Derating of Current Sensors: What is it and why?

In an article about DC current measurement and bandwidth I wrote that "the derating of a current sensor and the sensor's bandwidth have nothing to do with each other".

The bandwidth of a current sensor defines how quickly the sensor can detect and respond to changes in the measured current. Like with an oscilloscope, the bandwidth of a current sensor also defines at which current the output signal of the sensor is -3dB lower than the input current signal.

The derating of a current sensor describes which input current a current sensor can handle at which frequency.

Let's look at the frequency derating curve of HIOKI's CT6877A zero-flux current sensor. The banner specs of this sensor state a maximum input current of 2000A and a bandwidth of 1MHz. The current is specified for continous use at a maximum temperature of 85°C (orange line). As you can see from the blue line, the maximum current is higher when used in an environment with a lower temperature.

Following both the blue and the orange line, the maximum input current of the sensor is flat until 300Hz, at which point it starts to drop. At 10kHz the maximum input current has dropped to 600A, and at 1MHz the maximum input current is just 10A.

What is the reason for the derating of maximum input currents?

You can get a very good indication why derating is necessary from the above derating curve: The current represented by the blue line, which shows the derating for a maximum ambient temperature of 65°C, is higher than the current represented by the orange line, which shows the derating for a maximum ambient temperature of 85°C. That's because heat is the main reason for derating.

The ambient temperature, of course, only plays a minor role. More relevant causes are heat generation on the circuit and heat generation caused by higher frequency eddy currents.

When does derating actually matter?

If you measure 50Hz AC currents with a couple of harmonics or even DC currents from a battery, then derating might not be a huge concern for you. The story is different if, for example, you need to measure currents of an inverter output.

The above image shows the active power of an inverter output over a wide frequency band. As you can see, the highest amount of power and therefore also currents are in a frequency range where current derating typically hasn't started yet. Especially for power analysis it is absolutely crucial to also precisely measure currents in the range of the switching frequency of an inverter.

Of course, these currents are much lower than the currents of the modulated wave and it's harmonics, but with the developments of new inverter technologies like SiC and GaN it is crucial to ensure that the derating of current sensors allows to measure these currents.

Switching frequencies are getting higher - what about the sensors?

The derating requirements for a current sensor which was launched ten years ago were not the same as for a current sensor which was launched recently.

The CT6844A is a current probe with a maximum input current of 500A. The predecessor of this sensor was called CT6844-05 with the same maximum input current of 500A. The maximum input current at 20kHz of the older CT6844-05 is 100A.

That is still an impressive number - especially given the fact that this is a current clamp with a split magnetic core and not a push-through sensor. The CT6844A, however, has a maximum input current at 20kHz of around 250A.

As a conclusion, derating describes the maximum input current of a current sensor over it's bandwidth. The derating curve of the sensor helps users to ensure that the higher frequency currents they measure can be handled by the current sensor.

The current sensors used as example in this article are based on zero-flux technology. Find out in this article where the name "zero flux" actually comes from.