DMM Mastery

Study an annotated diagram

The digital multimeter is the instrument you'll touch a hundred times a day. Open visuals/03-test-equipment.html ↗ for the annotated front panel and probe-placement diagrams. This module makes every mode second nature and — more importantly — teaches the in-circuit traps that separate accurate techs from part-swappers.

1. The cardinal rules

1. Voltage: across. Current: through. Resistance: power off.
2. In current mode the meter is nearly a short circuit. Putting an ammeter across a supply is the classic blown-fuse (or blown-board) mistake. Always break the circuit and insert the meter in series — and put the probes back in the V jacks when done. Leaving a probe in the A jack and then measuring "voltage" shorts whatever you touch.
3. Never measure ohms/continuity/diode on an energized circuit. Isolate all sources, including batteries and stored energy, using the approved procedure. Verify absence of voltage with an appropriate instrument and method before resistance work. Discharge components only with the specified equipment, sequence and verification: resistance, voltage rating, peak power, pulse energy and insulation all matter. A generic resistor value or a universal residual-voltage threshold is not a safe procedure. Never short a capacitor with a tool.
4. Trust, but verify the meter: before declaring something open/dead, touch your probes together (should beep, ~0Ω) and check the meter battery. Before declaring "no voltage," verify the meter on a known-live source. "Test your tester" is a habit that prevents false diagnoses.

2. Mode by mode

DC Volts (V⎓) — your primary mode

• Black probe to ground (clip it — free hand matters), red probe to the point under test.
• Auto-range is fine for bench work. The meter's ~10MΩ input barely loads most circuits.
• What you're usually doing: verifying rails (against the applicable drawing or test limits; there is no universal ±5% acceptance rule), checking regulator in vs out, measuring across a component to see what it drops.
• Millivolts matter: mV mode reveals voltage drops along traces and across closed contacts/fuses — a "0Ω" path carrying current shows a few mV; a corroded one shows a lot more.

AC Volts (V~)

• AC mode is valid only within the meter's specified frequency, crest-factor and input limits. This course's independent practice excludes mains and aircraft power. Those measurements require task-specific electrical-safety training, rated equipment and the approved procedure.
• DMM AC bandwidth is low (≈ kHz) — it cannot meaningfully measure high-frequency ripple or signals; that's scope work.

Resistance (Ω)

• Power off, caps discharged. For a trustworthy value, isolate the component — lift one leg or one end of an SMD, or rely on the in-circuit rules below.
• In-circuit interpretation rules:
  • Reading ≈ marked value → probably fine.
  • Reading lower than marked → maybe fine (parallel paths) — isolate to confirm before condemning.
  • Reading higher than marked → genuinely suspect: the part or its joints are open/degraded. Parallel paths can't raise a reading.
  • A reading that climbs slowly → you're charging a capacitor on the net; normal.
  • Reverse the probes and re-measure: a big difference means semiconductor junctions are in the path — interpret accordingly.

Continuity (beeper)

• Beeps below ~30–50Ω (meter-dependent). For: fuses, traces, connector pin-to-solder-tail, cable cores, "is this net connected to that pin?"
• Trace verification against the schematic is a core repair task: net on schematic → beep every point that should be common. No beep where there should be one = open trace/via/joint. Beep where there shouldn't = short.
• The beeper is fast but crude — it can beep through a 30Ω fault. For "is this joint really good," use ohms or mV-drop under load.

Diode test (the secret weapon)

• Sources ~1mA, displays forward voltage. Healthy silicon junction: 0.5–0.7V forward, OL reverse. Schottky: 0.15–0.45V. LED: 1.6–3.3V (lights faintly). Germanium (old equipment): 0.2–0.3V.
• A low reading in both directions may be a shorted junction or a parallel path. OL both ways may be normal for a bidirectional TVS or a device beyond the meter's test voltage; it does not universally mean open. Check the device type, datasheet and isolated/reference measurement. A diode-mode test does not verify TVS breakdown or clamping performance. IC input protection structures vary: signature comparisons are clues on a matched, de-energized assembly, not proof that every pin has rail diodes.
• Rail short-finding: diode test (or ohms) from each power rail to ground, power off. A healthy 3.3V rail might read a few hundred Ω to a few kΩ (and charge caps); a dead short reads near 0V/0Ω — then the hunt begins (section 4).

Capacitance

• Out-of-circuit (or one leg lifted) only — in-circuit, everything else on the net adds to the reading. Good for "is this 100µF cap still 100µF," useless for ESR (the more common electrolytic failure) — that needs an ESR meter or scope-and-ripple judgment.

Current (µA/mA/A)

• Break the circuit, insert meter in series, correct jack, correct range. The meter's internal shunt drops some voltage ("burden voltage") — on low-voltage circuits this can disturb operation.
• On the bench you'll more often get current from the bench supply's readout or by measuring mV across a known resistor in the path (I = V/R) — both avoid breaking the circuit.
• Supply current is a diagnosis in itself: a board that should draw 150mA drawing 2A has a short; drawing 5mA has something not starting. Know the expected draw (test spec) before powering any UUT (unit under test).

3. Measurement discipline

• Do not use a one-hand habit as authorization for energized work. Electrical risk depends on accessible voltage, available energy, environment and equipment. Follow the facility's electrical-safety procedure; do not defeat protective earth, interlocks or required barriers.
• Probe on the pin/pad, not the trace — fine probes slip; a slipped probe bridging two IC pins can kill a good board. Use micro-grabbers/hooks on fine-pitch work, or probe at vias and test points.
• Conformal-coated boards: probe at designated test points, or use approved coating-piercing technique — and log any coating penetration for repair.
• Write numbers down. "5V rail was 4.82V" is data; "rail looked okay" is not.

4. Technique: finding a short to ground on a rail

The dreaded "rail measures 0.3Ω to ground" with 60 capacitors on the net:

1. Look and smell first — burnt part, discolored mask, cracked MLCC.
2. Divide by geography: if the rail crosses removable links (0Ω resistors, ferrite beads, connectors), open them and re-measure each side — instant half-split.
3. Millivolt-gradient localization (supervised shop technique): only use an authorized injection procedure with all normal sources isolated and reviewed limits for the entire net, connected devices, backfeed paths, current, voltage, energy and duration. There is no universal 1–2 A setting. Record voltage drop along known conductors with a defined return/reference; follow current flow toward the branch containing the fault. A small difference can be contact error, not localization.
4. Thermal observation: use approved non-contact imaging within the authorized power and temperature limits. A warm regulator may be supplying a fault elsewhere; warmth alone does not identify a failed component. Do not use a fingertip or flammable solvent to locate a powered fault. Cooling or mechanical stimulus requires a specific approved method and limits.
5. Last resort: lift components on the net one at a time, starting with the statistically guilty (MLCCs, tantalums, TVS diodes, the main IC).

5. Technique: junction signature comparison (no schematic needed)

With a known-good board: power off, diode mode, black probe on ground; touch every pin of a connector/IC on both boards and compare readings. Mismatch = pursue that net. This is one of the fastest ways to localize damage on complex digital boards and pairs perfectly with ATE failure tickets.

6. Self-check

1. Why must the meter never be in current mode across a power supply? Reveal answer
2. In-circuit, R12 (10kΩ marked) reads 14kΩ. Suspect or fine? Reveal answer
3. Diode test on D3 reads 0.002V both directions. Reveal answer
4. The 3.3V rail reads dead short. Name three localization techniques in order of preference. Reveal answer
5. What does 0.187V on a diode test across an unknown diode, measured out of circuit, suggest? Reveal answer

Next: 05 — Oscilloscope Mastery