E4: Amateur Practices

An Extra is expected to own and understand the instruments on a serious workbench. This group introduces the big four, the oscilloscope, the spectrum analyzer, the antenna analyzer, and the frequency counter, and a few habits that make their readings trustworthy.

The oscilloscope: seeing a signal as a picture

An oscilloscope ("scope" for short) draws a graph of a signal's voltage over time, so you can literally see the shape of a waveform. The modern kind is a digital storage oscilloscope, which captures the signal by taking thousands or millions of quick voltage snapshots per second and reconstructing the picture from them. The part that takes those snapshots is the analog-to-digital converter (ADC), a chip that turns a smoothly varying voltage into a stream of numbers.

Because a digital scope works by sampling, the thing that limits the highest frequency it can accurately display is the sampling rate of its analog-to-digital converter. Picture trying to film a spinning wheel: if your camera does not take frames fast enough, you simply cannot capture how fast the wheel really spins. A scope must sample much faster than the signal it is watching.

When the sampling rate is too slow for the signal, you get a famous trap called aliasing. Aliasing is when undersampling fools the instrument into drawing a signal that is not really there. The visible effect on a digital scope is that a false, jittery low-frequency version of the waveform is displayed, a slow, wandering ghost that looks nothing like the real fast signal. (That spinning-wheel-going-backwards illusion in old movies is aliasing.)

Using a scope probe well

A scope probe is the cable-and-tip you touch to the circuit. Two practical points the exam wants:

• Compensation. A standard probe contains a small adjustable capacitor that must be tuned to the particular scope, a one-time tweak called compensating the probe. You do it by feeding the probe a clean square wave (a signal that snaps between two levels with flat tops) and adjusting the probe until the horizontal portions of the displayed wave are as nearly flat as possible. Overshoot or droop on those flat tops means the probe is mis-compensated.
• Keep the ground lead short. Good practice is to minimize the length of the probe's ground connection. A long ground lead acts like a little antenna and inductor that adds ringing and noise to fast signals, distorting what you see.

The spectrum analyzer: seeing signals by frequency

Where a scope shows voltage versus time, a spectrum analyzer shows signal amplitude on the vertical axis and frequency on the horizontal axis. Instead of a waveform, you see a row of peaks, one for each frequency present and how strong it is. That makes it the perfect tool for finding signals that should not be there.

Two uses to remember: a spectrum analyzer is the instrument used to display spurious signals and/or intermodulation distortion products generated by an SSB transmitter. ("Spurious" means unwanted extra signals; "intermodulation distortion products," which we will study in E4D, are extra signals created when two tones mix in something that is not perfectly linear.) When you transmit and your spectrum shows ugly extra peaks beside your main signal, the analyzer is telling you the transmitter is misbehaving.

The antenna analyzer

An antenna analyzer is a small instrument that sends a low-power test signal into your antenna system and reports back how well it is working. Compared with a plain SWR bridge (an older device that only shows standing wave ratio, a measure of how well your feed line and antenna are matched), the big advantage is that antenna analyzers compute SWR and impedance automatically. You do not have to transmit a real signal or do the math by hand.

An antenna analyzer can measure a whole range of useful things, in fact all of these: SWR, impedance (including its resistance and reactance parts), and the frequency where the antenna is resonant. So when a question asks what it can measure and lists several, choose "all these choices are correct."

Note that SWR itself can be measured several ways, by an SWR bridge, by a directional/SWR meter, and by an antenna analyzer, so the question "which is used to measure SWR?" also answers to "all these choices are correct."

Frequency counters and prescalers

A frequency counter measures a signal's frequency by counting cycles in a known time window. Counters have an upper frequency limit. To measure a signal that is too fast for the counter, you put a prescaler in front of it. A prescaler is a digital divider that cuts the frequency down by a fixed factor (say, divides by 10 or 100). Its purpose is to reduce the signal frequency to within the counter's operating range, after which you multiply the reading back up.

One scope trigger detail

To measure the small leftover hum on a linear power supply's output, called ripple (the small AC variation that survives on top of the DC output), the most effective scope trigger mode is Line. Line triggering synchronizes the display to the AC power-line frequency, which is exactly what ripple is tied to, so the ripple waveform stands still on the screen instead of sliding around.

Owning good instruments is only half the job; the other half is knowing how far to trust them and how to use them correctly. This group covers what limits accuracy, how to calibrate, how to measure quality factor (Q), and the modern language of S parameters and vector network analyzers.

What limits accuracy

• Frequency counters. The factor that most affects a frequency counter's accuracy is its time base accuracy. The "time base" is the internal clock (usually a crystal oscillator) that defines the counting window. If that clock is off by a part in a million, every frequency you read is off by the same proportion. The counter can only be as accurate as its own clock.
• Voltmeter loading and "ohms per volt." An analog voltmeter draws a tiny current from the circuit it measures, which can disturb the reading. Its sensitivity, given in ohms per volt, tells you how little it loads the circuit. The meaning to memorize: the full-scale reading of the voltmeter multiplied by its ohms-per-volt rating is the input impedance of the voltmeter. A higher ohms-per-volt number means higher input impedance and less disturbance of the circuit.

Directional power meters: forward and reflected

A directional power meter (or wattmeter) measures power flowing toward the load and power bouncing back separately, called forward power and reflected power. The power actually delivered to the load is simply forward minus reflected. So if a meter between a transmitter and its load reads 100 watts forward and 25 watts reflected, the load is absorbing 75 watts (100 minus 25). Easy subtraction, but a very common exam item.

Measuring Q

Q, short for quality factor, describes how "sharp" or low-loss a tuned circuit is; a higher Q means a narrower, sharper resonance peak. One practical way to find the Q of a series-tuned circuit is from the bandwidth of the circuit's frequency response. A high-Q circuit responds over only a narrow band of frequencies; a low-Q circuit responds over a wider band. Measure that bandwidth (relative to the center frequency) and you have the Q.

S parameters: describing a circuit by its ports

At RF, the cleanest way to describe how a device passes and reflects signals is with scattering parameters, almost always shortened to S parameters. Think of a device as a box with connection points called ports (port 1 is typically the input, port 2 the output). Each S parameter is a ratio that tells you what comes out of one port when you drive another. The two subscripts on an S parameter represent the port or ports at which measurements are made, written as S(out)(in).

• S21 is the signal coming out of port 2 when you drive port 1, which is the forward gain (or forward transmission) of the device.
• S11 is the signal reflected back out of port 1 when you drive port 1, which represents the input port return loss or reflection coefficient, equivalent to VSWR. In plain terms, S11 tells you how well the input is matched, the same information SWR gives you.

The vector network analyzer (VNA)

A vector network analyzer is the instrument that measures S parameters across a range of frequencies. "Vector" means it captures both the size and the phase (timing) of signals, not just the size. A two-port VNA can measure a great deal, for example a filter's frequency response, and in general the answer to "what can a VNA measure?" is "all these choices are correct" (impedance, return loss, gain, and more).

Before a VNA's readings mean anything, you must calibrate it at the ends of your test cables using three known standards: a short circuit, an open circuit, and a 50-ohm load. These three references let the instrument cancel out the effects of its own cables and connectors. (Remember them as "short, open, and 50 ohms.")

Measuring transmitter intermodulation distortion

To measure the intermodulation distortion (IMD) of an SSB transmitter, the standard method is the two-tone test: you modulate the transmitter using two audio (AF) signals having non-harmonically related frequencies and observe the RF output with a spectrum analyzer. Two clean tones go in; if the transmitter is perfectly linear, only those two come out, but any nonlinearity creates extra products you can see and measure on the analyzer. The tones are chosen non-harmonic on purpose so the distortion products land where they are easy to spot.

Now we get to the heart of what makes a receiver good. This group is about hearing weak signals: how much noise the receiver itself adds, how it rejects unwanted frequencies, and how cleanly it converts signals. Many of these ideas live inside the superheterodyne ("superhet") receiver, the dominant design that mixes the incoming signal with a tunable local oscillator (LO) to shift it to a fixed intermediate frequency (IF), where the real filtering and amplification happen.

Noise floor, noise figure, and MDS

Every receiver has a faint background hiss below which it cannot hear, called the noise floor. The lower the noise floor, the weaker a signal you can still detect.

• The -174 dBm landmark. A receiver noise floor of -174 dBm represents the theoretical noise in a 1 Hz bandwidth at the input of a perfect receiver at room temperature. This is a law-of-physics minimum, the quietest any receiver could possibly be in a 1-hertz-wide slice. Real receivers are noisier than this ideal.
• Bandwidth and noise. Noise power grows with bandwidth, so a wider receive filter lets in more noise. Specifically, widening the bandwidth multiplies the noise by the bandwidth ratio. Going from 50 Hz to 1,000 Hz is a 20-times-wider window, which raises the noise floor by 13 dB. (Ten times wider is 10 dB; the extra doubling to 20 times adds 3 dB, totaling 13 dB.)
• Noise figure. The noise figure of a receiver is the ratio in dB of the noise generated by the receiver to the theoretical minimum noise. It is a single number for how much worse than perfect a receiver's own electronics are. A lower noise figure is better.
• MDS. The MDS stands for the minimum discernible signal, the weakest signal a receiver can just barely detect above its noise floor. (The exam phrases this two ways: "what does MDS stand for?" answers "minimum discernible signal.")

Selectivity and choosing bandwidth

Selectivity is a receiver's ability to separate the signal you want from signals on nearby frequencies. A key tool is variable bandwidth. The advantage of having a variety of receiver bandwidths to select from is that you can set the receive bandwidth to match the modulation bandwidth, maximizing signal-to-noise ratio and minimizing interference. A narrow CW signal wants a narrow filter (which keeps out noise and neighbors); a wider voice signal wants a wider filter so it is not muffled.

A related control is the IF Shift. Its purpose is to reduce interference from stations transmitting on adjacent frequencies. IF Shift slides the receiver's passband sideways relative to your tuned signal, letting you move an interfering neighbor out of the passband while keeping your desired signal centered in the audio.

A roofing filter is a narrow filter placed very early, right after the first mixer in the IF chain. A narrow-band roofing filter improves blocking dynamic range by attenuating strong signals near the receive frequency, meaning it knocks down powerful nearby signals before they can overload the later stages. (We define "blocking dynamic range" fully in E4D.)

Rejecting images and out-of-band signals

A superhet has a weakness called an image: a second, unwanted frequency, located twice the IF away from the wanted one, that the mixer will also convert straight into the IF if it is not filtered out first.

• A good reason to choose a high IF for a superhet HF or VHF receiver is that it is easier for front-end circuitry to eliminate image responses. The higher the IF, the farther the image frequency sits from the wanted signal, so a simpler input filter can reject it.
• To eliminate interference from strong out-of-band signals (signals outside the band you are listening to), the effective circuit is a front-end filter or preselector, a tuned filter at the very input that passes your band and rejects everything far away.

The capture effect (FM)

In an FM receiver there is a built-in winner-take-all behavior: when two signals are on the same frequency, the stronger one takes over and the weaker disappears entirely. The term for this suppression in an FM receiver of one signal by another stronger signal on the same frequency is the capture effect. It is why an FM repeater conversation does not garble the way an AM signal would, the strongest signal simply captures the receiver.

Phase noise and reciprocal mixing

An oscillator is never perfectly pure; it has a little smear of phase noise, tiny random jitter in its frequency that spreads a faint noise skirt around the oscillator's signal. This causes two related problems:

• Reciprocal mixing. This is local oscillator phase noise mixing with adjacent strong signals to create interference to desired signals. A strong neighbor, mixed with the LO's noisy skirt, gets smeared across your listening frequency as noise even though it is not on your frequency. The cure is a cleaner, lower-phase-noise local oscillator.
• SDR master clock phase noise. In a software-defined radio, excessive phase noise in the master clock oscillator has a similar effect: it can combine with strong signals on nearby frequencies to generate interference.

SDR overload and front-end attenuation

A software-defined radio (SDR) digitizes signals with an ADC, and that converter has a maximum input it can represent. An SDR receiver is overloaded when input signals exceed the reference voltage of the analog-to-digital converter. Beyond that point the converter clips and produces gross distortion.

One simple defense on the crowded low bands is to switch in some input attenuation (a few decibels of loss at the antenna input). On the lower HF bands this reduces receiver overload with little or no impact on signal-to-noise ratio because atmospheric noise is generally greater than internally generated noise even after attenuation. In plain terms, those bands are so noisy from nature that knocking everything down a bit does not make weak signals any harder to hear, but it does keep strong signals from overwhelming the front end.

This group is about the upper end of a receiver's abilities: how big a range of signal strengths it can handle at once, and what goes wrong when a strong signal pushes it past its limits. The recurring villain here is nonlinearity, any circuit that does not respond in perfect proportion to its input, because nonlinearity is what creates spurious mixing products.

Dynamic range and blocking

Dynamic range is the span between the weakest signal a receiver can detect and the strongest it can tolerate before performance falls apart. One specific measure is blocking dynamic range, defined as the difference in dB between the noise floor and the level of an incoming signal that will cause 1 dB of gain compression. ("Gain compression" means the receiver's amplification starts to sag because a strong signal is overdriving it; 1 dB of sag is the agreed test point.) A bigger blocking dynamic range means the receiver can keep hearing a weak signal even when a strong one is nearby.

What does poor dynamic range look like in practice? It produces spurious signals caused by cross modulation, and desensitization from strong adjacent signals. In other words, strong neighbors create phantom signals and deafen the receiver.

Desensitization and how to fight it

Desensitization (often "desense") is the reduction in receiver sensitivity caused by a strong signal near the received frequency. A powerful nearby transmitter effectively turns down your receiver's hearing. A direct way to reduce the likelihood of desensitization is to insert attenuation before the first RF stage, knocking the strong offender down before it can overload the front end. (Yes, you attenuate your own weak signal too, but you gain far more by keeping the front end out of overload.)

Intermodulation: where extra signals are born

Intermodulation (IMD) is the creation of new, unwanted signals when two or more signals mix together in something nonlinear. The fundamental cause to remember: intermodulation is caused by nonlinear circuits or devices. A perfectly linear circuit will never make these products; only nonlinearity does.

• Between two nearby repeaters, intermodulation interference is created when the output signals mix in the final amplifier of one or both transmitters. One transmitter's signal leaks into the other's final amplifier, mixes there, and generates new spurious frequencies.
• To reduce or eliminate that repeater intermod, install a properly terminated circulator at the output of the repeater's transmitter. A circulator is a device that lets power flow out toward the antenna but routes any signal coming back in from the other transmitter into a load instead of into the final amplifier, so it never gets a chance to mix.

A subtlety the exam loves: odd-order intermodulation products (the third order, fifth order, and so on) are of particular interest because the odd-order products of two signals in the band being received are also likely to be within that band. Even-order products usually land far away where filters remove them; odd-order ones land right next to the signals that made them, where you cannot filter them out.

Third-order intercept point

The third-order intercept point, given in dBm, is a single figure of merit for how resistant a receiver (or amplifier) is to making those troublesome third-order products; higher is better. The careful definition: a third-order intercept level of 40 dBm means that a pair of 40 dBm input signals will theoretically generate a third-order intermodulation product that has the same output amplitude as either of the input signals. It is a theoretical, extrapolated point, real devices overload long before reaching it, but it lets you compare receivers fairly.

The preselector

A preselector is a tunable filter at the front of a receiver. Its purpose is to increase the rejection of signals outside the band being received, protecting the sensitive front end from strong out-of-band signals that could cause overload or intermod.

MDS as a power level, and link math

The exam asks you to turn signal levels in dBm into actual power and to add up a link budget. Two anchor facts make this easy: 0 dBm is 1 milliwatt, and every 10 dB changes power by a factor of ten.

• A minimum discernible signal of -100 dBm equals 0.1 picowatts. Counting down from 0 dBm (1 mW) in steps of 10 dB: -30 dBm is 1 microwatt, -60 dBm is 1 nanowatt, -90 dBm is 1 picowatt, and -100 dBm is one-tenth of that, 0.1 picowatts.
• Received signal level. With transmit power +40 dBm, a transmit antenna gain of 6 dBi, a receive antenna gain of 3 dBi, and a path loss of 100 dB, you simply add the gains and subtract the loss: 40 + 6 + 3 - 100 = -51 dBm. (A "dBi" is antenna gain compared to an ideal point source; "path loss" is how much the signal weakens traveling between the antennas.)
• Link margin. The link margin is how much signal you have to spare above the minimum needed for a reliable contact. With +40 dBm transmit power, 10 dBi antenna gain, 3 dB cable loss, 136 dB path loss, a receiver MDS of -103 dBm, and a required signal-to-noise ratio of 6 dB: first find the received power, 40 + 10 - 3 - 136 = -89 dBm. The signal needed is the MDS plus the required SNR, -103 + 6 = -97 dBm. The margin is the received power minus what is needed: -89 - (-97) = +8 dB.

These look intimidating, but the trick is always the same: add the gains, subtract the losses, all in decibels. No multiplication, just addition and subtraction.

The best receiver in the world is useless if your own house, car, and neighborhood drown it in noise. This final group is about identifying real-world interference, reducing it, and protecting the station. Two recurring concepts are common-mode current and the single point ground panel, so watch for them.

Reducing noise in the receiver: blankers, notches, and DSP

Modern receivers fight noise in several ways. Know what each tool does, and its side effects:

• A noise blanker removes impulse noise, short, sharp bursts such as ignition pops or electric-fence ticks, by briefly muting the receiver during each spike. Its downside: strong signals may be distorted and appear to cause spurious emissions, because a strong wanted signal can trigger the blanker and get chopped up.
• A digital noise reduction (DSP) uses digital signal processing to lower many kinds of background noise. The types it can often reduce are, in fact, all of these common kinds, so on that question choose "all these choices are correct."
• An automatic notch filter (ANF) automatically nulls out steady interfering carriers (single unmodulated tones). The catch when receiving CW: it can cause removal of the CW signal as well as the interfering carrier, because a Morse signal looks like a steady tone to the filter. So turn the ANF off for CW.

Suppressing noise at its source

Often the best fix is to choke the noise where it is made:

• Automobile charging-system noise that travels along the wires (conducted noise) can be suppressed by installing ferrite chokes on the charging system leads. A ferrite choke is a ring of magnetic material that you clamp around a wire to block high-frequency noise riding on it.
• RF interference from a line-driven AC motor is suppressed with a brute-force AC-line filter in series with the motor's power leads. A "brute-force" filter is simply a heavy-duty power-line filter that blocks the noise from getting onto the household wiring.

Recognizing interference by its fingerprint

Different sources leave different signatures on the receiver, and the exam expects you to match them:

• Computer network equipment causes the appearance of unstable modulated or unmodulated signals at specific frequencies.
• Switch-mode power supplies (the small, efficient supplies in nearly every modern gadget) cause a series of carriers at regular intervals across a wide frequency range, evenly spaced spurs marching across the band.
• Corroded metal connections nearby can act as accidental mixers: they explain local AM broadcast band signals combining to generate spurious signals on the MF or HF bands, because the corroded connections mix and reradiate the broadcast signals. This is sometimes called the "rusty bolt effect."
• Intermittent loud roaring or buzzing AC line interference can come from several power-line faults at once, so its causes are all of these choices.

Common-mode current: the sneaky troublemaker

Common-mode current is current that flows the same direction on all the conductors of a cable at once, rather than the intended balanced down-one-and-back-the-other flow. It is the current that flows equally on all conductors of an unshielded multiconductor cable. Because it is not balanced, it turns the cable itself into an unintended antenna. That is exactly why shielded cables can radiate or receive interference: common-mode currents on the shield and conductors. The usual cure is, again, a ferrite choke (a common-mode choke) on the cable.

Grounding and lightning protection

For both noise control and safety, a serious station routes every entering cable through one common entry point called the single point ground panel, a metal panel where all feed lines, control cables, and their protectors connect to a common ground.

• An AC surge protector should be installed on the single point ground panel, right where the power enters with everything else, so all the protective devices share one ground reference.
• The purpose of a single point ground panel is to ensure all lightning protectors activate at the same time. If every protector shares one ground at one point, a lightning surge raises them all together to the same voltage, so destructive voltage differences do not build up between your cables and fry the equipment in between.