In-circuit resistance measurement:
pitfalls and tips
Why phantom resistance ruins
your readings
In-circuit resistance measurement is unreliable because a powered board contains active semiconductors that conduct current in unexpected directions. Your multimeter's test voltage (typically 2.0V to 5.0V in resistance mode) can forward-bias diodes, partially activate transistors, and create false parallel paths through capacitive loads.
Measuring a resistor that sits between a power rail and GND while the board is powered will show artificially low resistance. The multimeter's DC source superimposes on the existing voltage gradient, and parasitic leakage paths through nearby components artificially reduce the measured value.
Example: A 10 Ω sense resistor measures 6.8 Ω in-circuit because the parallel impedance of surrounding biased circuits contributes a phantom shunt. Measure the same resistor out-of-circuit: 10.0 Ω exact.
Diodes and forward biasing during
resistance mode
Multimeters apply a low test current when measuring resistance—typically 0.5 mA to 1 mA on the 200 Ω scale. Any diode in the measurement path will conduct, presenting a forward-drop resistance (typically 0.3 Ω to 0.7 Ω for silicon diodes, 0.1 Ω to 0.3 Ω for Schottky).
Protective diodes across loads, schottky diodes on power rails, and ESD clamps all become conductive pathways. If you measure across a protection circuit containing a BAT60A Schottky, you measure the diode's forward resistance, not the circuit integrity. Similarly, TVS diodes like SMAJ36CA (used on USB rails) will conduct under test voltage and appear as shorts.
Capacitive loading and DC measurement
instability
Large bulk capacitors create two problems: they store residual charge that taints readings, and their impedance changes during the measurement window. When your multimeter applies test voltage, capacitors charge, creating transient current flow. The ohmmeter's display may drift for 1–2 seconds while the cap settles.
On power delivery circuits with 100 µF to 470 µF tantalum or ceramic bulk caps, measure resistance between two nodes that are directly across a capacitor: the capacitive path can register false low resistance (typically 0.5 Ω to 3.0 Ω) until the cap fully discharges through your meter's input impedance.
Best practice: After powering off, press both leads together and short them to GND for 3–5 seconds to bleed residual charge. Then wait for the board to sit quiescent for at least 30 seconds before probing. High-impedance measurements (above 100 kΩ) need even longer settling time.
| Bulk Cap Value | Settling Time Typical | Safe Probe Delay |
|---|---|---|
| 47–100 µF | 2–4 sec | 20 sec minimum |
| 220–470 µF | 5–10 sec | 30 sec minimum |
| 1000+ µF | 15–30 sec | 60 sec minimum |
Parasitic conduction paths and
multi-component shunting
Dense boards with integrated passive networks, ferrite beads, and low-value coupling resistors create unexpected parallel impedance. When you measure between test point TP_VDDIO and TP_GND, you are not measuring a single path—you are measuring the combined impedance of dozens of components in parallel.
A ferrite bead on the input to an LDO (e.g., BLM21PG121SN1) presents 2–5 Ω impedance at DC. Bypass capacitors appear as near-zero impedance. A 100 nF ceramic cap in parallel with a 10 Ω resistor will measure closer to 0.8 Ω due to capacitive effects and the diode stack in ISL6259 or TPS51125 buck controllers.
Measurement technique: Isolate the target component or trace. If possible, lift one end of the component using a fine soldering iron (apply 350 °C tip, dwell 2–3 seconds) to break the parallel path, then measure. For high-impedance networks, use the highest ohms range on your meter and trust only the final stable reading after 3–5 seconds of settlement.
Practical measurement workflow and
documentation
Establish a repeatable in-circuit resistance protocol:
- Power down and wait. Kill AC, wait 60 seconds. Verify no LED indicators glow.
- Discharge capacitors. Use a resistive load (a 1 kΩ resistor probe from power to GND) or short main rails to GND with tweezers. Confirm multimeter shows 0V DC on the rail.
- Select the correct range. Use the 2 kΩ or 20 kΩ range for component-level work; use 200 Ω or lower only for direct trace-to-trace shorts.
- Probe with light pressure. Avoid bridging adjacent pads. A spring-loaded probe tip or tweezers with fine points prevents accidental shorts.
- Record the stable reading. Wait 3–5 seconds for the display to stabilize. If it drifts more than ±10%, note it and re-measure after longer settling.
- Cross-validate. Measure the same rail at a different test point. Resistance should be within 5–15% if the trace is continuous.
Documentation example:
TP_VCORE to GND: 18.3 Ω (typical for a power delivery stage with bulk caps). If you measure 0.2 Ω, suspect a short or failed LDO output stage (check the LP8550 or equivalent controller for stuck switching).
Never rely on a single resistance reading to declare a board "good" or "bad." Cross-reference against schematics, compare with known-good boards if available, and always perform follow-up voltage and signal checks on the affected rail.