Oscilloscope probe compensation is the adjustment that makes a divider behave the same way at high frequency as it does at DC. Get it wrong and every square edge on screen is exaggerated or rounded by the probe itself. The trimmer is not a factory setting you can ignore; it belongs to one probe on one input.
What the compensation trimmer is actually adjusting
A 10x probe is a resistive divider, and a resistive divider alone would be accurate only at DC. Every scope input has capacitance, every cable has capacitance, and those capacitances form a second, frequency-dependent divider working in parallel with the resistive one. The two dividers only agree if their time constants match. The adjustable trimmer inside the probe body or compensation box sets a small capacitance so that the capacitive ratio equals the resistive ratio. When they match, the division ratio is flat across frequency and a square wave passes through unchanged. When they do not, low-frequency and high-frequency content are divided by different amounts, so the shape of every fast transition is altered before the scope ever digitizes it. This is a passive, mechanical adjustment, and nothing about it is automatic.
How to read the calibrator square wave
Scopes provide a square-wave output on the front panel for this purpose. Clip the probe to it, set a comfortable timebase, and look at the corners. A flat top with square corners means the compensation is correct. A top that slopes upward into a sharp spike at each transition means the probe is overcompensated, passing too much high-frequency content. A top that sags away from a rounded corner means it is undercompensated. Turn the trimmer with the plastic tool supplied, not a metal screwdriver, because your hand and a metal blade both add stray capacitance and shift the very thing you are adjusting. Move in small steps, remove the tool, and look again. Aim for flat rather than for whatever looks fastest; a spike is not a crisper signal, it is an error.
Why compensation belongs to one probe on one input
Input capacitance varies between scope models, between channels on the same scope, and slightly with the input coupling and attenuation settings. A probe compensated on channel one is not guaranteed to be correct on channel two, and a probe compensated on one instrument is definitely not guaranteed on another. Good practice is to compensate each probe on the channel it will be used on, at the start of a session, and again if the probe has been swapped between inputs. Compensation also drifts over time as the trimmer ages and as cable handling changes stray capacitance, so it is worth a quick check whenever measurements start disagreeing with expectations. Some probes include a dedicated low-frequency and a high-frequency adjustment, and those must be set in the order the manufacturer specifies rather than by trial.
What bad compensation does to real measurements
Overcompensation exaggerates edges, producing overshoot and ringing that are not present in the circuit. Chasing that artifact wastes hours, and worse, it can lead someone to add damping components to fix a problem that never existed. Undercompensation does the opposite, smoothing genuine overshoot into something that looks acceptable and hiding a real signal-integrity issue. Both distort amplitude measurements on pulses, because the peak the scope measures is not the peak the circuit produced. Rise-time readings are affected too, since the shape of the transition has changed. None of this shows up on a slow sine wave, which is why a probe can appear perfectly fine for routine voltage checks and be badly wrong the moment you look at a digital edge. Compensate before you trust any waveform shape.
When the flat top will not come no matter what
Sometimes the trimmer reaches its end of travel and the top still slopes. That usually means the probe and the scope input are mismatched: probes are designed for a particular range of input capacitance, and a probe intended for one family of instruments may simply not be adjustable onto another. Check the probe’s specified compensation range against the scope’s input capacitance, both of which are published. A damaged cable, a cracked compensation box, or a corroded trimmer can also prevent adjustment, and in those cases the probe should be replaced rather than forced. Distortion that changes when you flex the cable points to a mechanical fault. A probe that cannot be compensated is not a probe with a small error; it is an instrument returning shapes you cannot interpret, and it belongs out of service.
What to check before you act
- Compensate each probe on the channel and attenuation setting you will use.
- Use the supplied plastic tool, not a metal screwdriver.
- Adjust for a flat top, not for the sharpest looking corner.
- Replace any probe that cannot reach flat within its trimmer travel.
Common questions
Does a 1x probe need compensation?
Most 1x-only probes have no trimmer because there is no divider to balance, but they also load the circuit far more heavily. Switchable probes must be compensated in their 10x position.
How often should I recompensate?
At the start of a measurement session, after moving a probe to a different channel or instrument, and any time a waveform shape looks unexpected. It takes under a minute.
The practical takeaway
A scope probe is part of the circuit and part of the measurement, and compensation is what makes it neutral. Clip onto the calibrator, adjust for a genuinely flat top with the plastic tool, and repeat it whenever the probe moves. Any overshoot you see after that is worth investigating; any overshoot you see before it is not evidence of anything.
References and further reading
- NFPA: electrical safety resources and code background
- CDC/NIOSH: electrical hazards and safe work practices
- FCC: frequently asked questions on radio frequency signals
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
