Electrical Engineering › Section I · Sinusoidal waveforms and phasors › Topic I.4
AC meters and instruments
Point a multimeter at an ac signal and it shows a single number, but which number depends on how the meter works. Knowing whether a meter responds to the average or to the true rms value, and what an oscilloscope really shows, keeps you from trusting a reading that is quietly wrong.
Turning a wave into one number
The topic "Ammeters and voltmeters" dealt with steady dc readings. An ac signal swings both ways, so its full-cycle average is zero, and a meter must do something extra to show a useful figure. Nearly every ac meter is marked to show rms, the value that matters for power. But many meters do not actually measure rms. They measure something easier and scale it, and the scaling is only correct for a pure sinusoid. The kind of meter therefore decides whether a reading on a distorted waveform can be trusted.
Average-responding meters
The simplest ac meters pass the signal through a full-wave bridge rectifier of four diodes, which flips every negative half cycle upward. The meter then responds to the average of the rectified wave. For a sinusoid that average is \(0.637V_m\), while the rms value is \(0.707V_m\). The ratio of the two, the form factor of a sine wave, is
So the scale is marked to show 1.11 times the rectified average. Moving-coil analogue meters and many low-cost digital multimeters (DMMs) work this way. A meter using a half-wave rectifier sees half the average and multiplies by 2.22 instead.
Worked example: an average-responding meter on other shapes
An average-responding meter is used on a square wave alternating between +10 V and −10 V, then on a triangle wave of 10 V peak. 1. Square wave: the rectified wave sits at 10 V, so the average is 10 V and the meter shows \(1.11 \times 10 = 11.1\,\text{V}\). The true rms is 10 V, so the reading is about 11% high. 2. Triangle wave: the rectified average is \(V_m/2 = 5\,\text{V}\), so the meter shows \(1.11 \times 5 = 5.55\,\text{V}\). The true rms is \(10/\sqrt{3} = 5.77\,\text{V}\), so the reading is about 4% low. 3. Conclusion: the 1.11 factor is right only for sinusoids.
True-rms meters
A true-rms meter computes the rms value of whatever waveform it is given, either with an rms-converter circuit or by sampling the signal and calculating the root of the mean of the squares. Older instruments did this with the electrodynamometer movement, whose deflection depends on the square of the current. A true-rms reading is the one to trust on distorted waveforms such as the output of a variable-speed motor drive or a dimmer. Two limits remain. Each meter is accurate only up to a stated frequency, and only up to a stated crest factor. Many true-rms meters on their ac range also block the dc part and report the rms of the ac part alone; some offer an ac + dc setting.
Crest factor
The crest factor of a waveform is its peak value divided by its rms value. For a sinusoid it is \(\sqrt{2} \approx 1.41\); for a square wave it is 1; for a triangle wave it is \(\sqrt{3} \approx 1.73\). Narrow, spiky pulses have much higher crest factors. The current drawn by many electronic power supplies flows in short bursts near each voltage peak, and its crest factor is often well above that of a sinusoid. A meter's manual states the highest crest factor it can handle at a given reading; above it, the peaks overload the input circuit and the reading falls short.
Clamp meters and frequency counters
A clamp meter measures current without breaking the circuit. Its hinged jaws close around a single conductor, which acts as the primary winding of a transformer, so an alternating current induces a proportional current in a coil in the jaws. Clamp around one conductor only: around a whole cord, the equal and opposite currents in the two conductors cancel and the meter reads close to zero. Clamp meters that also read dc use a Hall-effect sensor instead, because a steady current induces nothing. A frequency counter counts cycles over a precise time interval and displays the frequency digitally, often to many more significant figures than an oscilloscope can resolve.
The oscilloscope
An oscilloscope draws the waveform itself, voltage against time, so you can see its shape, peak-to-peak value, period and any dc level. Most have two or more channels, so two waveforms can be compared on one time axis. The vertical scale is set in volts per division and the horizontal scale in seconds per division. The input coupling switch matters. On DC coupling the whole signal reaches the screen, including any dc level. On AC coupling a series capacitor blocks the dc part, so the trace is centred on zero. On GND the input is disconnected and the trace marks the zero-volt reference line.
Worked example: a signal with a dc level
On DC coupling at 2 V per division and 0.2 ms per division, a sinusoid spans 6.8 divisions from peak to peak, one cycle covers 5 divisions, and the centre of the wave sits 1.5 divisions above the GND line. 1. Peak-to-peak: \(6.8 \times 2 = 13.6\,\text{V}\), so \(V_m = 6.8\,\text{V}\). 2. Period: \(5 \times 0.2 = 1\,\text{ms}\), so \(f = 1\,\text{kHz}\). 3. DC level: \(1.5 \times 2 = 3\,\text{V}\). 4. Rms of the ac part: \(6.8/\sqrt{2} = 4.81\,\text{V}\). 5. Total rms: \(\sqrt{3^2 + 4.81^2} = 5.67\,\text{V}\). An ac-coupled true-rms meter would show 4.81 V; on AC coupling the trace would drop 1.5 divisions.
Mains supplies and safety
US homes receive 120 V at 60 Hz, with 240 V across the two hot conductors for large appliances; most of the world uses about 230 V at 50 Hz, and Japan uses both frequencies in different regions. Higher voltage carries the same power with less current: an 1800 W heater draws 15 A at 120 V but only 7.83 A at 230 V. Wiring loss goes as \(I^2R\), so the 120 V circuit loses about 3.7 times as much in the same wire. The price is higher shock risk and stricter insulation. Treat every live circuit with respect: isolate and lock out power before work, and use meters rated for the category of circuit being measured.
Common mistakes to avoid
First, trusting an average-responding meter on a square, triangle or distorted wave; its 1.11 scaling assumes a sinusoid. Second, assuming a true-rms meter is right at any frequency or crest factor; check its stated limits. Third, clamping around a whole cable, which cancels the reading. Fourth, reading an oscilloscope's peak-to-peak value as the peak, or as rms. Fifth, forgetting that AC coupling hides the dc level, and that most ac meter ranges ignore it too, so the true total rms can be larger than either shows.
Key takeaways
Average-responding meters rectify the signal and display 1.11 times its average, the sine-wave form factor, so they are correct only for sinusoids: a square wave reads about 11% high and a triangle about 4% low. True-rms meters compute the rms of any shape, within stated frequency and crest-factor limits, and often report the ac part only. Crest factor is peak over rms: 1.41 for a sine. Clamp meters measure current around one conductor by transformer action, or with a Hall sensor for dc. Oscilloscopes show the waveform; DC coupling includes the dc level and AC coupling removes it.
Flashcards
Flashcard
Rectifies the signal and scales its average to read rms for a sine
Test your understanding
Question 1 of 5
An average-responding meter is connected to a square wave alternating between +20 V and −20 V. What does it display?
Question 2 of 5
What is the crest factor of a pure sinusoid?
Question 3 of 5
A clamp meter is closed around a two-conductor extension cord feeding a 10 A heater. What does it read?
Question 4 of 5
On DC coupling a sinusoid's trace is centred 2 divisions above the GND line at 5 V per division. What happens when you switch to AC coupling?
Question 5 of 5
A trace spans 3.2 divisions peak-to-peak at 5 V per division. What rms voltage should a true-rms meter show for this sinusoid?
