HOW-TO GUIDE
Why Does an Oscilloscope Waveform Shift Vertically?
By the OhmSavvy Research Team ·
Safety first: never move the ground clip to fix a baseline
- On a mains-powered bench scope, every probe ground clip is connected to mains earth. Moving the clip to a point that isn't at earth potential to "fix" a baseline shorts that point to earth and can destroy the probe, the scope or the circuit.
- Never float the scope by removing its earth connection. For signals that aren't referenced to earth, such as anything on the mains side of a power supply, use a rated differential probe or an isolation transformer on the device under test.
The short answer
A trace sits where it does for two reasons: the signal's voltage relative to the probe ground, and where the scope has been told to draw 0 V. Anything that changes either one moves the waveform up or down. These are the usual suspects, roughly from most to least common:
Vertical position
Someone turned the position knob, or the offset moved with a V/div change.
DC coupling
The signal has a DC component, so it sits above or below ground. Correct behavior.
AC coupling
DC is removed and the trace drifts back to center after any level change.
Probe ratio
Channel set to 1x with a 10x probe, or the reverse, scales every level.
Ground reference
Ground clip loose, missing, or on a point that isn't the circuit's 0 V.
Real drift
The signal's own DC level is changing: supply sag, warm-up, coupling capacitors.
1. The vertical position or offset control
Each channel has a vertical position knob that slides the trace and its ground marker (the small arrow at the left edge) up or down together. It doesn't change the signal or the reading, because levels are measured from the marker, not from the center line. Moving it is useful: dropping the ground marker 2 divisions lets a 0–5 V logic signal fill the screen at 1 V/div. But if someone else used the scope last, or you nudged the knob, the trace can appear to have jumped.
On many scopes, pressing the position knob returns the marker to the center line. Some models also hold the offset in volts rather than divisions, so the trace moves when you change V/div. Always read levels from the marker, as explained in how to read an oscilloscope.
2. DC coupling shows the signal's DC level
With DC coupling, the scope shows everything, including any steady voltage the signal sits on. An audio stage biased at half its supply, a sensor output with a 2.5 V midpoint, or ripple on a 5 V rail all appear lifted off the ground line. That isn't a shift; it is the true voltage. Switching to AC coupling removes the DC part and centers the waveform.
AC coupling is the right tool for small variations on a big DC level, which is exactly how ripple voltage is measured. For anything else, start on DC so you see where the signal really sits.
3. AC coupling re-centers the trace, slowly
AC coupling works by putting a capacitor in series with the input, which forms a high-pass filter with the input resistance. It doesn't remove DC instantly. When the signal's DC level steps, for example when a circuit powers up or a load switches on, the trace jumps by the size of the step and then drifts back towards the center line as the capacitor charges. The cutoff frequency, and so the speed of that drift, varies by model; your scope's datasheet lists it.
AC coupling also centers a waveform on its average, not on its low level. A 0–5 V square wave at 50% duty averages 2.5 V, so it shows as ±2.5 V. Change the duty to 20% and the average drops to 1 V, so the same signal now shows as +4 V and −1 V: the whole trace moves up even though the real levels never changed.
This is why logic and PWM signals belong on DC coupling. Very slow square waves on AC coupling also show tilted tops, because the coupling capacitor starts to discharge within each half cycle. More on square-wave set-up in how to measure a square wave.
4. The probe attenuation setting doesn't match the probe
A 10x probe divides the signal by ten before it reaches the scope. The channel's probe setting tells the scope to multiply it back. If the two disagree, every level is scaled by ten, so a DC level appears much closer to ground, or much further from it, than it should. It looks like a vertical shift, but it is a scale error.
| Probe switch | Channel setting | 5 V rail reads as | Fix |
|---|---|---|---|
| 10x | 10x | 5 V | Correct |
| 10x | 1x | 0.5 V | Set the channel to 10x |
| 1x | 10x | 50 V | Set the channel to 1x |
| 1x | 1x | 5 V | Correct, but less bandwidth than 10x |
A quick check is to measure a known voltage, such as a battery or the scope's calibration output, and compare. While you are there, check the probe's compensation.
5. Ground reference problems
The scope measures the voltage between the probe tip and the probe's ground. If that ground isn't the circuit's 0 V, the baseline is wrong:
- Ground clip not connected. The return current finds its way through mains earth and other wiring, so the trace drifts, picks up hum and sits at an unpredictable level.
- Ground clip on a busy return wire. If the clip is on a ground wire that carries load current, that wire's voltage drop appears as an offset. For example, 2 A through 0.05 Ω of ground wiring is 2 × 0.05 = 0.1 V, enough to lift a low-voltage signal visibly. Clip to ground right next to the signal.
- Ground loops. When the circuit is also earthed through another instrument, small currents between the earths can shift and wobble the baseline. The fixes are covered in why a waveform looks noisy.
Remember that on a bench scope all channel grounds are connected together and to earth, so two probes clipped to two different "grounds" short them to each other.
6. Auto/AutoSet changed the settings
The Auto (or AutoSet) button picks V/div, timebase and trigger to fit whatever it finds, and on many scopes it also resets the vertical position. If you had carefully placed the ground marker or set a large offset, one press puts the trace somewhere else. Exactly what Auto changes, including whether it touches coupling, varies by model. After using it, check the channel menu before trusting the levels.
7. The signal itself is drifting
Sometimes the shift is real. Common causes are a supply rail sagging under load, a battery discharging, a circuit warming up after power-on, and coupling capacitors inside the circuit charging after a level change, which produces the same slow drift as the scope's AC coupling. Mains hum mixed into a signal can also look like the baseline jumping between captures at slow timebases.
Prove it is the signal
Troubleshooting table
01
Coupling
Set DC unless you need AC
02
Probe ratio
Match the channel to the probe switch
03
Position
Re-zero the ground marker
04
Ground
Short tip to clip: flat line?
05
Signal
Confirm drift with a multimeter
| Symptom | Likely cause | Fix |
|---|---|---|
| Trace offset but readings from the marker are right | Vertical position moved | Re-zero the position; read from the ground marker |
| DC level missing, trace centred on ground | AC coupling | Switch to DC coupling |
| Trace jumps then creeps back to center | AC coupling settling after a level change | Use DC coupling |
| Trace moves when duty cycle changes | AC coupling centers on the average | Use DC coupling |
| Levels 10 times too small or too large | Probe ratio mismatch | Match the channel setting to the probe |
| Baseline wanders and hums | Ground clip loose, missing or on a busy return | Clip to ground next to the signal |
| Shift appears after pressing Auto | AutoSet reset the position | Re-check the channel menu |
| Same drift on scope and multimeter | The signal is really drifting | Investigate the circuit |
Scopes in our reviews
All of these causes apply to every digital scope. If you are still choosing one, compare the options in our best beginner oscilloscopes roundup, or start with how to use an oscilloscope.
Bench scopes in our reviews
50MHz, 4 channels, 1 GSa/s. Suited to hobbyists, students and general electronics.
100MHz, 4 channels, 1 GSa/s, for projects that will get more demanding.
100MHz, 2 channels, 1 GSa/s; an oscilloscope and waveform generator in one unit.
As an Amazon Associate I earn from qualifying purchases. Prices and availability are shown on Amazon.
Frequently asked questions
Why is my DC voltage showing as 0 V on the oscilloscope?
The channel is almost certainly AC coupled. AC coupling blocks the DC part of the signal, so a steady 5 V rail settles onto the ground line and reads about 0 V. Switch the channel's coupling to DC and the trace moves up to its true level. Also check that the probe ratio matches the probe switch.
How do I reset the vertical position to zero?
On many digital scopes, pressing the vertical position knob returns that channel's ground marker to the center line. Others have a menu option to zero the offset. A full default setup also resets it, but that clears every other setting too. The exact method varies by model, so check your scope's manual.
Why does my waveform jump when I change volts per division?
It depends on how your scope stores the position. Some models hold the offset in divisions, so the trace stays put on screen; others hold it in volts, so the same offset needs a different number of divisions at the new V/div and the trace moves. Neither is a fault. Your manual describes how your model behaves.
Can a damaged probe make the trace shift?
Yes. A broken ground lead or a worn ground connection at the BNC leaves the probe without a proper return, so the trace can wander, pick up hum and sit at the wrong level. Touch the probe tip to its own ground clip: a good probe shows a flat line on the ground marker. Swapping in a second probe confirms it.
Should I run self-calibration if the trace sits off zero?
It is worth doing if the trace rests a fraction of a division away from the ground marker with the input on GND coupling or nothing connected. Let the scope warm up first, often around 30 minutes, unplug every probe, then start self-calibration from the utility menu. It corrects small internal offset errors; it will not fix a faulty probe or a genuine DC level on the signal.
