Specifications

55
Good measurement practices
These are provided as guidelines of good practice, but may not be true in all situations.
1. Divide the bandwidth by 10 to get a rough idea of the fundamental frequency of an arbitrary
periodic signal that your scope will be able to reasonably reproduce.
2. Time measurements are generally relative with oscilloscopes -- thus, you usually subtract two
times measured on the screen to get a time difference. This is because there's no inherent
notion of time t = 0 (unless you define the trigger event as t = 0). This principle may not be
applicable to digital scopes, as they have more sophisticated time measurement features than
analog scopes.
3. Voltage measurements are referenced to power line ground for the majority of non-battery-
operated scopes.
4. If you measure voltages and times on the scope's screen, make the signal take up as much
space on the screen as possible in both directions to get the best measurement resolution and
accuracy.
5. Scope probes become part of the circuit they're measuring and thus load the circuit, especially
at higher frequencies. The impedance of a passive scope probe can drop by three or four orders
of magnitude from DC to the probe's rated bandwidth.
6. An uncompensated scope probe is a measurement error waiting to happen: the measured
waveform may be distorted and you may get incorrect amplitude measurements.
7. A scope probe's rated maximum voltage can drop by two or more orders of magnitude from DC
to its rated bandwidth.
8. Never connect the ground lead of a scope probe to anything except a point that is at ground
potential or an isolated point in a circuit that can be at ground potential.
9. Do not use a grounding lead/clip on a scope probe at frequencies above roughly 100 MHz. Use
the probe's spring clip (bayonet-style) or a coaxial probe socket on a printed circuit board.
10. For a non-battery-operated scope, do not float the scope above ground potential by the use of a
3-to-2 wire adapter or an isolation transformer. If you need to float a battery-operated scope,
consult the scope's user manual as to whether this is appropriate.
11. Learning the screen annotations for your digital scope well (and the control locations and
positions of your analog scope) will help you quickly understand the measurement conditions.
12. Every time you connect a probe to a scope, compensate it and make sure the probe's
attenuation is set properly in the corresponding channel's menu. If appropriate, make sure the
probe's attenuation switch is in the position you need.
13. If you use coaxial cables for making connections, get in the habit of using pass-through
terminations for the scope's inputs, as this may help avoid measurement mistakes in the future.
14. Noise can't be distinguished from the signal unless the noise spectrum and the signal spectrum
have an intersection of zero. The best strategy is to not let the noise into the system in the first
place if possible.
15. If you're using an arbitrary waveform generator for stimulation of a circuit or system, put in a
short, tall pulse in the waveform if the system will allow it. This pulse can make it easier to
trigger a scope to get a stable display. If this isn't possible, but your scope supports e.g. pulse
triggering, you can put in a low amplitude narrow pulse and try to trigger on that waveform. If
your arbitrary waveform generator has a marker signal output, use the marker signal as the
external trigger for the scope.
16. You can estimate the standard deviation of a signal using a digital scope. Here are two possible
ways. First, turn on persistence and measure the overall envelope of the waveform for a time
interval. The maximum and minimum values of this envelope will give you the range. You can
then estimate the standard deviation, either by assuming the distribution or using a non-
parametric method. The second way is to measure the RMS value. The RMS value of a
waveform is the same as the population standard deviation of the waveform with the DC
component subtracted out.
All content Copyright © B&K Precision Corporation, except where otherwise noted.
Copying/reprinting/republishing without express written consent prohibited.