Catto / Topic Map · Analog Electronics I Year 4
Analog Electronics I · 120 h · Topic 7 of 8

The oscilloscope

The instrument that shows the shape of a signal, not just its value. A multimeter says “5 V”; the oscilloscope tells you whether those 5 V are clean DC, a sine wave, a square wave with overshoot, or DC with 200 mV of hum riding on top.

Measurement and instrumentation Timebase Trigger Probes Waveforms

01What the screen shows

An oscilloscope draws a graph of voltage versus time. The vertical axis is voltage and the horizontal axis is time; the screen is divided into a grid of divisions (typically 8 vertical × 10 horizontal), and two knobs set what each division is worth:

  • VOLTS/DIV — how many volts each vertical division represents.
  • TIME/DIV (or SEC/DIV) — how much time each horizontal division represents.

Every oscilloscope measurement comes down to counting divisions and multiplying.

2 V/div 0.5 ms/divon-screen amplitude5.0 divvisible cycles5.0period reading1 msSeveral full cycles and a comfortable amplitude: this is how to measure.5 V/div 0.5 ms/divon-screen amplitude2.0 divvisible cycles5.0period reading1 msWith more volts per division the wave ends up tiny and reading accuracy is lost.2 V/div 0.1 ms/divon-screen amplitude5.0 divvisible cycles1.0period reading1 msFaster timebase: a single cycle fits, useful for examining an edge in detail.2 V/div 2 ms/divon-screen amplitude5.0 divvisible cycles20.0period reading1 msSlow timebase: many cycles fit and the shape of each one can no longer be made out.The signal is always the same: 5 V peak and 1 kHz. The only thing that changes is the instrument controls.
Figure 1. The controls applied to the same signal, animated: 5 V peak and 1 kHz seen with different volts per division and timebase settings. The signal does not change; how you look at it does.
Vpp=vertical div.×VOLTS/DIV T=horizontal div.×TIME/DIV,f=1T

02The cathode-ray tube

The analog oscilloscope is based on a cathode-ray tube (CRT). Although new instruments today are digital with an LCD screen, understanding the CRT explains the origin of all the controls that are still around.

cathode + filament G focus anodes electron gun Y plates (signal) X plates (timebase) screen with a phosphor layer glows where the beam hits brightness is set by grid G and focus by the anodes
Figure 2. Cathode-ray tube. The gun emits and accelerates a beam of electrons; two pairs of plates deflect it; the phosphor on the screen turns the impact into light.
  • Electron gun. A cathode heated by a filament emits electrons. The control grid (G) regulates how many get through: this is the INTENSITY control. The focus anodes concentrate the beam into a point: this is the FOCUS control.
  • Y plates (vertical). They receive the amplified signal. They deflect the beam up or down in proportion to the voltage.
  • X plates (horizontal). They receive the timebase, which sweeps the beam from left to right at constant speed.
  • Screen. Coated with phosphor, it glows where the beam hits and keeps that light for a moment (persistence), which lets you see a continuous curve instead of a single dot.

03The timebase

If a ramp that rises linearly and drops back abruptly (a sawtooth) is applied to the X plates, the beam travels across the screen from left to right at constant speed and returns instantly. This sweep turns the horizontal axis into a time axis.

max 0 sweep (left → right) retrace (the beam is blanked) the slope of the ramp is set by the TIME/DIV knob
Figure 3. Sawtooth timebase. During retrace the beam is turned off (blanking) so that the return line is not visible.
X-Y mode

If another signal is applied to the X plates instead of the timebase, the screen plots one signal against the other. This is useful for viewing the characteristic curve of a diode (voltage versus current) or for Lissajous figures, which are used to compare two frequencies: if the ratio is exact, the figure stands still.

04Synchronization: the trigger

This is the most important concept of the oscilloscope and the one that causes the most problems. If each sweep started at an arbitrary moment of the signal, each trace would be drawn shifted relative to the previous one and the screen would be a smear. The trigger (trigger circuit) makes every sweep start at the same point of the wave, so the picture stays still.

🔬 Instrument simulator A two-channel analog oscilloscope and a function generator, with the front-panel knobs working. You can move the trigger level until you lose it and watch the image drift away, which is exactly what this section explains. ›
levelLevel within the signalThe trigger always finds the same point of the wave and all the sweeps overlap: the imagestays still.levelLevel within, different pointWith a different level the image also stays still, but it starts at another part of the cycle.levelLevel above the peakThe trigger never finds that level: the sweep free-runs and the image drifts nonstop.The trigger decidesat which point of thesignal each sweepstarts.If they all startthe same way, theimage looks still.
Figure 4. The trigger, animated: with the level inside the signal, all sweeps start at the same point and the image stays still. With the level outside, the sweep free-runs.
ControlWhat it doesHow to use it
LEVELVoltage at which the sweep is triggered.It must lie within the range of the signal. If it goes beyond the peak, it never triggers.
SLOPEWhether it triggers on a rising (↑) or falling (↓) crossing.Changes the point where the trace starts. Useful for viewing the edge you care about.
SOURCEWhich channel it takes its reference from: CH1, CH2, EXT or LINE.When comparing two signals, always trigger on the same one: otherwise the measured phase shift is false.
MODE · AUTOIf there is no trigger, it sweeps anyway.Default mode: you always see something, even if it is just a line. This is the one to start with.
MODE · NORMALSweeps only when there is a valid trigger.Blank screen if it does not trigger. Better for slow or sporadic signals.
MODE · SINGLETriggers only once and freezes.For capturing a one-time event: a motor starting, a switching spike.
HOLDOFFDead time after each trigger.Stabilizes complex signals that cross the level several times within a single cycle.
Recipe for when “nothing shows up”
  1. Trigger mode on AUTO, source set to the channel being used.
  2. Channel coupling on DC and vertical position at the center.
  3. VOLTS/DIV at a large value (5 V/div) and TIME/DIV at a middle value (1 ms/div).
  4. Check that the channel is turned on and the probe is properly connected, with the ground clip on the correct point.
  5. Only then adjust the trigger level and fine-tune the scales.

05Coupling and probes

Input coupling

DC

Passes everything: DC and AC. It is the default mode and the one that shows the signal as it really is.

AC

Inserts a capacitor: it blocks DC and leaves only the variation. Useful for viewing the ripple of a 12 V power supply at a sensitivity of 50 mV/div.

GND

Disconnects the input and ties it to ground. It is used to mark where zero is on the screen before measuring.

Probes

ProbeAttenuationInput ZWhen to use it
×1None1 MΩ · 100 pFSmall, low-frequency signals. It loads the circuit more.
×10Divides by 1010 MΩ · 15 pFThe normal working position. It loads the circuit much less and reaches a higher bandwidth.
Two things to do with the ×10 probe
  • Tell the oscilloscope. If the probe is on ×10 and the instrument thinks it is ×1, all voltage readings come out 10 times too low. On digital scopes this is set in a menu; on analog ones you have to multiply in your head.
  • Compensate it. With the tip on the oscilloscope's own square-wave calibration output, adjust the probe's trimmer screw until the square wave has square corners. If they look rounded or show overshoot, the probe is out of compensation and distorts every fast signal.
The ground clip

The probe ground is connected to the oscilloscope ground and, through the power cord, to the third conductor of the installation (the earth wire). Consequences:

  • Connecting the ground clip to a point that is not ground shorts that point to earth. In a circuit powered directly from the mains without an isolation transformer, that can destroy the equipment or the oscilloscope, or be dangerous for the operator.
  • With two channels, the two ground clips are tied together. You cannot measure two points with different grounds at the same time.
  • To measure floating signals, use an isolation transformer or a differential probe. Never cut off the earth pin of the oscilloscope's plug.

06What is measured and how

QuantityHow it is obtained
Peak-to-peak voltageDivisions between the maximum and the minimum × VOLTS/DIV.
Peak voltageHalf of the previous one, if the signal is symmetrical.
RMS value (sine wave)VRMS = Vp / √2 = 0.707 · Vp. Only for a pure sine wave.
DC componentSet the channel to GND to mark zero, switch to DC and measure how far the trace shifted.
Period and frequencyT = divisions of one cycle × TIME/DIV. Then f = 1/T.
Rise timeTime between 10% and 90% of the edge amplitude.
Duty cycle(time high / period) × 100%.
Phase shiftWith two channels: measure Δt between the zero crossings and apply the formula below.
φ=ΔtT×360° Phase shift between two signals of the same frequency. If Δt is one quarter of the period, the phase shift is 90°.
Worked example · Complete reading

With VOLTS/DIV = 2 V (probe on ×10, correctly configured) and TIME/DIV = 20 µs, the signal spans 3 divisions peak to peak and one cycle measures 5 divisions.

  • Vpp = 3 × 2 V = 6 V → Vp = 3 V
  • VRMS = 3 / 1.414 = 2.12 V (if it is a sine wave)
  • T = 5 × 20 µs = 100 µs → f = 1/100 µs = 10 kHz

If the probe were on ×10 and the instrument set to ×1, the reading would have been 0.6 V: an error of one order of magnitude, and the most frequent mistake when starting out.

Instrument bandwidth

tr≈0.35BW A 20 MHz oscilloscope cannot show edges faster than 17.5 ns: it rounds them off. Rule of thumb: the bandwidth must be at least 5 times the frequency of the signal you want to see with its correct shape.

07Analog and digital

A digital oscilloscope (DSO) does not deflect a beam: it samples the signal with an A/D converter, stores the samples in memory and draws them on an LCD screen. That changes everything for the better, except for one detail you need to know about.

What digital gains
  • Freezes the image and lets you save or export it.
  • Pretrigger: shows what happened before the trigger. Impossible in analog.
  • Automatic measurements (Vpp, frequency, RMS, duty cycle).
  • Cursors, math between channels, FFT.
  • Captures one-time events with SINGLE.
What to watch out for
  • Aliasing. If the sampling rate is insufficient, the screen shows a signal of a false frequency, a lower one, that looks perfectly real.
  • With few samples per cycle, the shape looks angular even if the signal is a sine wave.
  • Very fast, isolated events (glitches) can fall between two samples and not show up.
Sampling theorem

To reconstruct a signal you must sample it at more than twice its maximum frequency (Nyquist). In practice oscilloscopes use 5 to 10 samples per cycle. If the screen shows a suspiciously slow signal, speed up the timebase: if the displayed frequency changes, it was aliasing. This theorem reappears in Telecommunications II and in the A/D conversion topic of Year 5.

08Measurements in amplifier circuits

This is the direct application of the instrument in this subject, on the circuits of the small-signal amplifiers topic.

What you look forHowWhat it indicates
GainCH1 at the input, CH2 at the output. Av = Vout(pp) / Vin(pp).Compare it with the calculated value. If it differs a lot, check CE and the load.
Phase inversionBoth channels at once.In a common-emitter stage the output is at 180°. If it is not, the stage is not what you think it is.
Clipping and Q-pointRaise the input until the output flattens.In a common-emitter stage, if it clips at the top first, the Q-point has shifted toward cutoff; if it clips at the bottom, toward saturation.
Power supply rippleChannel on AC coupling, 50 mV/div, across the filter capacitor.You see the sawtooth at 100 Hz and measure its peak-to-peak amplitude.
DistortionCompare the shape of the output with a sine wave.Asymmetries, flattening or the step of crossover distortion in power stages.
Frequency responseFixed input amplitude, sweep the frequency.You find the cutoff frequencies (where the output drops to 0.707 of its maximum, that is, −3 dB).
Parasitic oscillationsFast timebase (1 µs/div) with no input signal.If a signal of hundreds of kHz appears, the amplifier is oscillating: decoupling is missing.

09In the lab

Lab 1 · Setup and probe compensation

Connect the probe to the calibration terminal (1 kHz, 2 Vpp on most instruments). Adjust intensity, focus, position and scales until you see two or three complete cycles. Check that the reading matches what the label on the terminal says. Adjust the probe's compensation trimmer until you get square corners; sketch in your folder the three possible shapes (undercompensated, correct, overcompensated).

Lab 2 · Measuring a sine wave from the generator

Function generator at 1 kHz, 4 Vpp. Use the oscilloscope to measure the amplitude, the period and the frequency. Calculate the RMS value and compare it with a multimeter on the AC range. Repeat with a square wave and a triangle wave at the same amplitude: a multimeter without True RMS will give values different from the correct one, and that is where you see why.

Lab 3 · DC and AC superimposed

On the rectified and filtered power supply from the semiconductors topic: with DC coupling you see an almost straight line at the DC level; switching to AC and raising the sensitivity reveals the sawtooth ripple at 100 Hz. Measure its peak-to-peak amplitude and compare it with the value calculated from Vr = I/(f·C).

Lab 4 · Phase shift in an RC circuit

Series circuit with R = 1 kΩ, C = 100 nF, driven with 1 Vpp at 1 kHz. CH1 at the input, CH2 across the capacitor (with both grounds at the same point). Measure Δt between the zero crossings and calculate φ. The theoretical result is φ = −arctan(2πfRC) = −32°. Repeat at 10 kHz and check that the phase shift approaches 90°.

10Common mistakes

SymptomUsual cause
All voltages read 10 times too low (or 10 times too high)The probe is on ×10 and the instrument is set to ×1, or the other way around.
The image “runs” or looks blurryTrigger set incorrectly: the level is outside the range of the signal, or the trigger source is the wrong channel.
Blank screenTrigger mode on NORMAL with no valid trigger. Switch to AUTO.
The square wave shows rounded corners or “ears”Probe out of compensation. Adjust it using the calibration terminal.
A slow sine wave appears that should not be thereAliasing on a digital oscilloscope. Change the timebase and see whether the displayed frequency changes.
Connecting the ground clip trips the residual-current device (RCD) or burns componentsThe clip was connected to a point that is not ground. Mains-connected equipment requires an isolation transformer.
The signal is full of 50 Hz noiseGround lead too long, or ground clip far from the measurement point.
A horizontal line is displayed even though there is a signalCoupling on GND, or the channel is turned off, or the timebase is too slow for that frequency.

11Self-assessment

With VOLTS/DIV = 0.5 V and TIME/DIV = 2 ms, a signal is 6 divisions tall and one cycle measures 4 divisions. What are Vpp, T and f?

Vpp = 6 × 0.5 = 3 V. T = 4 × 2 ms = 8 ms. f = 1/0.008 = 125 Hz.

What is AC coupling for, and when should you NOT use it?

It is for viewing a small variation riding on a large DC level (the ripple of a power supply, for example), because it blocks the DC and lets you raise the sensitivity. It is not suitable when the absolute level of the signal matters, nor with very slow signals or signals with a low duty cycle, because the coupling capacitor distorts them.

What exactly does the trigger circuit do?

It makes each sweep always start at the same point of the signal: when it crosses a given voltage level with a given slope. That way successive traces overlap exactly and the image looks still.

Why is the ×10 probe preferable even though it attenuates the signal?

Because it presents 10 MΩ and about 15 pF instead of 1 MΩ and 100 pF: it loads the circuit much less, especially at high frequency, where the capacitance of the ×1 probe can completely alter the circuit's operation. The attenuation is compensated by the instrument's scale.

Two 1 kHz signals are 250 µs apart. What is the phase shift?

T = 1 ms. φ = (250 µs / 1000 µs) × 360° = 90°.

What is aliasing and how is it detected?

It is a false signal, of a lower frequency than the real one, that appears on a digital oscilloscope when the sampling rate is insufficient. It is detected by changing the timebase: if the displayed frequency changes, it was aliasing; a real signal keeps its frequency.

The output of an amplifier flattens only at the top. What does that indicate?

That the Q-point has shifted and the signal hits one end of the load line sooner. You need to recalculate the biasing to center the Q-point and recover the maximum symmetrical swing.

Why can't the ground clips of the two channels be connected to different points?

Because internally they are tied together and to the earth of the plug. Connecting them to points at different potentials shorts those two points through the oscilloscope, with a risk of damage to the circuit, the instrument and the person.

What bandwidth does an oscilloscope need to display a 1 MHz square wave properly?

At least 5 MHz, and preferably 10 MHz or more: a square wave contains odd harmonics of the fundamental, and if the instrument cuts them off, the edges look rounded. With the formula tr = 0.35/BW, a 20 MHz instrument shows edges of down to 17.5 ns.

Development of the topic “The oscilloscope” of Analog Electronics I (Year 4), based on the “Curriculum Proposal – Second Cycle of the Technical-Vocational Track, Secondary Education – Electronics,” Ministry of Education of the Province of Córdoba, DGETyFP. Back to the Topic Map · catto.ar