EMC issues with open-loop Hall effect current sensor

Hi all,

we have a system that uses an open-loop hall effect sensor to measure the current (LEM DHAB s/133). During the EMC tests, precisely the IEC61000-4-3 and IEC61000-4-6, at frequencies of 120 MHz and 67 MHz we have bad read values of current (error >> 100% of real current). We're almost sure that the harness is ok because we've replaced the hall sensor with a divider and the read value was stable despite the EMC disturbances.

This is the output of the all effect current sensor with the coupled noise when I = 0A.

In this case (real current = 0 A and RF disturbance ON), we can see the RF noise but the DC value is correct (about 0 A)!

This is the output of the all effect current sensor with the coupled noise when I = 10 A.

In this case (real current = 10 A and RF disturbance ON), we can see the RF noise but the DC value of the output is NOT correct (about 58.8 A instead of 10 A)! This is the cause of the bad reading.

The bad read values seems due to a sort of ""polarization effect" of the hall effect sensor caused by the RF disturbance.
Anyone has some experiance on these sensors with EMC tests ?

Thank you all!

In case of doubt the effect is caused by RF overload of the signal output. Presumed you are sure that the connected signal processing electronic can't cause an error. Question is if the specified current injection limits in the datasheet are exceeded.

I would send a detailed specification of the test setup to the manufacturer.

I would send a detailed specification of the test setup to the manufacturer.

Click to expand...

I've already asked to the manufacturer, but thecnical support is not available by phone. I'm waiting their response by email.
The Hall sensor is directly connected to an opamp without any resistance towards Vcc or GND so i think the current injected to the hall sensor is small enougth since the op-amp has an high input impedance.
The dataheet of the hall effect sensor specifies 10 mA as absolute maximum ratings.