Yes, the ubiquitous 4 way antenna switch normally bundled together with a 4 way switch….. but we don’t want to give you that !!! 

   I’ve seen this design on the Internet and its bazillion variations, some are plain rip-offs and other are quite enterprising but none seem to have the best of all worlds. I hope to alter that here with my “version” which is basically tinkered together ideas under one roof.

   Firstly, the design incorporates a Bias T. This enables the remote switch to be powered over a single coax and no extra “and expensive” cables needed. That then poses a problem? How does one go bout switching the device? One uses ones own WiFi network and a ESP32- Wroom 32U (the external antenna version). The idea is basic but the code naturally has a few gimmicks to separate it from other designs. The power on sequence slowly checks each relay for activity and then switches to the next and repeats this 10 times, then it passes a second time, going between each relay faster 10 times. We get a visual indication of the switching and satisfy ourselves the switch is on and receiving code from the ESP32 Wroom. When the forst power on after uploading the sketch we are presented with a prompt to use out PC or mobile phone to look for G5STO_Switch in our WiFi network list. Once connected we can then go to our web browser and open 192.168.4.1 in the URL bar. Then click on our SSID and input the password. Once that is done then the password is stored in the EEPROM and the the device reboots. Now you will need to find the device IP on your network but this can be easily done with the ESP32-Wroom plugged into the Arduino IDE or Putty and examine the output. It should tell you what IP in the Serial Monitor field what IP address has been assigned to the device.

We then return to out browser and input the new IP address in our URL and then we are presented with the GUI for the antenna switch. The IP address should remain with the device until flashed with an update or your Routers DHCP alters it but you can obviously enter your router and use the LAN IP function to reserve the IP address via MAC address from changes and assign it its own IP.

 Now that you have accomplished this menial task you can go ahead and select buttons to control each antenna port. 

   I have not presented the code just yet as I have some testing to do on the switch and its impedance and return loss etc but quietly confident its minimal.

Lastly, the switch is unique in the way the relays are driven. The Bias T at the headend has a relay driver IC. This then switches the power to the relevant relay, a bit like a common anode with each relay path brought to ground. This is OK for WiFi but alas it will not work for a manual switch. So the design benefits from the upper and lower planes acting as with a ground of 12v source to maintain a common anode or common cathode. In Wifi mode we are naturally focused on a common anode. Shorting the upper and low planes allows us to use a normal powered switch with the relays all now shorted to ground and power applied to the other terminal. 

  Ok that works in principle and why the blocking components must be reversed when using a manual switch. eg. snubber diodes are now facing the opposite direction as will the LED’s need to be to illuminate. All other components are non polarised and will work the same.

 

Final comment……the code also features a robust interlock mechanism with a break before make structure. This prevents transmission into a direct short on infinite VSWR and also helps to prevent two antennas being engaged at once and also giving unwanted high VSWR. In truth the design could be improved upon with locking relays that remain in the state after power off rather than all relays going open circuit. In essence I should have added a 5th relay which would be N/C or normally closed with all other relays N/O normally open. A small, dummy load could be placed inside the remote box and save the radio from a bad situation should the remote power be lost or the ESP32-Wroom fail or crash. In that situation a small 50w dummy load made up of non inductive carbon resistors or even a thick film resistor rated at 50 ohm and 100w would work well. This naturally adds to the expense, space as we now require a substantial heat sink. V2 of this design will do just that and I cant wait until someone rips me off!!!! as the Aluminium casting will be machined with an incorporated heatsink acting as not only a weather and RF shield but dissipate heat too.