← Back to courseVisual Reference

Test Equipment Visual Guide — DMM · Oscilloscope · ICT

Annotated reference figures for the aerospace CCA repair bench. Every callout number in a diagram matches the numbered legend below it.

1. Digital Multimeter — Annotated Front Panel

Fluke-87-style handheld DMM. Display on top, rotary function dial in the center, four input jacks at the bottom. Learn the dial cold — picking the wrong function or jack is the most common new-tech mistake.

AUTO V ⎓ 0.583 HOLD RANGE REL Δ Hz % OFF V~ V⎓ mV⎓ Ω µA·mA A A mA/µA COM V·Ω FUSED 10A FUSED 440mA CAT III 1000V · CAT IV 600V TRUE-RMS MULTIMETER · 10 MΩ INPUT 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
  1. LCD display. Reading, units, AUTO-range flag, bar graph. If the units annunciator doesn't match what you think you set, stop and re-check the dial and jacks.
  2. HOLD / RANGE / REL Δ buttons. HOLD freezes the reading so you can look away from the probes. RANGE forces manual ranging. REL Δ zeroes out the displayed value — press it with the leads shorted before any low-ohms check to subtract lead resistance.
  3. V⎓ — DC volts. The repair-bench workhorse: power rails, reference voltages, voltage drops across components. Voltage is always measured ACROSS a component (in parallel), with the circuit powered. Typical input impedance is 10 MΩ, so the meter barely loads the node it touches.
  4. V~ — AC volts (true-RMS). AC line and transformer secondaries; also a rough check for AC ripple riding on a rail (a scope shows ripple far better — see Section 6d).
  5. mV⎓ — DC millivolts. Small drops: current-sense resistors, fuse drop, connector and solder-joint resistance under load.
  6. Ω — resistance. Only on UNPOWERED, discharged circuits — the meter sources its own small current and any external voltage corrupts the reading (or damages the meter). Discharge bulk caps first.
  7. Continuity beeper / diode test (speaker + diode symbols). Beeper: fast unpowered go/no-go on traces, vias, fuses, and joints. Diode test sources ~1 mA and displays forward voltage: silicon junction 0.5–0.7 V, Schottky 0.15–0.45 V, LED 1.6–3.3 V. Low reading in both directions = shorted junction; open (OL) in both = blown junction.
  8. Capacitance. Sanity-check on a removed cap (value, not ESR — a cap can read full value and still be bad; use an ESR meter or scope ripple for that).
  9. µA / mA — low current ranges. Sleep currents, LED currents, 4–20 mA loops. Uses the 440 mA-fused jack.
  10. A — high current (10 A) range. Motor, heater, supply input current. Uses the 10 A-fused jack. Current is measured THROUGH the circuit — in series.
  11. COM jack (black). Common/return for every measurement. Black lead lives here, always.
  12. VΩ jack (red). Volts, ohms, continuity, diode, capacitance, frequency.
  13. mA/µA jack (red, 440 mA fused). Low-current input. In a current jack the meter is a near-short — if you then probe across a supply you blow the fuse at best. Move the lead back to VΩ when you're done.
  14. A jack (red, 10 A fused). High-current input. Same near-short rule: never place a current-jacked meter across a voltage source.
  15. CAT safety rating. CAT II = receptacle-connected equipment, CAT III = distribution panels/fixed equipment, CAT IV = utility service entrance. Meter and leads must meet or exceed the environment you probe. Bench CCA work is low-energy, but the same meter goes to the aircraft — respect the rating.

2. DMM Measurement Setups — Where the Probes Go

Four canonical setups on a simple battery + resistor circuit. The geometry is the rule: voltage across, current through, ohms isolated and unpowered.

(a) Voltage — ACROSS

ACROSS — parallel + 9 V R1 1 kΩ 9.00 V DMM · V⎓ Circuit powered · meter in parallel with R1
Voltage is measured ACROSS a component, circuit powered. The 10 MΩ input barely disturbs the node, so you can probe almost anywhere safely.

(b) Current — THROUGH (in series)

THROUGH — series break opened + 9 V R1 8.20 mA DMM · mA (fused jack) all circuit current flows through the meter
Current is measured THROUGH: open the circuit and bridge the break with the meter in series. In a current jack the meter is a near-short — never place it across a supply; that's an instant blown fuse (or worse).

(c) Resistance — isolated, power OFF

POWER OFF switch open 9 V one leg lifted 1.002 kΩ DMM · Ω No power · caps discharged · part isolated from the circuit
Resistance and continuity only on UNPOWERED, discharged circuits. For a value you can trust, isolate the component — lift one leg so nothing else is in parallel.

(d) Continuity — beeping through a trace

POWER OFF copper trace under test 0.2 Ω DMM · continuity BEEP! Beeps below ~30 Ω — trace, via, fuse, or joint is intact
Continuity beeper: the fastest unpowered check that a trace, via, fuse, or solder joint is intact end-to-end. Eyes on the probes, ears on the meter.
In-circuit resistance caveat. On a populated board, parallel paths and semiconductor junctions make in-circuit resistance readings read LOW — the meter sees R1 in parallel with everything else connected to those two nets. A low in-circuit reading is a clue, not a verdict: lift a leg (or remove the part) and re-measure before you condemn it.

3. Oscilloscope — Annotated Front Panel

A 2-channel bench scope. Everything on the front panel belongs to one of three questions: how tall (VERTICAL), how wide (HORIZONTAL), and when to draw (TRIGGER).

BENCHSCOPE DS-2102 · 100 MHz · 2 CH · 1 GSa/s RUN CH1 1.00V CH2 off M 250µs Trig CH1 ↗ 1.52V CH1 CH2 EXT TRIG PROBE COMP ~3V 1kHz USB RUN / STOP SINGLE VERTICAL POSITION CH1 CH2 COUPLING: AC · DC · GND VOLTS / DIV 5 mV – 5 V per div HORIZONTAL POSITION TIME / DIV 2 ns – 50 s per div TRIGGER LEVEL MENU SRC: CH1·CH2·EXT EDGE: ↗ rise / ↘ fall MODE: AUTO · NORM · SINGLE 1 7 2 6 8 3 4 5
  1. Display. The waveform plus on-screen readouts: V/div, time/div, trigger settings, automatic measurements. Anatomy in Section 4.
  2. CH1 / CH2 BNC inputs (+ EXT TRIG). Probes connect here; channel color rings match the trace colors on screen. EXT TRIG lets an outside signal (e.g. a test-set sync pulse) tell the scope when to draw without using a display channel.
  3. VERTICAL group — "how tall." VOLTS/DIV scales the trace; POSITION slides it up/down; CH buttons turn channels on/off. Coupling: DC shows everything, AC blocks the DC so you can zoom in on ripple riding on a rail, GND shows where 0 V sits. Bench habit: set V/div so the signal fills 3–6 divisions.
  4. HORIZONTAL group — "how wide." TIME/DIV sets seconds per division (zoom in time); POSITION slides the capture left/right around the trigger point. Slow for power rails, fast for clock edges.
  5. TRIGGER group — "when to draw." A stable picture means the scope starts each sweep at the same event. LEVEL sets the voltage threshold; SOURCE picks which input is watched; EDGE picks rising or falling. MODE: Auto free-runs so you always see something (good for browsing), Normal waits for a real event (good for intermittent signals), Single captures one event and stops (good for glitches and power-up transients).
  6. Probe-compensation terminal. Built-in ~3 Vpp 1 kHz square wave. Hook every probe here and trim it before trusting any measurement (Section 5).
  7. RUN/STOP and SINGLE. Freeze a live trace to study it; SINGLE arms a one-shot capture — your tool for catching the glitch that only happens once at power-on.
  8. USB port. Save screenshots and waveform data — required evidence for repair records and failure-analysis writeups.

4. Scope Screen Anatomy — Reading the Display

Everything you need to turn a picture into numbers: count divisions, multiply by the per-division settings shown at the bottom of the screen.

T T 1⏚ Vpp(1) = 3.00 V Freq(1) = 10.00 kHz CH1 1.00 V M 25.0 µs Trig CH1 ↗ 1.50 V 1 2 3 4 5 6 7 8
  1. Graticule. 10 horizontal × 8 vertical divisions; the center axes carry minor ticks at 0.2-division steps. All scope math is "count divisions, multiply by the per-div setting."
  2. V/div readout. Vertical scale for CH1 — here 1.00 V per division.
  3. Time/div readout. Horizontal scale — here 25.0 µs per division.
  4. Trigger LEVEL arrow (right edge). The voltage threshold the trigger fires at — currently 1.50 V, mid-swing of the signal, where it belongs.
  5. Trigger POINT marker (top edge). The instant in time the trigger fired; everything left of it is pre-trigger history — gold for seeing what happened before a fault event.
  6. Ground reference arrow (left edge). Where 0 V sits for CH1. Always know where ground is before reading amplitudes.
  7. Automatic measurement readouts. Vpp, frequency, etc. Trust them only when the trace is clean and fills a good part of the screen.
  8. The trace — worked example: the wave is 3 divisions tall at 1 V/div → 3 Vpp. One period spans 4 divisions at 25 µs/div → 100 µs → 1/100 µs = 10 kHz. Matches the automatic readouts — that cross-check should become reflex.

5. Probes — The Most-Ignored Part of the Measurement

A 10:1 passive probe. The probe is part of the circuit the moment it touches the board — know what it adds and what it can short.

10:1 COMP to scope CH input short ground spring on the probe barrel correct — flat tops under — rounded over — peaked Probe compensation on the cal terminal — trim until the square wave is actually square. 1 2 3 4 5 6 7
  1. Probe tip. Sharp point for pads and vias. The tip plus its ground return is a loop — keep the loop small.
  2. Spring-hook attachment. Slides over the tip to grab a lead or test point hands-free. Remove it for fine-pitch work — hooks slip and short adjacent pins.
  3. Ground lead with alligator clip. Convenient, but the long loop adds inductance — it's the usual cause of the ringing in Section 6(b).
  4. Compensation trimmer. Small adjustable capacitor. ALWAYS compensate a probe on the scope's cal terminal when you pick it up or move it to another channel — see the three inset waveforms: flat tops = correct, rounded = under-compensated, peaked = over-compensated. An uncompensated probe lies about every fast signal.
  5. BNC connector. Quarter-turn bayonet to the scope input.
  6. 10:1 attenuation. Divides the signal by 10. Why: it loads the circuit far less (~10 MΩ and a few pF instead of 1 MΩ and ~100 pF) and extends the voltage range. The scope channel must be set to 10X (most sense it automatically) — otherwise every reading is 10× off.
  7. Short ground spring. For fast edges, ditch the alligator lead and use the spring on the barrel to a ground point right next to the signal — the small loop kills most ringing.
CRITICAL SAFETY — the ground clip is earth ground. The scope's ground clip is tied straight to mains earth through the power cord. Clip it to a live or elevated node and you have shorted that node to earth — through your probe, the scope, and possibly through you. On floating or off-ground measurements use a differential probe. Never "fix" this with a ground-lift cheater plug.

6. Waveform Gallery — What Good and Bad Look Like

Pattern recognition is most of scope work. Eight traces you will meet on the repair bench.

(a) Clean 3.3 V clock

3.3 V 0 V

Healthy digital clock: crisp edges, flat tops, full 0-to-3.3 V swing, even period. This is your baseline.

(b) Severe ringing / overshoot

overshoot + ring

Same clock with ringing on every edge — classic long ground lead on the probe, or a real impedance/termination problem. Re-probe with the ground spring before blaming the board.

(c) Runt pulses / dips

runt runt

Pulses that never reach a valid logic high: bus contention (two drivers fighting) or a weak/dying driver. Receivers downstream see garbage intermittently.

(d) Rail ripple (AC-coupled)

AC coupled · 50 mV/div ≈140 mVpp sawtooth on the 5 V rail

DC rail viewed AC-coupled at high vertical sensitivity. Sawtooth ripple this large points at a failing bulk capacitor (lost capacitance / high ESR) after the regulator or rectifier.

(e) UART frame (8N1)

idle (high) start D0D1D2 D3D4D5 D6D7 stop byte 0x4D, LSB first

UART frame: line idles high; start bit pulls low; 8 data bits LSB-first; stop bit returns high. If you can read this by eye you can debug most serial links.

(f) I²C transaction (SCL + SDA)

SCL SDA START ACK START: SDA falls while SCL high · 9th clock: ACK

I²C: open-drain bus — both lines need pull-ups and idle high. START = SDA falls while SCL is high; on the 9th clock the addressed slave ACKs by holding SDA low. No ACK? Check pull-ups, address, power to the slave.

(g) Flatline where a clock should be

expected clock 0 V — dead

Flatline at 0 V where the datasheet says a clock lives: dead oscillator. Check its power pin, enable pin, and the crystal/load caps before condemning the IC it feeds.

(h) Slow, rounded rising edges

RC charge curve

Sharp falls but lazy exponential rises: the line is fighting an RC. Suspect a missing/wrong pull-up, excess capacitance on the net, or a partial short loading it. Timing margins quietly evaporate.

7. Teradyne In-Circuit Test (ICT) — Bed of Nails

ICT presses the assembled board down onto hundreds of spring pins and measures every component, one at a time, in seconds. You'll spend a lot of time turning its failure tickets into actual repairs.

press head / hold-down gate U12 R47 C8 vacuum probe plate — custom-drilled per board design fixture receiver — relays · measurement bus · digital drivers/sensors ↓ to Teradyne test head ↓ 1 2 3 4 5 6 7 8
  1. UUT (unit under test). The assembled CCA, components up, pressed down onto the pin field.
  2. Test pads / vias. Dedicated landing targets on the board's underside — one per net the test engineer needs access to.
  3. Spring-loaded pogo pins. Each pin has an internal spring (shown in the cutaway) and a crown tip that bites through flux residue. Bent, dirty, or worn pins cause false opens — remember that when a ticket looks impossible.
  4. Probe plate (fixture). Pins are press-fit into a plate drilled specifically for this board design — the "bed of nails."
  5. Press head / vacuum. A gate press (or vacuum drawing the sealed cavity down) forces the board onto hundreds of pins with even pressure.
  6. Vacuum seal / fixture wall. Gasket seals the cavity so vacuum can pull the UUT flat onto the pin field.
  7. Fixture wiring. Every pin is wired to a specific test-system channel — the fixture is a giant, board-specific adapter.
  8. Receiver → Teradyne test head. Relay matrix routes any pin to the measurement bus (analog tests) or to digital driver/sensor channels (vector tests).

Guarding — how ICT measures one resistor inside a connected circuit

ICT measurement channel source → R1 → measure (virtual ground) S — source pin M — measure pin R1 — the part being measured R2 R3 A B C GUARD pin — 0 V measured current — only R1's current reaches the measure pin stray current through R2 is swallowed by the guard ≈0 V across R3 path → ≈0 mA — can't corrupt R1
Guarding: the tester measures R1 between nodes A and B while a third pin drives node C to 0 V. Current sneaking through R2 is absorbed by the guard, and with both ends of the R3 path near 0 V essentially no current flows through it into the measure node — so the parallel path R2 + R3 can't fake R1's value. This is why ICT can measure in-circuit where your DMM reads low (Section 2 caveat).

What ICT tests, and how to read a failure ticket

What the system checks, component by component: opens, shorts, wrong value, missing part, backwards part (diodes, polarized caps, ICs) — each measured individually using guarding. Beyond analog measurements it runs digital vector tests (drive a pattern into an IC's inputs, sense its outputs) and boundary scan (IEEE 1149.1 / JTAG), which shifts test data through the scan chain inside the chips themselves — the main way to verify solder connections under BGAs and other packages no pogo pin can reach.
The failure ticket gives you: the device designator (e.g. R47), the test name, the measured value vs. nominal ± tolerance (e.g. "measured 142 Ω, expected 1.00 kΩ ±2%"), and the pin/net numbers involved. That is your starting map to the exact spot on the board.
Repair-tech mindset: the ticket tells you WHERE to look, not always WHAT is wrong. A "failed R47" may really be a cracked solder joint, a bent or dirty pogo pin, a board not seated flat in the fixture, or a neighboring part loading the net. Verify with the DMM at the component before swapping parts — pulling a good part wastes time and risks pad damage on an expensive aerospace assembly.