Lightning Protection For The Flexradio Transceiver

What is needed is an automatic system that grounds the antenna when not in use.

Given: When there are no users on, the antenna needs to be grounded. A remote coax switch normally grounds antennas when not in use but this installation only has one antenna and no remote coax switch. This project will perform that function.

Certain parts are already in place. A Polyphaser lightning arrestor offers protection but when a lightning event occurs a Polyphaser will let through 1000 volts and destroy equipment. A relay that grounds the antenna when the power is off is already installed but now it must be activated manually over the Web. Here is a white-board sketch proposing an automatic solution. It uses Node Red software.

The Weeds

Node Red

Node Red will listen over the LAN for a user to be logged in and then activate the grounding relay. It sounds complicated but the flow below shows how simple it really is. A Raspberry Pi running Node Red is already on site being employed to log the usage. For this new project all that needs to be done is to add a new flow.

Explanation: On the left is a node that queries the Flexradio for information including any clients who might be logged in. In the center is a “function” node that looks only at the data field for the client’s identification. If the field is empty, null, the function outputs a boolean “false”. If there is a client logged in, the function outputs a boolean “true”. On the right is a node labeled “PIN: 16”. PIN 16 is a digital output pin on the Raspberry Pi where a relay module will be connected. The contacts on the relay control the existing antenna grounding relay. The debug nodes above are used only for testing.

Here is a look at each of the configurations of the nodes mentioned above. First, the “flexradio-discovery” node. There is very little here except the port number for the radio. Written by Steven Houser, the Flexradio nodes are quite easy to use, with default settings of “automatic”. The node will find the radio on the LAN and use port 4992 to poll the data.

Below is the function node. The function node contains a Javascript Function. In the first line the command examines the data “client_stations”. If the data is not empty, that is, does not equal a null, it returns a payload of true. Else it returns a payload of false.

Note: How was this function code written? With the help of the new tool, chatGPT, of course. (Yes, chatGPT does more than write essays) Here’s how:

Finally, the output node “PIN 16” tells the Raspberry Pi to make pin 16 (GPIO 23) high or low. The Raspberry Pi nodes are easy, like the Flexradio nodes. Just click the radio button for the pin to use. It even has a sketch of the pin layout that matches the layout of the Pi itself.

Hardware

This is a Raspberry Pi with the relay module attached, ready to be moved to the remote site.

Close up of the relay module. It is a little unusual because it is designed for 3.3 volts so it can match what the Raspberry Pi uses on it’s GPIO pins. https://www.amazon.com/gp/product/B07XGZSYJV/ref=ppx_od_dt_b_asin_title_s00?ie=UTF8&psc=1

Status: Currently the Node Red flow has been tested in the lab and seems to work without a hitch. The flow has been deployed into the existing Raspberry Pi at the remote location. Next, a site visit to install the 3.3 volt relay between the Pi and the antenna grounding relay, is scheduled for Monday, May 8, 2023. Update: …….Installed at the remote site, complete with heat shrink to protect the relay module.

The existing grounding relay, now activated automatically, is mounted on the single-point-ground-panel where the cables enter the building:

A week with several thunderstorms in May and no lightning damage. Of course, the big improvement is, no one had to worry about manually activating the grounding relay.

Concerns

Node Red might suffer a crash which would prevent users from having the use of an antenna. The current Raspberry Pi has been running 4 months without failure, logging away with Node Red. And there is a backup copy of the Node Red flow that could be used to restore this Pi or a replacement Pi if needed. Hopefully the odds are in our favor on this one, too.

New Thoughts On Grounding

Specifically, grounding for the DC circuits, and why is equipment being damaged?

It’s been 6 years since the last article on this blog about grounding. No damage from lightning (knock on wood) all these thunderstorm seasons later, is a good testimonial to the lightning ground. However, lately a few devices have been destroyed or damaged not by lightning but by something else. Considering there have been no thunderstorms the damage probably has been caused by voltage spikes from somewhere. But from where? A ground loop? A relay with no diode? One inadvertent source of a ground loop has been discovered. The negative side of the DC power bus has been connected to the Single Point of Ground since 2017. No noticeable problems have been observed due to this, probably because of luck.

Dereck Campbell, KF5LGW, has a good explanation of why connecting the two together is a bad idea. Note that Dereck is talking about an AC to DC power supply and the remote system has only a DC battery supply. Read around the references to AC line in the explanation below, please. Pay attention mostly to the right hand side of Dereck’s drawings.

The objective is to isolate AC and DC power systems and remove your radio equipment from the ground loop by relocating the ACEG to the same point the shack uses. To understand what is happening, let us draw the circuit out and see what is happening.

FIG 3A shows how 2-wire and 3-wire systems are incompatible. Look closely at the 2- wire system chassis (ground) and DC Negative circuit conductor and the same conductors bonded together inside your 12-volt radio equipment. That is not compatible with 3-wire systems. As you can see the red arrows result forcing normal DC operating current on ground conductors.

The result is DC flowing on equipment grounds, coax shields, and GES. Look what happens when I induce the same AC Line-to-Ground fault inside the DC power supply in FIG 3B. Follow the fault current paths like before. One path is on the ACEG conductor like you planned, and the other unplanned path goes through your radio equipment needlessly and can cause significant damage to your radio equipment and coaxes.

Take note. The 14 AWG ACEG conductor from the breaker panel to the wall outlet can be 30 to 100 feet in length. The unintended secondary path created is through your radio equipment using a 6 AWG conductor of approximately the same distance going back to the same point as the 14 AWG. They are in parallel. Which of those two parallel paths do you think will carry the bulk of the fault current? The 6 AWG or 14 AWG? All you need to know is how Ohm’s law works; no math required. Isolate the 2- systems, and you eliminate the problem!

Lost in a ground-loop creates two more issues. One a minor annoyance generating RFI/EMI. You have a piece of wire (you radio ) bonding the two ground electrodes together with a common-mode current flowing through your EGP. The second problem can be extremely hazardous if lightning strikes nearby. Equalization current path is right through your equipment, acting as a piece of wire.

Now, look at Fig 4A and Fig 4B. I removed the bonding jumper hiding inside the DC power supply. Follow the red arrow current again. What happened? AC and DC systems are isolated. No DC is flowing on any ground conductors. Removing the bonding jumper broke the DC galvanic bond across the transformer inside your DC Power Supply. AC and DC systems are isolated, allowing you to interface your home 3-wire system to your radio’s 2-wire system. The DC system can be a Grounded System, but do not use your DC power supply chassis ground because the transformer’s primary side is part of the AC system, including the chassis.

Why Is Equipment Being Damaged?

One possible explanation is the DC negative bus being grounded but that hasn’t caused known issues for many years. Could the damage be from “back-emf” from an inductance? When a relay is activated and then eventually deactivated there is a voltage spike from the coil which is back-emf caused by the inductance of the coil. A diode across the coil will short-out that back-emf voltage spike. There are several high amperage relays in the remote station that have large coils and large voltage spikes. Those relays all have diodes installed and it is unlikely those relays are the cause of the damaged equipment. What has changed lately is the West Mountain Radio 4005i power controllers have been replaced by a KMTronic 8 relay unit and a DC power strip. In the past, each device was powered individually by a port on the 4005i’s. Instead, now several devices are lumped together and get their power from the DC power strip through a relay on the KMTronic. Power to that strip is turned on and off by the KMTronic unit. Some of the devices are in the shack and others are out at the end of long runs of power cable. In-the-shack devices are in parallel with those long cables on the DC power strip. One tuner is 100 feet away and one security camera is also 100 feet away. That long cable may have enough inductance to induce a back-emf voltage spike when the KMTronic deactivates it’s relay. The plan is to add 1N4001 diodes across the long cables and then to add 14 volt Zener diodes across the inside-the-shack devices. It is hoped the diodes will provide protection from spikes and stop the damage to equipment. Or isolating the DC negative leads from the green-wire ground bus (single point of ground) will. Updates to come.