How to Test an RCD: Test Button and Instrument Testing

How do you test an RCD? Correct testing has two distinct steps: pressing the built-in test button, which proves the mechanical trip linkage and test-resistor circuit only, and using a calibrated RCD tester to inject a controlled residual current and measure the actual trip time in milliseconds against IEC 61008-1 and IEC 61009-1 limits. Skip the second step and a device that "passes" the button test can still fail to trip within the time that matters during a real earth fault. This article covers what the test button checks, how ramp and timed instrument tests work, the trip-time limits at IΔn and 5×IΔn, commissioning versus periodic test intervals, and the mistakes that produce false readings on site.

What the Test Button Actually Checks

Pressing the test button routes a small current through a resistor wired across one pole, deliberately unbalancing the toroidal core by roughly the rated IΔn. If the sense winding, trip relay, and mechanical latch are all functional, the device opens. That is the entire scope of the check.

The test button does not verify the actual trip threshold. A worn or drifted device might still trip on the button test at a much higher unbalance than its nameplate IΔn, and the button gives no indication of how long the trip takes. It also does not exercise the earth path or the real fault current pattern a hand-to-earth shock would produce.

What we see in the field: electricians sometimes treat a passing button test as proof the RCD is "fine" and skip instrument testing entirely, especially on older boards. The button confirms the device isn't dead. It says nothing about whether IΔn has drifted 40% high, which happens on units that have sat in a damp panel for a decade.

Instrument Testing: Measuring Actual Trip Time

A dedicated RCD tester (also called an earth-leakage tester) injects a precisely controlled residual current at a chosen phase angle and multiple, then times the trip electronically. This is the only method that produces a number you can compare against the standard.

Ramp Test at IΔn

The ramp test increases the injected current slowly from zero until the device trips, recording the current at which it operates. This confirms the actual pickup point sits close to the rated IΔn rather than assuming the nameplate value is still accurate. A 30 mA RCD that ramp-trips at 45 mA has drifted out of tolerance and should be replaced, even if it passes a fixed-current timed test.

Timed Test at 5×IΔn

The timed test injects a fixed multiple of IΔn — typically ×1 and ×5 — and measures trip time directly. This is the number that matters for shock protection: a 30 mA device must clear a fault fast enough that the current-duration product stays below the let-go and fibrillation thresholds used to derive the IEC limits.

S-type (selective) devices have a different table: a minimum non-trip time as well as a maximum trip time, since they are meant to hold long enough for a downstream device to clear the fault first. Testing an S-type RCD against the general-type limits above will flag a false fail — check the datasheet for the correct time band before condemning a selective device.

Test Button Result vs Instrument Test Result

Commissioning Tests for New Installations

On a new board, every RCD and RCBO circuit should get a full instrument test before handover: ramp test to confirm actual IΔn, timed test at ×1 and ×5, and — where the installation is TT — a loop impedance test to confirm the earth electrode resistance keeps touch voltage within the RA × IΔn limit for the fitted sensitivity. See our RCD earthing system guide for how that interacts with TT, TN, and IT arrangements.

Record the results per circuit: device location, IΔn, type (AC/A/F/B), measured trip time at IΔn and 5×IΔn, and date. This record is what a commissioning certificate actually needs — a note that "the test button was pressed" is not equivalent documentation and will not satisfy most inspection regimes.

For circuits carrying variable frequency drives, EV chargers, or PV inverters, confirm the fitted type matches the load before testing. A Type AC tester result on a circuit that needs Type B tells you nothing useful if the fitted device is the wrong type in the first place — check selection against our Type A vs Type B RCD guide.

Periodic Testing Intervals

Monthly test-button checks are standard guidance for domestic and light-commercial installations — the user presses the button, confirms it trips, and resets. This is a user-level check, not a substitute for periodic instrument testing.

Periodic instrument re-testing intervals depend on the installation category and local code: many commercial and industrial maintenance schedules call for full timed and ramp testing at each fixed electrical installation condition report cycle, commonly every 3-5 years, or annually in higher-risk environments such as construction sites and medical locations. Cumulative cable leakage on long final circuits with many electronic loads can push a healthy RCD closer to nuisance-trip territory over time, so environments with a lot of switched-mode equipment benefit from shorter re-test intervals even if the code minimum is longer.

Common Testing Mistakes and False Results

Testing the wrong pole or phase reference gives a result that looks like a fail when the device is fine — on 4P RCCBs, confirm the tester's line and neutral leads land on the same pole pair the toroid monitors.

Ambient leakage from other circuits sharing the same RCD-protected group can add to the injected test current and produce an apparently early trip. This is common on boards where one 30 mA RCD covers the whole panel rather than being split per circuit — some electricians accept this on small domestic boards, but on industrial panels with multiple VFDs or long cable runs it produces trips that look like an instrument fault when the real cause is cumulative background leakage.

Testing through an upstream RCD without disabling it produces nuisance trips of the wrong device and can be mistaken for a faulty downstream RCBO. On boards with selectivity between an S-type main and instantaneous final RCBOs, test the final device first, in isolation, before testing the upstream S-type at its own current level.

Testers with a stale calibration certificate produce results you cannot defend on an inspection report. Calibrate per the tester manufacturer's interval, typically annually, and record the certificate number alongside the test results.

Conclusion

The test button and the instrument test answer different questions. The button confirms the device can still trip at all. The instrument test — ramp plus timed, at IΔn and 5×IΔn — confirms it trips at the right current within the IEC 61008/61009 time limit, which is the actual safety requirement. Commission every circuit with a full instrument test and record the numbers, keep the monthly button check as a user-level habit between periodic tests, and rule out wrong-pole connection or shared leakage before condemning a device that fails on site. For selecting the right sensitivity and type before you ever reach the test bench, see our RCD selection checklist, and for units that need replacing after a failed test, browse Stoklink's residual current devices and RCBOs. For the broader terminology and device family, start with the RCD protection guide or the RCCB vs RCBO differences comparison. If trips keep recurring after a passing test, work through our RCD tripping causes and fixes guide next.