Oscilloscope calibration: what we actually check
Bandwidth, rise time, vertical gain, timebase and trigger. A walk through an oscilloscope calibration and why each test matters for the measurements you make.
Oscilloscopes are among the most versatile instruments on any bench, and among the least understood when it comes to calibration. A scope that displays a clean waveform is not necessarily measuring it correctly. Here is what happens to your oscilloscope when it comes into our lab.
Preparation and self-calibration
Before any measurement, the scope is left to warm up and stabilise in our controlled environment, typically for at least 30 minutes, following the manufacturer’s recommendation. Most modern digital scopes also have an internal signal path compensation or self-calibration routine, which we run after warm-up. This routine corrects for small internal offsets but does not replace external calibration: it cannot tell whether the scope’s internal references are correct.
Vertical: DC gain and offset
We apply precise DC voltages from an oscilloscope calibrator to each channel at several volts-per-division settings and compare the scope’s measurement with the applied value. This checks DC gain accuracy, typically specified at around ±1.5 % to ±2 % of full scale on general-purpose instruments.
We also check offset accuracy and the position of the trace with no signal applied. Errors here affect every amplitude measurement you make.
Bandwidth
Bandwidth is the frequency at which a sine wave’s displayed amplitude has fallen by 3 dB (to about 70.7 %) of its low-frequency value. We use a leveled sine wave source, which holds its amplitude constant across frequency, and sweep up to the scope’s rated bandwidth on every channel.
A scope that has lost bandwidth will under-report the amplitude of fast signals and round off edges, which can hide ringing and overshoot.
Rise time
Closely related to bandwidth, rise time is checked with a fast-edge pulse generator whose own rise time is much faster than the scope’s. For many scopes there is a useful approximation:
Rise time ≈ 0.35 ÷ bandwidth
So a 500 MHz scope should show a rise time of roughly 700 ps or faster. Significant deviation from the specification usually indicates a problem in the front-end amplifier.
Timebase accuracy
Every time measurement a scope makes depends on its internal timebase. We apply a time-marker signal from a reference traceable to our GNSS-disciplined frequency standard and measure the scope’s timing error, typically specified in parts per million.
Trigger
We check trigger sensitivity at low and high frequencies, on internal and external trigger inputs where fitted. A scope that will not trigger reliably on small signals is effectively unusable for many tasks, even if every other parameter is in specification.
Input impedance
The 1 MΩ and 50 Ω input impedances are measured directly. An out-of-tolerance 50 Ω termination introduces amplitude errors and reflections that can be mistaken for problems in the circuit under test.
What about probes?
Passive probes have their own attenuation and compensation characteristics, and active probes have offset, gain and bandwidth specifications. If you rely on a probe for quantitative measurements, send it with the scope and ask for it to be calibrated as a system. A perfect scope with a damaged probe is still a poor measurement.
