An oscilloscope current probe is a transducer with its own state, not a passive clip. Residual magnetism left in the core from a previous measurement shifts the baseline, so the probe reads a current that is not there. Degaussing and zeroing before each session is part of using the instrument, not an optional refinement.
Why the core remembers the last thing it measured
A current probe that can see DC contains a magnetic core with a sensing element in the gap. Passing a large current through the core, particularly a current that saturates it, leaves residual magnetization behind when the current goes away. That leftover field looks to the sensor exactly like a small steady current in the conductor, so the displayed trace sits above or below true zero even with nothing clamped. The offset is not a calibration error and it will not be corrected by adjusting the scope’s vertical position, because it drifts as the core relaxes and changes again the next time you measure something large. Degaussing drives the core through a decaying alternating field that leaves it in a neutral state. Most probes automate this with a single button, and it takes seconds.
The order of operations that actually works
Open the jaws with nothing inside them, let the probe warm up for the interval the manual specifies, then run the degauss cycle, then zero the probe. Doing it in the other order wastes the zero, because degaussing itself disturbs the baseline. Close the jaws fully before zeroing; a partly closed or dirty jaw changes the magnetic path and therefore the reading. Only after both steps should you clamp around the conductor. Repeat the sequence whenever the probe has been used near its maximum rating, whenever the instrument has been moved between very different temperatures, and at the start of any session where the absolute value of a DC current matters. For purely AC measurements the offset matters less, but a drifting baseline still eats into the display range you have available.
Insertion impedance is small but not zero
Clamping around a conductor is often described as non-invasive, and compared with breaking the circuit for a shunt it certainly is. It is not free, though. A current probe couples to the conductor and presents an insertion impedance that the circuit sees, and that impedance rises with frequency. In a low-impedance power loop the effect is usually negligible. In a fast switching circuit, a sensitive gate drive loop, or a high-frequency resonant path, the added impedance can change the waveform you are trying to observe. Manufacturers publish an insertion impedance curve for this reason. It is also worth remembering that the probe needs a single conductor inside the jaws; clamp around a cable containing both supply and return and the fields cancel, leaving you measuring nothing but leakage.
Ratings that get exceeded without anyone noticing
Current probes carry several separate limits, and it is easy to stay inside one while violating another. There is a continuous current rating, which falls as frequency rises because core losses heat the probe. There is a peak rating for brief events. There is an amp-second or amp-hertz product limit on some models, which restricts large currents at low frequency. There is a voltage rating on the insulation of the conductor you are clamping, because the probe’s jaws are not designed to be an insulation barrier against bare conductors. Exceeding the current limits saturates the core, which both clips the reading and leaves the magnetization that forces another degauss. Exceeding the insulation rating is a safety failure. All of these are in the probe’s documentation and none can be inferred from the jaw size.
Reading the trace without fooling yourself
Current probes have a stated sensitivity in amps per volt or volts per amp, and many scopes let you enter it so the vertical axis reads in amps directly. If the scope’s probe setting does not match the probe’s attenuation switch, every current figure is wrong by that factor while the waveform shape looks entirely plausible. Check the setting against the probe’s label before recording anything. Orientation matters as well: the jaws carry an arrow indicating assumed current direction, and clamping backward inverts the trace, which can make a perfectly normal waveform look like a fault. Finally, a current probe’s bandwidth is usually well below the scope’s, so the current waveform may be noticeably smoother than the voltage waveform on the same event, and that smoothing is the probe, not the circuit.
What to check before you act
- Warm up, degauss with open jaws, then zero, then clamp.
- Clamp around one conductor only, with the arrow pointing along current flow.
- Confirm the scope’s probe sensitivity matches the probe’s switch position.
- Check continuous, peak and insulation ratings in the probe’s manual.
Common questions
How often does a current probe need degaussing?
At the start of every session and after any measurement near the probe’s rating. If the baseline moves when you open and close the jaws with nothing inside, degauss again.
Why does my current trace look slower than the voltage trace?
Current probes usually have lower bandwidth than the scope and its voltage probe, so fast features are rounded. Compare the probe’s rated bandwidth against the edge speed you are observing.
The practical takeaway
Treat a current probe as an instrument that needs preparing, not a clip you attach and read. Warm it, degauss it, zero it, then clamp one conductor the right way round, and verify the scope agrees with the probe’s sensitivity. A baseline that will not sit still is telling you the core is magnetized, not that the circuit is doing something strange.
References and further reading
- CDC/NIOSH: electrical hazards and safe work practices
- NFPA 70E: standard for electrical safety in the workplace
- OSHA: 1910.304 wiring design and protection requirements
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
