Oscilloscope Basics: Probes, Triggering, Bandwidth and Measurements

A multimeter gives you one number. An oscilloscope shows how a voltage changes with time, so you can see ripple on a power rail, the shape of a PWM signal, ringing on a switching edge or the timing between two signals. Modern digital storage oscilloscopes (DSOs) are affordable for college labs and hobbyists, but their many buttons can be confusing. This tutorial covers the controls you will use every day, how to read the screen, and the limits of the instrument and its probes.

Reading the Screen

The display is a grid, usually 10 divisions horizontally and 8 vertically. The vertical axis is voltage, set by the volts/div knob for each channel; the horizontal axis is time, set by the time/div (or sec/div) knob shared by all channels. Every measurement on the screen is simply divisions multiplied by the scale.

Voltage = number of vertical divisions x volts/div x probe factor*
Time    = number of horizontal divisions x time/div
Frequency = 1 / period
(*modern scopes apply the probe factor automatically if set correctly)

Worked example: reading a square wave

A square wave spans 3.3 divisions vertically at 1 V/div, and one full cycle spans 4 divisions horizontally at 0.5 ms/div. The high part of each cycle spans 1 division.

Amplitude = 3.3 x 1 V   = 3.3 V peak-to-peak
Period    = 4 x 0.5 ms = 2 ms
Frequency = 1 / 2 ms   = 500 Hz
Duty cycle = 1 / 4     = 25%
ControlWhat it doesTypical starting setting
Volts/divVertical scale per channel1 V/div for 3.3 V or 5 V logic
Vertical positionMoves trace up or downGround marker at a grid line
Time/divHorizontal scaleShow 2 to 5 cycles
Coupling (DC / AC)AC blocks the DC componentDC normally; AC to view small ripple on a DC rail
Trigger level and edgeSelects the point where the trace startsRising edge, level at half amplitude
Trigger modeAuto, Normal, SingleAuto to find the signal, Single for one-off events

Probes and Compensation

Most scopes come with switchable x1/x10 passive probes. In the x10 position the probe forms a 10:1 divider with the scope's 1 MΩ input, presenting about 10 MΩ to your circuit and much less capacitance than x1. It also gives far more bandwidth: x1 mode on a typical probe is limited to a few megahertz. Use x10 for almost everything and set the channel's probe setting to match, or every reading will be off by a factor of ten.

A x10 probe has an adjustable capacitor that must be matched to the scope input. Connect the probe to the scope's calibration terminal, which outputs a square wave (often 1 kHz), and adjust the small screw on the probe until the corners are square. Over-compensation shows overshooting spikes at the corners; under-compensation shows rounded corners. An uncompensated probe distorts every measurement of fast edges.

Triggering

Without a trigger, each sweep starts at a random point and the waveform rolls or smears. The trigger tells the scope to start each sweep when the signal crosses a chosen level in a chosen direction. Set the source to the channel carrying the signal, the edge to rising, and the level to about half the amplitude. If the trace still jitters, the level is outside the signal range or the source is wrong.

  • Auto mode: shows a trace even with no trigger, useful for finding DC levels and lost signals.
  • Normal mode: updates only on a valid trigger, good for intermittent signals.
  • Single mode: captures one event and stops, ideal for a power-on sequence or one button press.

Bandwidth, Rise Time and Sample Rate

A scope's bandwidth is the frequency at which a sine wave reads 3 dB low, about 70.7% of its true amplitude. A 100 MHz scope therefore under-reads a 100 MHz sine by about 30%. For accurate amplitude, choose a bandwidth several times higher than the highest frequency of interest. For edges, the relationship between bandwidth and rise time is useful:

Scope rise time:  tr_scope ~ 0.35 / BW
  50 MHz scope  -> 7 ns
  100 MHz scope -> 3.5 ns

Measured rise time (approximately):
  tr_measured = sqrt(tr_signal^2 + tr_scope^2)

Example: true edge 5 ns on a 100 MHz scope
  tr_measured = sqrt(25 + 12.25) = 6.1 ns

Sample rate is a separate specification. A scope needs several samples per cycle, and more for edges, to reconstruct a waveform; a common guideline is a sample rate of at least four to five times the bandwidth. Note that on many scopes the quoted maximum sample rate is shared between channels, so it drops when two channels are on.

Practical Measurements and Mistakes

  • Measure power-supply ripple with AC coupling at 10 or 20 mV/div, and use the short ground spring instead of the long ground lead to avoid picking up noise.
  • Use the built-in measurements (Vpp, Vrms, frequency, duty, rise time) but confirm them roughly with the grid.
  • A flat line at zero usually means the probe is at x1 while the scope is set to x10 (or the reverse), the channel is off, or coupling is set to GND.
  • Ringing on edges is often caused by the long ground lead itself; repeat the measurement with the ground spring before blaming the circuit.
  • Compare two signals by using channel 1 as the trigger source and reading the time difference with cursors.

Summary

Read voltages and times by multiplying divisions by the scale, keep probes in x10 and compensated, and trigger on a clean edge at mid-level. Respect the bandwidth and rise-time limits when judging fast signals, and never clip the ground lead onto a point that is not at earth potential. With these habits, the oscilloscope becomes the instrument that shows you what your circuit is actually doing.