Accuracy: approximately 2% error
Bandwidth: typically 80kHz
Range: ±25A* (linear over ±20A)
Noise: ~40mV peak-to-peak (equivalent to ~400mA)
Power supply: 9V battery, approximately 20 hours life
Resistance: ~2mΩ plus cable resistance
Isolation: 2.1kV RMS (suitable for use up to 250V AC)
Withstand current: 100A for 100ms
Other features: power indicator, auto-off to preserve battery life
* With alternative shunt IC, range increases to ±38A (linear over ±33A) with 66mV/A output
In the SILICON CHIP laboratory, we often need to hook our digital storage oscilloscope (DSO) up to mains-powered equipment to examine the current waveforms. The two most common ways to do this are with a shunt resistor and differential probe or a clamp meter. But both approaches have drawbacks.
A shunt resistor connected in series with one of the mains conductors (eg, Neutral) provides the best bandwidth and resolution but you need a differential probe (which can be expensive), even if you are measuring on the Neutral leg since Neutral is usually a few volts above or below Earth potential.
The resistor also limits how much current you can measure depending on its value. For example, a 0.1Ω 10W resistor limits you to measuring around 7A RMS (after de-rating by 50%). This option can also be quite unsafe as the wiring between the shunt and probe is connected directly to mains.
A clamp meter is safer since it doesn’t require any exposed mains wiring. But they tend to have a fairly low output voltage, eg, 1mV/A. This gives you lousy resolution and noise performance with scopes which usually have a maximum sensitivity of 5mV/div. Clamp meters also typically have quite limited bandwidth (eg, 10kHz) which is no good for loads with fast-changing current waveforms such as switchmode supplies.
Also, you need to separate out the mains conductors to use a clamp meter since if you just clamp it over the cable, the Active and Neutral currents are of identical magnitude and opposite in direction so the magnetic fields effectively cancel. So you need some kind of special cable or adaptor to measure mains current with a clamp meter.
With our adaptor, you get much higher bandwidth and resolution than a clamp meter (80kHz, 100mV/A) with better safety than a shunt resistor, no need for a differential probe and at a fairly low cost.
We use an Allegro ACS712 IC, which like a clamp meter operates on the Hall Effect principle but the whole shebang is within a single chip. One side of the IC contains a 1.2mΩ shunt which can handle a continuous current of at least 30A and pulses up to 100A for 100ms. On the other side is a fully isolated Hall Effect sensor and amplification circuitry.
There is no electrical connection between the two halves; sensing is purely based on the magnetic field generated by current passing through the shunt. The chip has an isolation rating of 1500VAC between the two halves so the output can safely be hooked up to a scope or other device even if you are measuring mains current at up to 250VAC.
There are three versions of this IC, designed for sensing currents up to ±5A, ±20A and ±30A. They are otherwise identical. For our prototype, we used the 20A version since its output is 100mV/A and this makes it easy to set up our scope to read out directly in amps (by telling it we have a 10:1 current probe). We run it from a 5V supply, giving readings of up to ±25A although linearity is a little degraded at the extremes.
The 30A version has an output of 66mV/A and can read up to ±38A. You can use this one if you prefer but then you may need a calculator to interpret the readings.
Power comes from a 9V battery because this is much more convenient than a plugpack when setting up a test. We fitted ours with a mains plug and socket for measuring the current drawn by mains devices however it could also have been fitted with DC connectors if that’s what we wanted to measure. The output is a BNC socket, making it easy to hook up to a scope. For connection to a DMM, we use a BNC plug to banana socket adaptor.
So that you can’t accidentally leave the unit on and drain the battery (easy to do!), we incorporated an automatic time-out which switches the unit off after about 15 minutes. If you want to use it for a longer period, you just have to remember to periodically press the power button to keep it on.