Measuring instruments and CAT categories — the right instrument for the job
Measuring instruments and CAT categories
Choosing the right measuring instrument isn't just a matter of convenience — using an instrument with the wrong CAT classification in the wrong place can cause it to explode in the event of an internal short circuit.
Instrument for the task
| Instrument | Measures | Typical use |
|---|---|---|
| Duspol (voltage tester with current draw) | Voltage present/absent, line-neutral, line-PE | Proving dead before working de-energised (step 3 of LOTO) — more reliable than a screwdriver-type phase tester because it takes a loaded measurement |
| Insulation tester (megger) | Insulation resistance between conductors and earth | Checking cable/winding insulation before commissioning or after repair |
| Low-resistance ohmmeter (micro-ohmmeter) | Very low resistance with high test current (≥ 200 mA) | Continuity of earth connections and protective conductors — an ordinary multimeter isn't accurate enough for this |
| Installation tester | Loop impedance Zs, RCD trip time, insulation resistance — often combined in a single instrument | Full NEN 1010 commissioning inspection and periodic NEN 3140 inspection |
| Phase-rotation indicator (draaiveldmeter) | Phase sequence (clockwise/counter-clockwise) | Checking rotation direction before connecting a three-phase motor, often at its own low test voltage |
| Clamp meter | Current without breaking the circuit | Quick check of load current per phase, for example when overload is suspected |
Insulation resistance — test voltage per circuit type
The test voltage and minimum acceptable insulation resistance for an insulation tester depend on the circuit's nominal voltage:
| Nominal circuit voltage | Test voltage (DC) | Minimum insulation resistance |
|---|---|---|
| SELV / PELV | 250 V | ≥ 0.5 MΩ |
| Up to and including 500 V (incl. 230/400 V) | 500 V | ≥ 1.0 MΩ |
| Above 500 V | 1000 V | ≥ 1.0 MΩ |
Always measure insulation resistance on a de-energised circuit that's disconnected from the mains — electronics (dimmers, frequency drives) must be disconnected beforehand, because the test voltage can damage them.
Note the difference between a new/modified installation and an existing one. The table above (≥ 1.0 MΩ at 500 V) applies to new or modified wiring, tested against NEN 1010. For periodic NEN 3140 inspection of an existing installation, the classic rule of thumb of 1000 Ω per volt of nominal voltage is often used instead — at 230 V that gives a lower limit of roughly 0.23 MΩ, with 0.5 MΩ commonly used in practice as a more comfortable margin. Heating elements (e.g. a boiler or a tunnel oven) often have a separate, lower normative minimum of 0.5 MΩ because of the nature of their insulation material. Always state explicitly in your report which standard and which threshold you applied, and why (new-build/modification versus periodic inspection).
Four practical pitfalls that can cause a measurement to fail or give a false reading:
- Disconnect all electronics: dimmers, electronic RCDs, LED drivers, frequency drives and sensors.
- Unscrew lamps and disconnect loads, not just switch them off.
- For an N-PE measurement, disconnect the neutral conductor from the busbar connection — otherwise you'll measure through other groups as well and get a value that's too low.
- Disconnect the surge protective device (SPD) before measuring — an SPD/varistor left in place shows a far too low insulation resistance because the test voltage partly leaks through it.
Continuous PE connection: calculating resistance
The low-resistance ohmmeter takes the measurement, but to assess the result it helps to first calculate an expected value using the standard formula:
R = ρ · L / A
where ρ (rho) is the resistivity (for copper: 0.0175 Ω·mm²/m), L is the length of the protective conductor in metres, and A is the cross-section in mm².
Example: a PE conductor 25 m long with a cross-section of 6 mm²:
R = 0.0175 × 25 / 6 ≈ 0.073 Ω
In practice, allow a margin of roughly one and a half times this value (contact resistances, terminal connections) before judging a measurement to be abnormal. Always measure with a test current of at least 200 mA — at lower currents, oxidation on a terminal can hide a poor connection.
What if the insulation resistance is not OK?
A failed insulation measurement is not a reason to proceed with energising "to see if it works anyway":
- Do not energise — leave the lockout (LOTO) in place.
- Report the deviation immediately to the work supervisor (WV).
- Localise the fault by splitting the circuit into sections (for example, per branch or connection point) and measuring each section separately, until the faulty section is isolated.
- Re-measure the repaired or replaced section.
- Only once the complete circuit again meets the required minimum insulation resistance may it be released for energising.
CAT categories — why this affects your own safety
The CAT classification (IEC 61010) of a measuring instrument says nothing about measurement accuracy, but everything about the energy the instrument can safely withstand in the event of an internal fault or transient overvoltage:
| Category | Application area |
|---|---|
| CAT I | Electronics, low-voltage circuits isolated from the mains |
| CAT II | Connection points on the "load" side: socket outlets, household appliances |
| CAT III | Fixed installation: distribution boards, sub-distribution panels, industrial equipment — the minimum level for work in a meter cupboard or distribution board |
| CAT IV | The origin of the installation: main connection, outside the meter, the connection to the grid operator's network |
Rule of thumb: the closer to the grid connection (and so the higher the possible short-circuit current), the higher the required CAT category. Using a CAT II instrument in a main distribution board is a real risk of injury from the instrument itself, quite apart from the measurement you're trying to take.
Practical rule of thumb
First choose the instrument based on the required measurement (voltage, insulation, continuity, loop impedance, phase sequence), then check that the CAT classification matches the location where you're measuring — not the other way around.