Oscilloscopes
The oscilloscope shows voltage versus time. Learn vertical and horizontal settings, triggering, bandwidth and rise time, sample rate, probes and compensation, safe grounding, and measurements including FFT.
A multimeter gives a single number; an oscilloscope shows the whole waveform — shape, timing, noise, glitches, ringing and the relationship between several signals. It is the most important tool on an electronics bench.
Modern digital storage oscilloscopes sample the signal with a fast ADC, trigger on events, and compute measurements and spectra automatically. Using one well means understanding what can distort what you see.
An analogy
An oscilloscope is a high-speed camera for voltage: bandwidth is the lens sharpness, sample rate is the frame rate, and the trigger is the moment you press the shutter.
Vertical, horizontal and trigger
Vertical controls set volts per division and offset for each channel; horizontal sets time per division; the trigger decides when to start drawing, so repetitive signals stand still. Edge trigger on a rising edge at a chosen level is the default; pulse-width, runt, serial-bus and other triggers catch specific events.
Coupling
DC coupling shows the full signal; AC coupling removes the DC component so you can zoom into small ripple on a large DC rail. GND coupling shows where 0 V is.
Bandwidth, rise time and sample rate
A scope's bandwidth attenuates high frequencies: at the rated bandwidth a sine is already 30 % (−3 dB) low, and fast edges look slower. Rule of thumb: choose a scope with ≥ 3–5× the signal's highest frequency, and rise time 3–5× faster than the edge you measure. The sample rate must be high enough to reconstruct the signal (sin(x)/x interpolation needs ≥ ~2.5× BW).
Probes
A 10× passive probe presents 10 MΩ ∥ ~10 pF and must be compensated (adjust its trimmer until a square wave has flat tops). The long ground lead forms an inductive loop that rings on fast edges — use the short spring ground. 1× probes have far lower bandwidth.
Safety and differential measurements
A bench scope's ground clip is connected to mains earth. Clipping it onto a mains-referenced or floating node shorts that node to earth — destroying circuits and endangering people. Use high-voltage differential probes or an isolated scope, and never 'float' a scope by defeating its earth.
Advanced analysis
Scopes decode serial buses (I²C, SPI, CAN, UART), compute FFTs (watch for aliasing and window choice), perform mask and eye-diagram testing, and power analysis (switching loss, SOA, harmonics). Segmented memory captures many rare events with little dead time.
Formulas
- Scope rise time:
t_r,scope ≈ 0.35 / BW - Measured rise time:
t_r,meas = √(t_r,signal² + t_r,scope² + t_r,probe²) - Frequency from display:
f = 1 / (divisions per period × s/div) - Sample rate guideline:
f_s ≥ 2.5–5 × BW
Worked example: Reading the display
Given: 2 V/div, 5 µs/div; Peak-to-peak spans 6.4 divisions; One period spans 4 divisions
Find: Amplitude and frequency.
- V_pp = 6.4 × 2 = 12.8 V → V_peak = 6.4 V, V_rms (sine) = 4.53 V
- T = 4 × 5 µs = 20 µs → f = 50 kHz
Answer: 12.8 V p-p, 50 kHz
Worked example: Rise-time error
Given: 100 MHz scope → t_r ≈ 3.5 ns; Actual edge = 2 ns
Find: Displayed rise time.
- t_meas = √(2² + 3.5²) = √16.25 = 4.03 ns
- Error = +100 %: the scope is too slow for this edge
Answer: Shows ≈ 4 ns — double the real value
Common mistakes
- Wrong: Clipping the ground lead onto a mains-live or floating point. — Right: Use a differential probe or isolated instrument; the ground clip is earthed.
- Wrong: Leaving the probe on 1× or uncompensated. — Right: Match probe attenuation in the scope settings and compensate each probe on its channel.
Applications
- Debugging digital systems: Checking clock integrity, bus timing and glitches, with protocol decoding.
- Power electronics: Measuring switching waveforms, ringing, dead time and efficiency with differential and current probes.
- Signal integrity: Eye diagrams and jitter analysis on high-speed serial links.
Key takeaways
- Set V/div, s/div and a stable trigger first.
- Bandwidth ≥ 3–5× signal content; t_r,scope ≈ 0.35/BW.
- Compensate probes and use short grounds.
- Never clip the ground to live or floating nodes — use differential probes.
Going further
- ENOB and vertical resolution: 8-bit scopes show 256 levels over the screen; high-resolution (12-bit) scopes and hi-res acquisition modes reveal small ripple on large signals. ENOB falls with frequency.
- FFT windowing: FFT on a scope requires choosing a window (Hann, flat-top, Blackman) to trade amplitude accuracy against frequency resolution and leakage.
- Current probes: Hall-effect clamp probes measure DC–MHz currents; Rogowski coils measure large, fast AC currents with no saturation — both needing deskew against voltage probes for power calculations.