Guidance on using an automated RCD tester with Zaptec Go

When using an automated RCD tester, it is possible that the installer will receive unexpected results.

Specifically, a trip current of >6mA may be observed during ramp testing. This is due to the way some automated testers carry out the test. It does not necessarily mean there is a failure in the RCD unit.

Affected test equipment

Unexpected results have been observed with both the Megger 1741+ and Fluke 1654B. Other testers may exhibit similar behaviour.

Technical explanation

IEC 62955 specifies a maximum of 6mA for a slow-ramp (1mA/s) DC test. Testing by TÜV has confirmed that Zaptec Go complies with this requirement:


IEC 62955 specifies the trip time at 6mA must be <10s. Testing by TÜV has confirmed that Zaptec Go also complies with this requirement:


IEC 62955 also specifies how the tripping time should vary with the applied DC current. At 60mA, the trip time must be <0.3s and at 200 mA the trip time must be <0.1s 

Because the required tripping time is strongly dependent on the current, the measured tripping current therefore becomes highly dependent on the applied ramp rate. The automated testing process may not always take this into account.

Some automated testers apply a much faster current ramp (>10mA/s) than used in the slow-ramp IEC test (1mA/s). This leads to an inaccurate measurement of the slow-ramp tripping current. This behaviour has been replicated with a Fluke 1654B RCD Tester, which applies a ramp rate of 11.3 mA/s. As a result, it (incorrectly) reports a trigging level of 9-10mA, instead of <6mA. The correct triggering level would be measured if the applied ramp rate was <1mA/s, as specified in IEC 62955.

Automated testers are effectively carrying out two separate measurements (trip time and trip threshold) during a single, rapid test sequence at a higher ramp rate than specified by the test standard. This introduces significant uncertainty into the measurement results and may lead to incorrect measurement of the tripping current.



Was this article helpful?
1 out of 1 found this helpful