A portable graphing multimeter can reveal whether a 24 VDC problem is a slow supply drift, a switching transient, a load-induced sag, or intermittent wiring resistance. The useful purchase is not simply the instrument with the largest advertised bandwidth. Select the input architecture, probes, recording depth, triggering, and safety rating around the complete measurement path.
What measurement path must be defined first?
Follow the signal from the power supply output, through protection and distribution wiring, to the load terminals. A measurement at only one point cannot distinguish a weak source from voltage lost in a conductor, connector, fuse holder, or return path.
| Test point | What it reveals | Preferred capture |
|---|---|---|
| Supply output | Source regulation, shutdown, or output transient | Voltage versus time |
| Load input | Voltage actually reaching the device | Voltage versus time, triggered by the symptom |
| Supply-to-load positive path | Dynamic voltage drop in protection and wiring | Differential voltage during load operation |
| Load return-to-supply return | Return-path resistance or common shift | Differential voltage during load operation |
| Load current | Whether voltage collapse follows inrush or a state change | Current probe or suitable external measurement method |
Two channels are valuable because one can monitor the supply while the other monitors the load. If both fall together, investigate the source or a shared upstream path. If the supply stays stable while the load voltage falls, move downstream through the wiring. Confirm this plan by listing the source, every connection hop, the load, and the intended reference point before choosing probes.
Which input arrangement can connect safely?
Layer one first. Battery operation removes a direct protective-earth connection from many handheld instruments, but it does not prove that every input is isolated. Oscilloscope channels may share a common terminal internally, and the meter input may have different limits from the scope input. Connecting two shared-reference probe clips to different potentials can short those points through the instrument.
| Arrangement | Primary concern | Selection check |
|---|---|---|
| DMM leads | Input rating, transient category, and fused current terminals | Read the marked terminal limits and manual |
| Single-ended scope probe | Probe reference may be common to another channel or an external port | Check channel-to-channel and input-to-port isolation |
| Differential probe | Both differential and common-mode limits must cover the circuit | Read the probe ratings, not only the scope rating |
| Instrument operating from its battery | Floating enclosure does not guarantee isolated inputs | Verify the internal input architecture |
| USB charging or PC connection | The cable can create a ground path | Disconnect USB during live measurements unless the documented isolation permits it |
The evaluated Hantek unit was convenient on battery power, but connecting USB for charging introduced a ground path. Treat every charger, communications cable, signal-generator terminal, and probe reference as part of the circuit. Confirm safe connection by drawing the resulting current path with every cable attached; no instrument lead may bridge two unintended potentials.
How much bandwidth and recording capability are useful?
A graphing DMM suits slow changes that remain visible at its update or logging rate. A scope is required when the suspected event is shorter than the meter can record, when event shape matters, or when trigger timing must be related across two points. Bandwidth alone does not define capture quality; sample rate, memory depth, trigger functions, vertical resolution, input range, and probe bandwidth all affect the result.
| Symptom | Capability to prioritize | Reason |
|---|---|---|
| Gradual 24 VDC drift | Trend logging and usable recording duration | The event develops slowly |
| Occasional voltage sag | Single-shot trigger and pre-trigger history | The instrument must retain events that occur before the observed failure |
| Fast spikes or ringing | Probe suitability, sample rate, bandwidth, and short wiring | The measurement setup can attenuate or create high-frequency detail |
| Supply-versus-load comparison | Two simultaneous channels | Time correlation separates source and path faults |
| Long unattended observation | Memory depth, logging, battery duration, and data export | A brief screen trace may miss or overwrite the event |
The selected Hantek configuration advertised 110 MHz bandwidth, two channels, graphing in meter mode, a built-in signal generator, PWM capability, and CAN bus display. Those features exceed the immediate need if the task is only slow 24 VDC trending, but two-channel triggering adds direct diagnostic value. Used Fluke 97 units were identified as an adequate lower-cost option; Fluke 99 and 196 units, Siglent SHS800 and SHS1000 instruments, a Hioki logger, and a Fluke 435 were also cited as alternatives with different cost and portability tradeoffs. Confirm suitability with the manual specifications for input ratings, channel isolation, capture memory, and export before purchase.
How should the first capture be configured?
Start with a known-safe voltage range that keeps the normal 24 VDC level on screen and leaves headroom for excursions. Use DC coupling when the objective is total bus voltage. Use a second, deliberately scaled view only when examining ripple; removing the DC component can make small disturbances easier to see but hides the absolute bus level.
| Setting | Initial choice | Adjustment rule |
|---|---|---|
| Vertical range | Display nominal voltage with transient headroom | Tighten only after a clean baseline capture |
| Time span | Cover one complete machine or load event | Shorten it to inspect the captured edge |
| Trigger source | Channel nearest the observed symptom | Use the other channel for upstream comparison |
| Trigger direction | Falling edge for a suspected sag | Use rising edge for an overvoltage event |
| Acquisition mode | Single capture for intermittent events | Use continuous display while establishing levels |
Do not select an exact trigger level before observing the healthy minimum and failed waveform. Place the trigger between those measured levels, then retain enough pre-trigger data to show what changed first. Confirm the setup by producing a deliberate, normal load transition and checking that both channels capture it without clipping or losing the event.
How should the probes follow the disturbance?
- Operate the instrument from its battery and remove USB or other external connections unless their isolation is documented for the measurement.
- Inspect probe condition, compensation where applicable, terminal assignment, and instrument mode before touching the circuit.
- Measure directly across the supply output and record the baseline.
- Move to the load terminals without changing the time reference. Capture the same operating transition.
- If the load trace differs, measure across one wiring segment at a time, including the return path. A voltage appearing across a connection during load current identifies where the path is dropping voltage.
- For a two-channel comparison, verify whether the channel references are common before attaching both clips. Use a properly rated differential method when the two references cannot be bonded.
- Repeat the event with short probe connections. A long ground lead can add inductive pickup and display ringing that is not present at the test point.
Graphing mode is useful for locating a slow event; scope mode can then resolve its edge and sequence. Confirm each move by reproducing the same machine state and retaining a labeled capture that identifies the exact test points.
How is the fault and the instrument verified end to end?
First prove that the instrument can display a stable known source and that channel offsets are reasonable with both inputs connected to the same reference. If a built-in signal generator is used for a functional check, keep it separate from the live 24 VDC circuit unless the manual explicitly permits that connection.
| Observation | Diagnostic direction | Verification after correction |
|---|---|---|
| Supply and load fall together | Source, shared protection, or upstream feed | Repeat the triggering load event at both points |
| Only load voltage falls | Distribution or return-path drop | Measure the repaired segment and load voltage simultaneously |
| Trace changes when USB is connected | Measurement-induced ground path | Repeat on battery with external cables removed |
| Ringing changes with probe-lead position | Probe-loop pickup or connection inductance | Repeat with shorter connections |
| No captured event despite a repeatable symptom | Trigger source, level, direction, time span, or memory configuration | Trigger successfully on a controlled normal transition before waiting for the fault |
Keep the same test points, scaling, trigger, load sequence, and instrument connection state for before-and-after captures. The repair passes when the original symptom no longer occurs and the source and load traces remain within the equipment limits read from their documentation throughout the repeated event.
FAQ
Can I use a graphing multimeter to find a 24 VDC dropout?
Yes, when the dropout lasts long enough for the meter's logging rate to record it. Use scope mode with a falling-edge single-shot trigger when the graph misses the event.
Can I connect a battery-powered scope anywhere in the circuit?
No. Battery power may remove a protective-earth path, but channels can still share a common reference and input limits still apply. Verify isolation and common-mode ratings in the manual before connecting probes.
Does USB charging change scope grounding?
It can. The tested Hantek arrangement acquired a ground path while connected for USB charging, so disconnect USB during live measurements unless the documented port isolation permits the connection.
Can I diagnose wiring loss with two scope channels?
Yes, compare supply and load voltage during the same event, provided the channel-reference arrangement is safe. If the supply remains stable while the load falls, measure successive positive and return-path segments.
Does a stable trace prove the 24 VDC fault is fixed?
Only if the original operating event has been repeated under the same measurement conditions. Perform the final verification at both the supply and load terminals with the original trigger, scaling, load sequence, and external-cable state.