A differential probe for oscilloscope measurements exists because both scope inputs share one ground, and that ground is usually bonded to the safety earth. Subtracting channel A from channel B does not change where the ground clips are connected. It is a useful trick with narrow limits, and it is not a substitute for a probe designed for the job.
Why a scope’s inputs share one ground
In a conventional bench oscilloscope the outer shell of every input connector is tied together and, through the power cord, to protective earth. That is a deliberate safety design: it keeps the chassis at earth potential and it gives every measurement a stable reference. The consequence is that the probe ground clip is not a neutral sensing wire but an earth connection you are attaching to the circuit. Clip it to a point that is not at earth potential and you create a fault path through the instrument. On a low-voltage board that may be a blown trace or a damaged scope. On mains-referenced equipment it is a hazard to the person holding the probe. This one property drives almost everything else about how scope measurements on power circuits have to be arranged.
What channel subtraction can and cannot reject
Using two probes and displaying A minus B genuinely removes the component common to both inputs, and for low-voltage work on a board it is often perfectly adequate. The limits appear quickly. Both probes still have their ground clips attached somewhere, so nothing about the earth connection has changed. The subtraction is performed after each channel has already digitized its input, so if a large common-mode voltage drives either channel beyond its range, the information needed for the subtraction is already clipped and lost. The two channels must also be closely matched in gain, offset and delay, because any mismatch converts common-mode signal into apparent differential signal. Small differences that are invisible on a large waveform become the entire result when you are looking for a few millivolts riding on a large common signal.
Common-mode rejection is a specification with limits
A true differential probe has two high-impedance inputs and an amplifier that responds only to the difference between them, with neither input referenced to the probe body. Its ability to ignore the shared voltage is stated as a common-mode rejection ratio, and that figure is not constant: it is highest at DC and low frequencies and degrades as frequency rises. Manufacturers publish a curve or a set of figures across frequency for exactly this reason. A probe with excellent rejection at power frequency may reject far less at the switching frequency of an inverter, which is often precisely where you are looking. Probes also state a maximum common-mode voltage, above which rejection is irrelevant because the input stage is overloaded. Read both specifications before deciding a probe suits the measurement you have in mind.
Floating the scope is the dangerous shortcut
The old workaround of defeating the ground pin on the scope’s power cord, or feeding it through an isolating transformer, is genuinely dangerous and should not be done. It lifts the entire chassis, every connector shell, and every exposed metal part of the instrument to whatever potential the probe ground is touching. Anyone who then touches the scope, a nearby earthed object, or a second instrument connected to the same circuit completes a path they never agreed to. It also removes the protection that would otherwise clear a fault. Measuring on energized mains-referenced equipment requires appropriate training, appropriate personal protective equipment, and instruments rated for the job, which for oscilloscope work means a properly rated differential probe or an isolated-input instrument. If you are not trained and equipped for live work, the correct answer is to stop, not to improvise.
Choosing a probe for the voltage you actually have
Differential probes carry ratings for differential voltage, common-mode voltage, and a measurement category that describes the transient environment they are built to survive. Those ratings apply to the probe together with the specific tip accessories supplied with it, which is why substituting generic leads invalidates them. Attenuation settings on the probe often change the voltage rating as well, so a probe safe at one setting is not automatically safe at another. Match the probe’s category rating to the installation you are working on rather than to the nominal voltage alone, because the transient a probe must survive depends on how close it sits to the incoming supply. When a probe’s ratings do not clearly cover the circuit, the probe is the wrong tool and the measurement needs to be redesigned.
What to check before you act
- Never defeat the scope’s earth connection to float the chassis.
- Check the probe’s common-mode voltage limit and its category rating.
- Confirm rejection performance at your signal’s frequency, not just at DC.
- Use only the tip accessories the probe’s ratings were established with.
Common questions
Is channel subtraction ever acceptable?
Yes, for low-voltage circuits where both ground clips can safely sit at earth potential and the common-mode signal is small. It is not acceptable as a way to measure across mains-referenced components.
Does a battery-powered handheld scope solve the problem?
Isolated-input instruments remove the shared-earth constraint, but they still carry voltage and category ratings that must match the circuit. Check those ratings rather than assuming battery power alone makes a measurement safe.
The practical takeaway
The ground clip on a scope probe is an earth connection, and no amount of on-screen math changes where it is attached. For small signals on low-voltage boards, channel subtraction is a reasonable tool. For anything referenced to the mains, use a correctly rated differential probe, confirm its ratings cover the circuit, and treat energized work as the trained, protected activity it is.
References and further reading
- NFPA 70E: standard for electrical safety in the workplace
- OSHA: Subpart S, electrical requirements including safe work practices
- CDC/NIOSH: electrical hazards and safe work practices
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
