Why does a circuit read zero when something small is switched on?

Nothing is broken. The load is simply below what that clamp can see.

An SCT-013 clamp has two floors, and both scale with its rating — not with the size of the load you care about:

• a detectable floor, the current below which a reading is just noise;
• a readable floor, the current from which the reading is actually worth something (error within 10 %), at roughly 0.5 % of the clamp's rating.

Source: rbAmp module specification, verified in bench laboratory tests with a resistive load at 50 Hz.

What this looks like in practice

A 5 W lamp draws about 22 mA. On an SCT-013-005 that sits right at the readable floor — you'll see it, roughly. On an SCT-013-100, whose floor is 0.50 A, it's far underneath: the channel reads about zero, and correctly so.

The module subtracts its own channel noise, which is why an unloaded channel sits near zero instead of drifting — helpful, and also why a tiny load can look like nothing at all.

Power is computed separately from current, so near the floor you can see current at about zero while power shows a fraction of a watt. There's no separate energy threshold: energy is integrated from active power, so a load below the readable floor still contributes — just a small and imprecise amount.

What to do about it

Fit a smaller clamp on circuits with small loads. The specification says the same thing: if you need good measurement quality at small currents, choose a sensor with a smaller rating. Size each clamp to its circuit's breaker rather than buying one rating for the whole panel — a 100 A clamp on a lighting circuit measures nothing useful.

And tell the app which clamp is on the channel: the sensor model is part of the point's settings, and the hub scales its reading by it.

Next

• Which SCT-013 do you have?
• Sensors and modules
• Questions → Support