you call COTS or commercial off the shelf radios (a concept that is a spin off of NASA's "worker harder and cheaper" idea where military and government departments used commercially available radios instead of custom designed ones - the $435 hammer and $640 toilet seat).

There is no comparison here. Commercial two-way radios use basically a slow speed "push-to-talk" operation, not a high speed time domain diversity switching operation like WiFi radios.

Any suggestions would be most likely way beyond the support of the UBNT forum. Are you prepared to delve into the tx/rx transition gap (TTG) algorithm for proper guard time ramp down delay.

[chrisjmeagher]

I didn't realize that circulator wasn't accurate enough. I've heard them used interchangeably. What I'm using now is a ferrite, three-terminal device that sends power only in one direction (ie, port 1 to 2, 2 to 3, and 3 to 1). We're using it with the SR5 and XR5, so the TX and RX are at the same frequencies.

As for the TX/RX transition gap, I'm expecting that any circuitry or switches that I tie to the TX/RX control will be fast enough to respect the guard times. For example, a lot of the switches I've used work on the nanosecond scale, so that should be okay.

I was just hoping that someone here has already tried to "hack" into the hardware to address as similar problem. For the most part, everything about the COTS radio is fine, except I'm trying to undo the nicety that was added to make them single antenna port radios (or learn when the TX/RX switch is thrown).

Edit: I saw some diagrams of other wifi radios where I guess they use a coupler to combine the TX and RX chains to a single antenna. In that case, I wouldn't be able to borrow off of any TX/RX switch control since there is no onboard TX/RX switch.

- Chris

[Dzintars]

You mean something like MASWSS0129
http://www.macom.com/DataSheets/MASWSS0129.pdf

[WHT]

Keep in mind the MASWSS0129 is a sequential switching switch, not a full path duplexer where the signals are passing in all appropriate directions all the time.