RF in the Audio: Curing RF Feedback on Your SSB Transmission

You key the microphone, and a friendly voice on the other end reports that your audio sounds harsh, buzzy or distorted, or the transmitter squeals on every word. Yet when you listen to the monitor, everything is fine. The cause is often RF in the audio: radio-frequency energy from your own transmitter has found its way back into the microphone circuit, where it is rectified and amplified together with your voice. The result is a signal that is dirty in a way that no equaliser can fix. The good news is that the problem has a short list of causes and a short list of cures, and you can diagnose it in an afternoon with a dummy load and a few ferrite cores. This guide explains the mechanism, shows how to test, and gives the fixes in the order that works best.

What Is Happening Inside the Circuit

A microphone input is a very sensitive amplifier. It is designed to handle a few millivolts, so even a small amount of RF voltage on the microphone lead is enormous in comparison. The input transistor or operational amplifier contains semiconductor junctions that behave like diodes. When RF reaches them, they rectify it: the junction conducts on one half of the RF cycle and not on the other, and the slow envelope of the RF appears as an audio voltage. This effect is known as audio rectification, and it is the same mechanism by which a nearby transmitter can be heard in a hi-fi amplifier, as described in the article on electromagnetic interference.

On a single-sideband transmitter the rectified envelope of your own signal is fed back into the audio chain. Because the audio path controls the RF output, a loop is formed: the audio makes RF, the RF returns as audio and the loop adds distortion, harsh peaks or a squeal. When the loop gain approaches one, the transmitter oscillates and you hear a howl, which is called RF feedback.

The Usual Path: Common-Mode Current

RF rarely reaches the microphone by radiating directly into the small wires. The usual path is common-mode current on the outside of cables. When an antenna is not perfectly balanced, or when the feedline is a convenient length for the wavelength, part of the transmitter current flows on the outer surface of the coax shield instead of staying inside. That current spreads to everything attached to the radio: the power lead, the microphone cable, the USB cable and the metal case. The microphone cable then acts as a small antenna and brings RF to the microphone input.

Several conditions make it worse:

• An unbalanced antenna fed with coax, such as an end-fed wire, an off-centre dipole or a vertical with few radials. The antenna pushes current onto the feedline.
• A shack close to the antenna, with the antenna or feedline running within a few metres of the operating position.
• High power, because the RF field grows with power, and the audio distortion appears only above a threshold.
• Cable lengths near a quarter wavelength or a half wavelength on a given band, which are resonant and pick up the most RF. This is why the trouble often appears on only one or two bands.
• A poor earth, so that the chassis of the radio sits at an RF voltage above the earth.

The article on common mode chokes describes the physics of the current on the shield and the basic remedy. For RF in the audio the chokes are the first and most effective cure.

Recognising the Symptoms

RF feedback has several characteristic signs, and recognising them helps to separate it from simple overdrive or a faulty microphone.

The last row is a useful reminder: not every harsh signal is RF in the audio. Overdriven audio, too much compression and a clipped microphone preamp create the same effect, and the article on SSB audio settings shows how to set the levels so that the ALC only touches the peaks.

Step One: Test With a Dummy Load

The quickest way to find out whether RF is involved is to remove the antenna from the picture. Connect a dummy load to the transmitter output, as explained in the dummy load guide, and transmit at the power where the trouble appears, using a second receiver or an SDR to listen to the audio. With the dummy load, the antenna does not radiate and the feedline does not carry the current of an unbalanced antenna. If the problem disappears, the antenna system is the source, and the chokes and the antenna balance are the cure. If the problem is still there, RF is probably entering from the radio itself or from the cables near the radio, or the cause is not RF at all.

Then run a second test. Reduce the power in steps and note the power at which the distortion starts. A threshold effect, clean at 20 watts and distorted at 60, is typical for RF feedback, while overdriven audio is distorted at any power. Finally, change the band. A fault that follows one band points to a resonant cable or antenna length.

Step Two: Chokes on the Coax

If the dummy load test shows that the antenna system is involved, place a common-mode choke on the coax where it enters the shack. A choke is a number of turns of the coax through a ferrite core, or a string of ferrite beads, and it adds a high impedance in the path of the common-mode current without affecting the signal inside the cable. For HF, the core material matters; mix 31 gives high impedance at 1 to 30 MHz, and the article on the ferrite mix guide tells you which to choose. A second choke at the antenna feedpoint, preferably a 1:1 current balun, stops the current at its source. Both together are better than either alone.

The number of turns matters more than the size of the core. A choke made of six to ten turns of coax on a large toroid of the right mix presents some thousands of ohms at the lower HF bands. A single pass through a snap-on core gives far less. If the problem is still present on a lower band, add turns. Common-mode chokes on the coax are the cure with the largest effect in most stations.

Step Three: Treat the Microphone and Control Cables

After the coax, treat the cables that connect to the radio. The microphone cable is the first target, because it leads straight to the sensitive input. Put one or two ferrite cores on it, close to the radio connector and near the microphone, and wind as many turns of cable as fit, three to five, on a mix 43 or 31 core. A snap-on core can be opened and closed around a cable that has a connector, which is convenient. Do the same with the USB cable, the CAT cable and the power lead, because each of them can carry RF to the radio or to the computer. A guide to the wiring of interfaces is the article on the digital interface, where ferrite cores on USB cables are a standard part of a quiet setup.

If the problem persists, bypass capacitors can help, but they must be small. A capacitor of about 100 pF to 1 nF from the microphone line to ground, placed as near to the connector as possible, shunts RF while leaving voice frequencies untouched, since the impedance of 1 nF at 3 kHz is still more than 50 kilohms. Use ceramic capacitors with short leads. Do not use a large value, which would cut the high frequencies of speech, and take care to follow the circuit diagram of your microphone or radio, because some inputs are biased for electret capsules. Capacitors are a fine-tuning step, and the ferrite chokes should be in place first.

Step Four: Shield, Earth and Layout

With the chokes in place, look at the shield connection of the microphone cable. Some microphones with an XLR connector connect the shield to pin 1 inside the connector instead of to the chassis, which allows RF on the shield to enter through the internal wiring. This is known as the pin-1 problem. A shield should be connected to the metal body of the connector, where the RF current can leave without passing through the amplifier. If you build a cable, solder the shield to the connector shell and use a shielded cable of good quality.

An RF earth can help but it is not a cure by itself. A short, thick connection from the chassis of the radio to a ground rod or a ground plane lowers the RF voltage on the case, but a long earth wire on 20 metres can itself act as a quarter-wave radiator. The aim is a low-impedance path, so the connection should be as short as possible. In the layout of the shack, keep the coax and the microphone cable apart, do not run them in parallel for long distances, and keep the antenna and the feedline away from the operating position. The article about shack layout gives practical advice.

Antenna Balance and Power Level

If the stubborn cases remain, the antenna itself is part of the problem. A dipole fed with a balun and with symmetrical legs is easy on the shack. An end-fed half-wave wire or a vertical without a proper ground puts a large current on the outside of the coax. Give such antennas a counterpoise of suitable length, or place the feedpoint choke where the coax leaves the antenna. In the last resort, reduce the power by a few decibels, since a drop from 100 to 50 watts often moves the system below the threshold, and the signal on the other end sounds better than at the higher power with distortion.

Remember the safety dimension as well. RF on the case or the microphone is the same RF that can cause burns when you touch metal parts during transmission. Cures for RF feedback also reduce this risk.

Checking the Result

After each change, test again in the same way: same band, same power, same microphone. Ask a local friend for a report, or record your signal with an SDR a few metres away and listen to the file. Change one thing at a time. If you add three ferrite cores, a capacitor and a new earth at once, you will not know what worked, and you may leave unneeded parts in the cable. When the problem is gone, test on other bands, because a cure for 20 metres may leave 15 metres untouched.

The Bottom Line

RF in the audio is the rectification of your own transmitter’s RF in the microphone input, usually brought by common-mode current on the outside of the coax and other cables. Find out with a dummy load test whether the antenna system is involved, then fit a common-mode choke on the coax at the shack entry and at the feedpoint, put ferrite cores on the microphone, USB and power cables, and add a small bypass capacitor only as a last step. Check the shield connection of the microphone, keep the cabling short and tidy and balance the antenna. Make one change at a time and test on several bands. With these steps your signal returns to the clean audio you hear in the monitor.

Frequently Asked Questions

How can I tell RF feedback from overdriven audio?

Reduce the power and test with a dummy load. RF feedback depends on power and on the antenna system. Overdriven audio is distorted at any power and any antenna, and it follows the microphone gain and compression settings.

Will a ferrite core on the microphone cable fix it?

Often, but not always. It helps when RF enters through the cable. If the main current flows on the coax, chokes there are needed first, and the core on the microphone cable is a complement.

Why does it only happen on one band?

A cable or feedline close to a resonant length on that band picks up the most RF. The same system may be quiet on the other bands.

Can a bypass capacitor on the microphone input cause problems?

A value of 100 pF to 1 nF is usually harmless for speech. A larger value can cut the high frequencies of your voice, so start small.

Does a better ground cure RF feedback?

It can reduce it, but a long ground wire may add RF at some frequencies. Chokes on the coax and cables are more reliable than a ground alone.

Can a headset help?

A headset with a short, well-shielded cable and a ferrite core is less sensitive than a desk microphone on a long cable, and many operators solve the problem by changing the microphone type.