1
$\begingroup$

I need a 3-port RF switch illustrated below. The idea is to use this switch with a sequencer to enable half-duplex communication on a UHF antenna. The switch has three ports: port A connected to the HPA, port B to the receiver, and port C to the LNA. Ideally, the needed connections are A->C (transmit path) and C->B (receive path). However, signals on the A->C can leak into B, something that I don't want. Do switches with these requirements exist? enter image description here

$\endgroup$

3 Answers 3

2
$\begingroup$

Your diagram is wrong - you need the LNA in the Receive path, switched out of the way when you're transmitting. (Unless the LNA has its own internal auto bybass). So the Tx/Rx switch is the first thing connected to the antenna.
For best receive sensitivity, EME hams use two coax runs. With the Switch and LNA mounted at the feed, the thick Tx coax from the PA to the switch, and then a thinner Rx cable from the LNA to the receiver.

Then you need a regular SPDT RF Relay, with the Antenna connected to the common port and the Tx and Rx to the selected ports.
You want what's called an absorptive switch which terminates the unused input - these have a second mechanism that connects the unused port to a load internally, this reduces the coupling between ports and ensures the LNA doesn't start oscillating or anything awful like that.

SMA relays typically have a power handling of up to hundreds of watts (depending on frequency, no hot-switching) and an isolation of >> 60 dB which should be enough to protect your LNA. Here are some from JFW, mini-circuits.

Teledyne was my go-to name but you know it will be expensive when the specifications include 500 G shock etc, I think they were a few thousand dollars.
But there are so many out there on ebay and at ham events. Try to find an older, solid case, made-in-USA/Germany/Taiwan model, rather than a brand new cheapie with the coil hanging out of one side. The quality of the plating is usually a clue, it should be perfectly smooth, slightly matt and slightly less bright than silver - the colour of solder, not chrome. (even if it's dirty). Thin cheap chrome or nickel plating looks perfectly shiny but is a little rough and wrinkly.

$\endgroup$
5
  • $\begingroup$ Brilliant. Thanks. I’m thinking of controlling the switch using a device that can be controlled by sending TCP/IP packets. I was thinking of perhaps using a PLC. Are there alternative solutions? TCP/IP is a must. $\endgroup$ Jun 8 at 7:32
  • 1
    $\begingroup$ An ESP8266 relay board running the Tasmota firmware is my current favourite if you want Wifi. A raspberry Pi or Arduino might work if you need physical ethernet. In all cases you need good sofware interlocks to prevent transmitting into the relay termination and burning it out. Also the ethernet cable and gadget will create RF noise which is very difficult to shield, so it would be best to keep it indoors with the radios and run only a relay power cable to the antenna feed. $\endgroup$
    – tomnexus
    Jun 8 at 13:51
  • $\begingroup$ Another quick question: I will be using N-type cables and was looking for absorptive switches with N-type connectors. Will this one do the work? digikey.se/sv/products/detail/radiall-usa,-inc./R570012000/… $\endgroup$ Jun 28 at 13:13
  • $\begingroup$ It looks OK, but you might limit your options and increase cost if you insist on N type. SMA is more common so consider them too. You just need the correct jumper cables from N to SMA in the box. $\endgroup$
    – tomnexus
    Jul 2 at 13:43
  • $\begingroup$ I just realized the mini-circuit switch has a 20W maximum power. I suppose it won't be able to work with my Kuhne 60W amplifiers. $\endgroup$ Jul 20 at 17:42
2
$\begingroup$

What you want is for path B to be terminated — to have a dummy load attached — while the switch is in transmit mode, rather than being an open circuit that can act as an antenna for high-impedance signal leakage.

I'm not greatly familiar with the terminology for RF switch products, but a quick bit of browsing suggests that you want a “transfer” switch. There doesn't seem to be complete agreement on what this means, but in the type I'm thinking of there are four ports, two pairs of which are connected, and activating the switch causes the other two pairs to be connected instead. In that way, you switch between the two configurations:

  • TX (A) to antenna (C) and RX (B) to terminator (D)
  • TX (A) to terminator (D) and RX (B) to antenna (C)

Another possibility is to use a conventional (as in, what non-RF electronics means by the name) DPDT switch (6 ports) or two SPDT switches (two of the type you already have in mind), such that the receiver can have a termination switched in. Using two switches could also allow the receiver switching to be physically separated from the transmit signal path, which might reduce leakage.


Note that even with the best possible switching setup, you may also experience transmit-to-receive leakage that is not through the switch or the feed line at all, but from the transmitting antenna to the receiver around the feed line (i.e. through the receiver's enclosure and shielding), or even out of the enclosure of the transmitter and in to the enclosure of the receiver.

$\endgroup$
1
  • $\begingroup$ +1 for mentioning floating connections. $\endgroup$
    – user21417
    Jun 7 at 15:28
0
$\begingroup$

I think what you want is an RF circulator.

Circulators help direct the flow of microwave or radio-frequency signals in RF systems. Usually 3 or 4-port device that transmits input signal from 1 port to the next, Figure 1 shows a normal circulator symbol. The ports are either input signal, output signal or termination and typically have a line impedance of 50Ω. Common connectors are SMA or N-type coaxial. The signal will transfer from port 1 to 2, 2 to 3, and 3 to 1. The ports are connected to a symmetrical Y junction that is coupled with magnetically biased ferrite material. Ferrite along with a magnet creates a flow that helps push the signal in the circular path we see and ensure the port-to-port assignments are followed. Circulators are designed to have minimal loss when transmitting an input signal from one port to the next. Figure 1

In your example, A is port 1, C is port 2, B is port 3. A->C, C->B.

$\endgroup$
1
  • 2
    $\begingroup$ Hello Steve and welcome! I hope you'll stick around. The problem with a circulator in this application is that any reflected power from the antenna goes straight into the receiver. Perhaps it could get you the first 20 dB of isolation, but after that you still need a switch or relay to protect the receiver. $\endgroup$
    – tomnexus
    Jun 7 at 4:58

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .