Resistance, continuity & electrical reference/grounding

In ordinary resistance mode, the digital multimeter provides its own test stimulus and measures the resulting electrical response. The exact current or voltage, compliance and range-switching behavior are instrument dependent.

That immediately explains why ordinary resistance and continuity tests are performed on a de-energized circuit: an external source would be superimposed on the meter's test conditions and can invalidate the result or violate the instrument's intended input condition.

“Power off” must also include stored energy. Equipment can retain energy after its supply is disconnected, so the appropriate discharge/verification procedure belongs to the equipment and instrument instructions.

With two 10 kΩ resistors still connected in parallel,

The meter has not “misread” either resistor. It has measured the equivalent resistance of the network between its probes.

A second example makes the diagnostic rule useful. A nominal 10 kΩ resistor that remains in parallel with 100 kΩ produces

Therefore an in-circuit value lower than the component marking can be completely consistent with a healthy component. To identify one resistor's isolated value, one terminal may need to be disconnected or the surrounding topology must be modeled.

For a simple two-wire low-resistance measurement, a first model is

where the lead and contact resistances are in the same current path as the unknown.

Suppose R_X=0.50 Ω and each lead contributes 0.20 Ω before contact effects:

Even before variable contact resistance, the reading is 80% above the desired 0.50 Ω.

A relative/zero function, when supported, can subtract a stable lead offset measured under the same conditions. It cannot magically remove contact changes, connector instability, temperature effects or an incorrect circuit model.

For low resistances, four-wire measurement separates the leads that force test current from the leads that sense voltage. The sense input is designed to draw very little current, so the voltage drops in the sense leads are correspondingly very small.

If a test current of 100 mA produces 50.0 mV directly at the device sense points:

The high-current force leads may still have significant resistance and voltage drop, but those drops are largely outside the sensed voltage pair. That is the essential Kelvin idea.

Four-wire measurement does not make every error disappear: contact geometry, thermoelectric effects, test current accuracy, heating and instrument specifications still matter. It specifically attacks lead/contact voltage-drop contamination of the voltage measurement.

Continuity mode typically applies a test stimulus and decides whether the measured condition is below an instrument-defined threshold, often with an audible indication. The threshold, response time, hysteresis and test stimulus are model dependent.

Therefore a beep means approximately “the instrument's continuity criterion is satisfied under this test condition.” It does not mean:

• the resistance is exactly zero;
• the path can safely carry arbitrary operating current;
• the path is the one you intended rather than a parallel route;
• the node is safely connected to protective earth.

Use resistance mode when the actual resistance matters; use continuity as a fast topology/checking aid within the meter's documented behavior.

The ground symbol used in a circuit diagram can denote a chosen circuit reference. Never infer a protective-earth connection from the word “ground” alone without the schematic and equipment documentation.

COM on a multimeter is the instrument's common measurement terminal. Its relationship to chassis and protective earth depends on the instrument architecture.

Battery-powered handheld meters are often designed with isolated inputs, but that does not mean “safe floating at any voltage.” Isolation and measurement-category ratings have limits. Bench instruments can have different relationships among input low, guard, chassis and protective earth.

Connecting several instruments can also create new conductive paths between nodes that were previously separate. Therefore the correct question is not “is COM ground?” but “what is COM connected to in this instrument, in this mode, according to the manual?”

The simple ohmmeter model assumes the network behaves like a resistance under the meter's test stimulus. Real circuits can violate that assumption:

• a nonlinear junction can respond differently when probe polarity or test range changes;
• external cables or instruments can provide parallel paths;
• high-resistance measurements can be altered by leakage through contamination, humidity or insulating surfaces;
• active circuitry can contain internal protection or bias paths even when the main supply is removed.

A resistance reading is therefore evidence about the network under the meter's test condition—not a universal identity tag for a component.

1. Two 10 kΩ resistors remain in parallel. What does the ohmmeter see?
2. A 10 kΩ resistor remains in parallel with 100 kΩ. Find the in-circuit equivalent.
3. A 0.50 Ω resistor is measured with two 0.20 Ω leads. Ignore contact resistance. What is the reading and relative error?
4. A four-wire test forces 100 mA and senses 50.0 mV across the device. Find resistance.
5. A continuity tester beeps on a 20 Ω path. List two conclusions you may not make from the beep alone.
6. Does declaring a node “0 V” connect it physically to protective earth? Explain.
7. Why is “DMM COM is earth” an unsafe general rule?
8. A resistance reading changes when the probes are reversed. What does that tell you about the simple linear-resistance assumption?

Chapter 05 will close MS-01 with accuracy, resolution, repeatability, calibration/traceability, uncertainty budgets and a reproducible DC measurement record.