To test a capacitor, set your multimeter with capacitance test capabilities to the Farad (F) setting, insert the black lead into the COM jack and the red lead into the V/Ω jack, discharge the capacitor completely, and place the probes directly across the terminals. A good reading is numerically within the manufacturer's stated tolerance—typically ±5% to ±20% of the rated microfarad (µF) value printed on the component casing. If the meter reads "OL" (Open Loop) or a value significantly below the lower tolerance bound, the capacitor has failed and must be replaced.
Meter Setup and Safety Categories (CAT Ratings)
Modern digital multimeters (DMMs) like the Fluke 117, Klein MM700, or Brymen BM235 integrate capacitance testing into the standard dial. Older or specialized meters might require moving the red lead to a dedicated "mA/µA/Cap" jack, but auto-ranging V/Ω jacks are the standard for 2026-era bench and field meters.
| Parameter | Required Setting / Connection |
|---|---|
| Dial Position | Capacitance (marked with F or the -( | |- symbol) |
| Black Lead Jack | COM (Common) |
| Red Lead Jack | V/Ω/Cap (Standard voltage/ohms jack on modern DMMs) |
| Range Selection | Auto-ranging (or start at the highest mF manual range and step down) |
Safety Categories (CAT Ratings) for Capacitor Testing
The safety category you need depends entirely on where the capacitor lives in the electrical system. Capacitors in HVAC systems, motor starters, and mains-tied power supplies can store lethal energy and are subject to high-energy transients from the grid.
- HVAC and Mains-Adjacent (240V/120V): You must use a meter rated for at least CAT III 600V or CAT IV 600V. This ensures the meter's internal blast shields and high-energy fuses can survive a transient spike if the capacitor arcs or fails catastrophically during testing.
- PCB and Board-Level (Under 50V): A CAT II rated meter is sufficient for testing decoupling caps, audio crossover networks, or low-voltage DC power supply filters.
For a detailed breakdown of transient overvoltage protections, refer to the Fluke guide on electrical safety categories based on IEC 61010-1 standards.
Never measure a capacitor in a live circuit, and never assume a disconnected capacitor is safe. A 45µF run capacitor charged to 370VAC stores enough joules to cause severe burns or fatal shock. Always discharge the capacitor first. Do not use a screwdriver to short the terminals (this damages the capacitor's internal metallization and can weld the screwdriver). Instead, use a dedicated discharge tool or a 20kΩ, 5-watt wirewound resistor on insulated alligator clips, held across the terminals for 5 to 10 seconds.
Expected Readings: Good vs. Bad Capacitors
What does a good reading look like numerically? A healthy capacitor will read within its printed tolerance band. Electrolytic capacitors (used for power filtering and motor starting) typically have a wide tolerance of ±20% or even -20%/+80%. Film and ceramic capacitors (used for timing, audio, and decoupling) usually have tighter tolerances of ±5% or ±10%.
Below is a data-dense reference table for the most common capacitors you will encounter in the field and on the bench.
| Component Type | Rated Value | Tolerance | Good Reading Range | Bad Reading (Replace) |
|---|---|---|---|---|
| HVAC Run Capacitor | 45 µF | ±6% | 42.3 – 47.7 µF | < 40 µF, > 50 µF, or OL |
| Motor Start Capacitor | 250 µF | ±20% | 200 – 300 µF | < 180 µF or Short (0.00 Ω) |
| Power Supply Filter (Electrolytic) | 1000 µF | ±20% | 800 – 1200 µF | < 750 µF or high ESR |
| Ceramic Decoupling | 0.1 µF (100 nF) | ±10% | 0.09 – 0.11 µF | OL (often below DMM resolution) |
| Audio Crossover Film Cap | 4.7 µF | ±5% | 4.46 – 4.93 µF | < 4.2 µF or physical bulging |
Note: "OL" (Over Limit) on a capacitance setting means the meter cannot detect a charge holding capability, indicating an open internal connection. If the meter reads near 0.00 µF or triggers the continuity beeper when set to Ohms, the capacitor has shorted internally.
Step-by-Step Probe Placement and Measurement
- De-energize and Isolate: Turn off the power, lock out the breaker, and verify the circuit is dead with a non-contact voltage tester or a proven CAT-rated meter. For HVAC systems, pull the disconnect block.
- Discharge the Component: Apply your 20kΩ discharge resistor across the terminals. For dual-run capacitors (common in HVAC, with HERM, FAN, and C terminals), discharge between C and HERM, then between C and FAN.
- Remove Parallel Paths (If Necessary): If you are testing a capacitor on a PCB or wired into a motor winding, remove at least one leg of the capacitor from the circuit. Leaving it connected allows parallel components (like inductors or other capacitors) to skew the reading.
- Connect the Probes:
- Polarized (Electrolytic): Place the red probe on the positive (+) anode and the black probe on the negative (-) cathode. While DMM capacitance test voltages are low (usually under 3V), observing polarity ensures the most accurate dielectric charging curve.
- Non-Polarized (Film/Ceramic): Probe placement does not matter. Place one probe on each terminal.
- Wait for Stabilization: The multimeter measures capacitance by applying a known DC test voltage and measuring the time it takes to charge. Large capacitors (e.g., 10,000 µF power supply filters) can take 5 to 15 seconds for the reading to settle. Wait until the numbers stop climbing.
Common Mistakes That Give Misleading Readings
Capacitance testing is highly susceptible to parasitic interference and operator error. If your readings seem erratic, check for these specific failure modes:
1. In-Circuit Testing Without Isolation
Measuring a capacitor while it is still soldered to a board or wired to a compressor winding is the most common cause of false "good" readings. Capacitance in parallel is additive ($C_{total} = C_1 + C_2$). If you are testing a 10µF cap, but it is in parallel with a 47µF cap, your meter will read ~57µF, leading you to believe the 10µF cap has swollen and failed when it is actually fine. Always lift one leg of the component.
2. Finger and Lead Parasitic Capacitance
When testing small values (under 1000 pF), the human body and the test leads act as capacitors. Standard silicone test leads introduce about 15 to 30 pF of parasitic capacitance. Touching the metal probe tips with your bare fingers adds another 50 to 100 pF.
The Fix: Use the REL (Relative) or NULL button on your multimeter. Short the probe tips together, press REL to zero out the lead capacitance, and then take your measurement. For sub-100pF measurements, a standard DMM is inadequate; you must use a dedicated LCR meter.
3. Ignoring Equivalent Series Resistance (ESR)
A standard multimeter with capacitance test functions only measures the ability of the component to store charge (µF). It does not measure Equivalent Series Resistance (ESR). An electrolytic capacitor can read a perfect 1000 µF on your DMM, but if the internal electrolyte has dried out, its ESR might have spiked from 0.05 Ω to 15 Ω. Under a high-frequency switching load (like in a PC power supply or LED driver), that high ESR will cause the capacitor to overheat and fail, even though your DMM says the capacitance is fine. For power supply diagnostics, pair your DMM capacitance check with a dedicated ESR meter.
4. Misinterpreting "Slow Charging" as a Bad Cap
Some technicians see the numbers slowly rolling up on the DMM display and assume the capacitor is "leaking" or bad. This is normal behavior for high-capacitance components. The meter is outputting a tiny constant current (often 10µA to 100µA) and timing the voltage ramp. A 50,000 µF capacitor will legitimately take half a minute to stabilize on a standard Fluke or Klein meter. Patience is a required tool for high-capacitance bench work.
For deeper theory on how dielectric absorption and leakage current affect these measurements, the All About Circuits capacitor testing chapter provides excellent bench-level context on the limitations of standard DMM test voltages.
