Re: Isolated AC line current sensors

On Tue, 22 Jul 2025 at 12:05, Harold Hallikainen <[email protected]>
wrote:

> I've been asked about sensing current on the AC line. Current transformers
> immediately come to mind. Also, I see there are Hall effect based sensors
> ...


*** WARNING ***

When using isolated current sensors you need to be sure to be certain (to
be sure) that the sensor can NEVER be destroyed by an input side fault (or
any other condition) in a manner that breaches isolation OR that the worst
case outcome is tolerable.
Ask me how I know :-) .

For example - if an isolated Hall effect sensor is rated at 20A full scale
and is measuring current from a source able to provide FAR more current
under fault conditions then, if a fault occurs, the sensor may be destroyed
and isolation may be broken.
In such a case either be sure that current adequate to melt the sensor
never occurs OR ensure that the output side either will tolerate a worst
case isolation breakdown, or that it will fail safe sacrificially.

The former may be achieved by fuse, breaker or current limiter, with an
operating time short enough for the Hall etc sensor not to know that
something bad almost happened.

It may be easier to achieve the latter as once you have an output signal
from the sensor it is usually easy enough  to couple it with a second stage
of isolation - digital is very easy and analog not too much harder. The
second isolation barrier need usually only tolerate voltage, which is
usually less demanding than currents high enough to melt a 20A Hall Sensor.

In my case I had a 24V LiFePO4 40Ah? battery good for untold short circuit
amps. Something now unrecollected happened and the Allegro 20A Hall sensor
turned into  a charred mess with Ohmic connection from 24 VDC into a data
collection system. No great 'collateral damage' resulted, but much worse
could have happened. In a system with higher voltages Murphy will usually
not be so kind.


          Russell McMahon