Home - About AR - Learning Material - Exams - Clubs - Posters

Amateur Radio Info & Exams - Amateur Practices 2 - Interference, DSP, Processors, and S-meters

The energy from amateur transmitters can cause interference to other equipment. This section covers some of the causes and solutions.

Interference to audio gear including telephones.

You may be aware of "crystal radios", which use a diode to rectify an RF signal, and directly drive an earpiece. In some cases, they were also used to drive an audio amplifier. If receiving an AM signal audio of reasonable quality is recovered. If the signal is SSB, then the audio is highly distorted, or Donald Duck like. Through "slope detection", it is possible that audio of moderate quality can be recovered from an FM transmission.

As an aside, while modern crystal sets are a "party trick" to demonstrate that the radio can operate only on energy from the transmitter, without a battery, in reality using a 1.5 volt cell, or something similar to forward bias the diode, and improve sensitivity was far from uncommon. Fox-hole or trench radios used both the battery and receiver "liberated" from a field telephone, as well as a razor blade "blued" with a selenium compound. Selenium is a semiconductor, so use used in both improvised and manufactured diodes.

The problem is that the diode and transistor junctions within an audio amplifier, or various parts in a traditional telephone can rectify the signal, meaning that Amateur radio signals can appear in broadcast radios, analogue TVs, audio amplifiers, and telephones.

This can be mitigated by placing ferrite sleeves or toroids on the input and output leads, placing small, usually ceramic, capacitors across the terminals of audio gear, and perhaps installing series inductors. For TVs high-pass filters can be plugged into the antenna socket before the antenna is connected.

A low-pass filter can be fitted to the Amateur transmitter, but if the problem is a TV with a cheap and nasty receiver section, then it suffers "fundamental overload", no matter how clean the amateur station's signal is.

Other sources

A range of equipment in households and businesses can cause interference to both Amateur Radio, and to things like reception of MW / AM broadcasts, shortwave radio, and other services.

This includes low quality switch-mode power supplies, variable frequency drives for swimming pool filters (sold with the claim they save power), and arcing thermostats switching heating and cooling equipment. Some are constant, some generate a periodic "blatt" or "buzz". Power hungry plasma TVs cause interference. Fly-by-night roof-top solar panel installer's low quality inverters are another source of interference on HF. Good quality units are available - ask nearby fellow hams what they have.

In the days of analogue TV commutation and some other mains related noise could place strips or bands of noise across TV images, as the field rate and mains frequency in many locations were the same or similar, so the peaks would occur in the same place in the screen scan.

Grounding

Many items of station equipment have a stud with a wing-nut, or other grounding point. These allow a braid or strap to be connected, or failing this, a stout wire with crimped eye terminals. Ideally, these should be connected to a common point, such as a strip of copper bus-bar, and this connected via braid to the station earth. This should consist of several ground rods, and should be "bonded" to the mains safety earth.

One problem is that if the strap is a quarter-wavelength long at the frequency on which you are transmitting, this will have a high impedance, meaning that if you touch a metal chassis while operating at high power, you may still get a shock.

Ground Loops

Suppose you have connected the audio connections of your desktop PC, grounded via the mains earth, to your transceiver, grounded via the station ground. The shield on the audio cable is thus connected to both Earths. This can cause induced hum to appear on the audio signal. The examiner suggests that using a single, common earth point is the solution.

One way to avoid hum and this is to place small audio transformers on both audio lines, and use an opto-coupler based circuit on any keying line, or perhaps a Darlington or a FET driving a small relay. (If using a fully tube set, the keying voltage can be quite significant).

Some hams use a "Pi" single board computer for digital, with the option to use a remote connection from elsewhere in your house, or beyond, to control it.

A somewhat counter-intuitive idea I heard recently was that Amateur stations are better to use Wi-Fi rather than blue cable for network connections in your station, if not your entire house. This prevents RF upsetting the network, and noise from the network from causing interference.

SSB and Band Edges

We have discussed the occupied bandwidth of SSB and other signals previously. You will remember that a USB signal will occupy about 3 kHz above the dial frequency on an Amateur transceiver. An LSB signal will occupy around 3 kHz below the dial frequency.

Assuming 3 kHz bandwidth, if we were operating LSB on 7.178 MHz then the spectrum used starts (just below) 7.178 MHz, and extends down to 7.175 MHz.

Conversely, if we were operating USB on 14.347 MHz, our signal would extend from (just above) 14.347 MHz, to 14.350 MHz. In both cases we add or subtract 3 kHz, or 0.003 MHz.

Beyond the exam, for AM we can also get to around 3.00 kHz from the edge of the band, unless your signal is wider, but for FM you need to avoid being within half the bandwidth of the signal from the band edge.

One distractor on two questions has 4 digits after the decimal. This relates to channels named for their centre frequency, placing the edges of the channel 1.5 kHz above and below the nominal frequency. This is NOT Amateur practice, except in part on 60 metres.

If unsure I suppose the trick is to mentally or physically draw a marker with the quoted carrier frequency, add or subtract the bandwidth according to the sideband used, and write this down on the appropriate side of the other marker. Remember that 3 kHz is 0.003 MHz. Then compare this to the answers offered. Likewise, you need to be above a lower segment edge; or below an upper segment or band edge, for the questions asked.

Off the exam, if you are using USB on the lower bands for some reason, you need to keep 3 kHz below the upper band or segment edge. This may be the case if using a type-approved or "commercial" radio, such as some Codan units; or some military or aviation gear.

In the case of the 80 metre band in Australia, the upper edge of the general use band is 3700kHz and you can have this on your display, as long as you are sure you are sure you are using LSB. Flip to USB, and you may cause interference to either MAF, or the military.

Guess-meters

Most transceivers have some form of display of the received signal. In most cases they are quite arbitrary, although, if you are dropping 10 kilobucks+ on a fancy transceiver, then you should expect a properly calibrated S-meter.

For HF the recommended S9 level is -73 dBm for HF, and -97dBm for VHF. 0 dBm is 1 milliwatt. At HF S1 is thus -121 dBm. On HF S9 is 50.2 microvolts into 50 ohms, and at VHF it is 5 μV. The HF S1 is 0.2 μV.

Proper S-meters require an increase of 6 dB to increase the display by a single S-unit. This means, to move from, say S8 to S9 you need 4 times the power, and this may have little benefit in readability. However, most meters are only marked with odd numbers, so S7 to S9 is 12 dB, or 16 times the power, so 100 watts to 1500 watts is just less than 2 points, but risks interference and other problems, not to mention being most pleasing to your money-grabbing electricity retailer.

Beyond S9 decibels are used, say 20 dB over S9, or "20 over 9", or just "20 over". This represents 100 times the signal strength. One might represent a station on a net from a few towns away, the other the other side of the same town. Someone a block away might be "40 over" or "60 over", but remember that they need to operate at a level heard by the bloke who moved a few hundred km away, and might only be a few S points, depending on conditions.

Remember also that this reading is at the radio's input socket, not the antenna itself, due to feedline losses.

However, even a non-calibrated S-meter, or "guess-meter" as they are sometimes termed, can be useful for comparing antennas, either by comparing the strength of a signal from a beacon receiving on different antennas, or by asking a station for signal reports as you transmit on one, switch, then transmit on the other. There is every change that stations at different distances, and/or using a beam or dipole versus a good vertical, will find different antennas to be better.

A good antenna tuner may allow you to switch antennas at the tap of a button.

Clearly good practice when the near neighbours are chatting is to reduce power to that needed to overcome any local noise. Using a quiet corner of a VHF or UHF band, typically using FM or SSB, is a good idea too.

Speech Processors

If you have been involved with things like church PA, with a reasonable system, you may have used a compressor or compressor-limiter to control the level of the lead vocals, and/or the pastor's speaking microphone, to keep the level reasonably consistent. Specialist processors are also used in broadcast audio, to provide "loud" audio, without over-modulating or over-deviating (although a transmitter should mute to avoid this).

Some transceivers include speech processors, or it is possible to add one between the microphone and transceiver. (Be aware of "power microphones" which used to be sold to CBers, they didn't just wind the level up to 11, but more like 15, and then operators wondered why other stations said that they sounded awful). These processors use a range of audio processing techniques to increase the average audio level on transmit, and so to "punch" through noisy conditions better. Again, judicious use of any adjustments is wise.

The downside of these is that they can do things like increase background noise, such as fans, traffic noise, and like; cause distortion; and intermodulation products, causing splatter outside your channel. I am not sure if they include a "noise gate" in the processor, to quieten it when there is no speech, so things like fan noise are not amplified, probably not.

Also, good settings for a club net or "rag-chewing" are different for DXing or contests, natural for nets, punchy for DX dog-piles.

Relevant Questions

These are actual questions from the General exam pool.

G4C01
Which of the following might be useful in reducing RF interference to audio frequency circuits?
A. Bypass inductor
B. Bypass capacitor
C. Forward-biased diode
D. Reverse-biased diode

Small capacitors, such as ceramics, placed across audio terminals will be low impedance to RF signals, so short them to ground; but high impedance to the audio signals, answer B.

G4C02
Which of the following could be a cause of interference covering a wide range of frequencies?
A. Not using a balun or line isolator to feed balanced antennas
B. Lack of rectification of the transmitter's signal in power conductors
C. Arcing at a poor electrical connection
D. Using a balun to feed an unbalanced antenna

An arcing electrical connection, such as in a worn, old thermostat can cause wide-band interference, answer C.

G4C03
What sound is heard from an audio device experiencing RF interference from a single sideband phone
A. A steady hum whenever the transmitter is on the air
B. On-and-off humming or clicking
C. Distorted speech
D. Clearly audible speech

Sideband signal being rectified in a telephone or audio gear causes distorted or "Donald Duck" speech to be heard, answer C.

The distortion is down to the SSB signal lacking a carrier to act as a reference for generating an audio signal of the correct frequency.

G4C04
What sound is heard from an audio device experiencing RF interference from a CW transmitter?
A. On-and-off humming or clicking
B. A CW signal at a nearly pure audio frequency
C. A chirpy CW signal
D. Severely distorted audio

The rectification of the CW signal generates only a DC signal in the device, which is heard as a clicking at it starts and stops (that is, as it is keyed), or as a hum if there is hum on the signal, answer A.

G4C05
What is a possible cause of high voltages that produce RF burns?
A. Flat braid rather than round wire has been used for the ground wire
B. Insulated wire has been used for the ground wire
C. The ground rod is resonant
D. The ground wire has high impedance on that frequency

If the braid or ground wire is a quarter wavelength long (or ¾λ), then it will appear as high impedance, answer D.

G4C06
What is a possible effect of a resonant ground connection?
A. Overheating of ground straps
B. Corrosion of the ground rod
C. High RF voltages on the enclosures of station equipment
D. A ground loop

This can cause a high RF voltages on the cases of equipment in the station, painful shocks, and the emission of words which may violate the regulations, answer C.

G4C07
Why should soldered joints not be used in lightning protection ground connections?
A. A soldered joint will likely be destroyed by the heat of a lightning strike
B. Solder flux will prevent a low conductivity connection
C. Solder has too high a dielectric constant to provide adequate lightning protection
D. All these choices are correct

A lightning strike will cause a very large current to flow in the wires, causing them to be heated, to melting point of the solder, if not vaporisation, answer A.

Further, while the solder is melted, the magnetic effect of the current may cause the wire to be repelled or attracted to other conductors, causing the joint to separate.

G4C08
Which of the following would reduce RF interference caused by common-mode current on an audio cable?
A. Place a ferrite choke on the cable
B. Connect the center conductor to the shield of all cables to short circuit the RFI signal
C. Ground the center conductor of the audio cable causing the interference
D. Add an additional insulating jacket to the cable

A split ferrite sleeve clipped around the cable, or loop several turns of the cable through a toroid can reduce the flow of RF currents in an audio cable, answer A.

G4C09
How can the effects of ground loops be minimized?
A. Connect all ground conductors in series
B. Connect the AC neutral conductor to the ground wire
C. Avoid using lock washers and star washers when making ground connections
D. Bond equipment enclosures together

You should bond all equipment enclosures together, answer D.

This should be to a single point, termed star grounding or star earthing. Avoiding star washers in silly, as these are designed to bite through surface treatments or oxidisation.

G4C10
What could be a symptom caused by a ground loop in your station's audio connections?
A. You receive reports of "hum" on your station's transmitted signal
B. The SWR reading for one or more antennas is suddenly very high
C. An item of station equipment starts to draw excessive amounts of current
D. You receive reports of harmonic interference from your station

Ground loops can cause hum on your station's signal, answer A, or on a received signal.

G4C11
What technique helps to minimize RF "hot spots" in an amateur station?
A. Building all equipment in a metal enclosure
B. Using surge suppressor power outlets
C. Bonding all equipment enclosures together
D. Low-pass filters on all feed lines

Bond all equipment cases together (and to ground), answer C.

G4C12
Why must all metal enclosures of station equipment be grounded?
A. It prevents a blown fuse in the event of an internal short circuit
B. It prevents signal overload
C. It ensures that the neutral wire is grounded
D. It ensures that hazardous voltages cannot appear on the chassis

This prevents dangerous voltages appearing in the cases of the equipment, answer D.

G4D01
What is the purpose of a speech processor as used in a transceiver?
A. Increase the apparent loudness of transmitted voice signals
B. Increase transmitter bass response for more natural-sounding SSB signals
C. Prevent distortion of voice signals
D. Decrease high-frequency voice output to prevent out of band operation

These compress the level of the voice, to increase apparent loudness, answer A.

The previous answer A claimed an increase in intelligibility, especially in poor conditions.

G4D02
How does a speech processor affect a single sideband phone signal?
A. It increases peak power
B. It increases average power
C. It reduces harmonic distortion
D. It reduces intermodulation distortion

The compression they use increases the average power, answer B.

G4D03
What is the effect of an incorrectly adjusted speech processor?
A. Distorted speech
B. Excess intermodulation products
C. Excessive background noise
D. All of these choices are correct

All of these, answer D.

"Excess intermodulation products" is a technical description of the previous answer B, Splatter. Remember, all are correct.

In the bad old days of CB, there were "Power Microphones", designed to amplify and compress the audio, to gain the maximum output on SSB. On AM they really would only have caused poor audio quality. I recall them being illegal in at lease soem areas, and most would have been wound up all the way, causing all the problems above.

G4D04
What does an S meter measure?
A. Carrier suppression
B. Impedance
C. Received signal strength
D. Transmitter power output

It is relative or absolute received signal strength, answer C.

G4D05
How does a signal that reads 20 dB over S9 compare to one that reads S9 on a receiver, assuming a properly calibrated S meter?
A. It is 10 times less powerful
B. It is 20 times less powerful
C. It is 20 times more powerful
D. It is 100 times more powerful

20 dB is 100 times. We remember that 10 dB is 10 times, and doing this twice is 100 times, so D gets the tick.

G4D06
How much change in signal strength is typically represented by one S unit?
A. 6 dB
B. 12 dB
C. 15 dB
D. 18 dB

This is 6 dB, answer A.

G4D07
How much must the power output of a transmitter be raised to change the S meter reading on a distant receiver from S8 to S9?
A. Approximately 1.5 times
B. Approximately 2 times
C. Approximately 4 times
D. Approximately 8 times

This is 4 times the power, answer C, also 6dB.

G4D08
What frequency range is occupied by a 3 kHz LSB signal when the displayed carrier frequency is set to 7.178 MHz?
A. 7.178 to 7.181 MHz
B. 7.178 to 7.184 MHz
C. 7.175 to 7.178 MHz
D. 7.1765 to 7.1795 MHz

The highest components of your voice, 3 kHz after filtering, pushes the signal DOWN to 7.175 MHz, so it is 7.175 to 7.178 MHz, answer C.

Look for the dial frequency, as the UPPER frequency in the answer, if LSB is being discussed. Note that this is the bottom of the phone band segment permitted for General licence holders.

G4D09
What frequency range is occupied by a 3 kHz USB signal with the displayed carrier frequency set to 14.347 MHz?
A. 14.347 to 14.647 MHz
B. 14.347 to 14.350 MHz
C. 14.344 to 14.347 MHz
D. 14.3455 to 14.3485 MHz

With USB, the 3 kHz components of the voice moves the signal up by 3 kHz, to 14.350 MHz, so the range is 14.347 to 14.350 MHz, and B gets you the banana.

Note that this the top edge of the band.

G4D10
How close to the lower edge of a band's phone segment should your displayed carrier frequency be when using 3 kHz wide LSB?
A. At least 3 kHz above the edge of the segment
B. At least 3 kHz below the edge of the segment
C. At least 1 kHz below the edge of the segment
D. At least 1 kHz above the edge of the segment

You must keep the dial at least 3 kHz above the lower edge of the voice segment, answer A.

This applies to 80 and 40 metres, noting that General has access to a limited portion of this. For 160 metres, while voice is legal across the band, check with the voluntary plan which indicates 1.843 MHz as the lowest frequency for voice. This happens to be the frequency of a low cost crystal, so AM, DSB, and Morse stations may be here. There are also IC sized oscillators available on 1.8432 MHz, which must be used with a low pass filter, which can be used as a QRP transmitter.

G4D11
How close to the upper edge of a band's phone segment should your displayed carrier frequency be when using 3 kHz wide USB?
A. At least 3 kHz above the edge of the band
B. At least 3 kHz below the edge of the band
C. At least 1 kHz above the edge of the segment
D. At least 1 kHz below the edge of the segment

In the case of USB it is 3 kHz below the upper edge of the band segment, answer B.

This applies to 20 metres, the subject of the previous version of this question, to 15m, and to the WARC bands. For 10 metres and up, the upper edge of the bands tends to be the domain of repeaters and other operations.


While US amateurs are not allowed to use voice on 30 metres, for VKs, a similar principle applies to the segment used for voice, with the following on the WIA Band Plan, noting the segment above is recommended for data: It is recommended that whenever possible SSB activity should not extend above 10.130 kHz. For USB this corresponds to an indicated suppressed carrier frequency no higher than 10.127 MHz.


On to: Amateur Practices 3 - Mobile & Portable HF

You can find links to lots more on the Learning Material page.


Written by Julian Sortland, VK2YJS & AG6LE, June 2024.

Tip Jar: a Jefferson (US$2), A$3 or other amount / currency. Thanks!