2
\$\begingroup\$

I'm designing a circuit that has 2 antenna options for one transceiver @ 1.6 GHz. One antenna is internal and is the default RF load. Another antenna is external - to be attached by the user if desired. If and when the external antenna is attached, an RF switch will change the path from the internal to the external antenna. Here's a simplistic sketch of what I'm thinking...

sketch of RF switch

I was thinking of using a directional coupler to measure the VSWR on my external antenna line and using the result to control the RF switch automatically.

Does this make sense? I'm not very experienced with RF designs... Are there any do's, don'ts or gotcha's? Best practices?

edits
My average transmitted power is 1.6W during a TX sequence. This is for a hand-held device (walkie talkie size)

\$\endgroup\$
5
  • \$\begingroup\$ If the switch is selecting the internal antenna then measuring the VSWR on the external antenna will be fruitless because there is no power going to it. \$\endgroup\$
    – Andy aka
    Dec 31, 2015 at 11:43
  • \$\begingroup\$ @Andyaka I had thought about this, and was considering a scheme for periodically checking for external antenna presence with an alternate source at very low power. There are still some kinks to work out... Can you suggest a better method for detecting the antenna presence? \$\endgroup\$ Dec 31, 2015 at 14:48
  • \$\begingroup\$ This appears to be something like a SWITCH 0/12v. Right? Switch link \$\endgroup\$
    – AllMac
    Jun 12, 2017 at 19:09
  • \$\begingroup\$ @AllMac It looks similar, but I can't read ukrainian. That looks like it is only for receiving. I needed to transmit. Also, I needed the circuitry to put on my own printed circuit board. \$\endgroup\$ Jun 13, 2017 at 22:54
  • \$\begingroup\$ For anyone who cares, I ended up with a working solution. An RF switch DPST, along with a GND detection circuit was used. When DC GND was detected, some transistors switched the RF switch. I didn't end up needing the VSWR reading. \$\endgroup\$ Jun 13, 2017 at 22:59

1 Answer 1

3
\$\begingroup\$

A few points which might be useful:

1) there are connectors with an RF switch in them, which are designed for exactly your situation. They pass RF to an internal antenna if there's no connector installed.

Of course they can't tell if there's a working antenna, only a connector plugged in. Also they're only available from tiny on-board connectors up to about mcx size, which is still pretty small, not rugged, not designed for many insertions, so this would only be useful for a one-off switching event, say if you had two installation options.

2) you may be able to detect the external antenna by its DC short circuit. Most antennas are DC shorted, it's good practice, and this would simplify your circuit a lot. Just a bias T, a power supply, and the RF relay. Don't forget protection diodes to prevent any spikes from ending up on the RF line. In fact, best to use a transistor and keep the voltage and current on the RF cable as small as possible.

3) detection by looking for return power means the radio will see an open circuit for a while, is this OK? If the wrong antenna is selected, the whole packet will be lost, what is the impact of this.

4) you'll need a supervisory crcuit to check regularly for the presence of an external antenna, every few minutes I suppose. You can only check this when the modem transmits, so you have a fairly tricky operating procedure. Think this through carefully.

5) your RF switch might end up being switched hot - is that a problem? Depending on the power levels that can damage a switch. It would be best for the transceiver itself to detect the antenna and make the switching decision. Not sure if this is possible for you though.

\$\endgroup\$
8
  • \$\begingroup\$ Just curious: would a physical double-throw relay work in this application? (I'm sure this is probably a bad idea, but I don't know much about RF, so I'm curious why it would be bad) \$\endgroup\$ Dec 31, 2015 at 5:26
  • \$\begingroup\$ Oh no, at 1.6 GHz you need a coaxial relay. They're available in all shapes and sizes. For power under 1 W there are also solid state switches, look at Mini-circuits for both. I'll comment on why a bit later. \$\endgroup\$
    – tomnexus
    Dec 31, 2015 at 5:40
  • \$\begingroup\$ @ryang a coaxial relay is a physical relay, but special care is taken to maintain the 50 Ohm impedance through the device, and keep any unconnected stubs nice and short. It will have a short rocker arm choosing between the two output terminals. A regular relay typically has little concern for the impedance of the signal path, and routes it up, through the contacts and back via a flexible cable. pictures here \$\endgroup\$
    – tomnexus
    Dec 31, 2015 at 8:05
  • \$\begingroup\$ If the power level is not too high, then a PIN diode switch may be up to the task: some examples at infineon.com/cms/en/product/rf-and-wireless-control/rf-diode/… \$\endgroup\$ Dec 31, 2015 at 10:02
  • 1
    \$\begingroup\$ Macom (to name but one manufacturer) make many switches good for up to 2W and more. An advantage of these is the infrastructure can be as rugged as you want as the switch is solid state. Selector guide macom.com/products/control-products/switches \$\endgroup\$ Dec 31, 2015 at 15:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.