Isolated current measurement for rectified AC?

[e100]

A google search didn't turn up much.
Can I just use an over-rated current transformer to avoid saturation issues, or is that never going to work?

Off the shelf Hall effect sensors only go up to about 30A. I don't think using a shunt to increase the range is viable because of tempco of the copper wires going to (and inside) the sensor. Using multiple Hall effect sensors in parallel to divide the current will only work if the readings can be synchronised exactly which will require multiple  ADCs.
Is there a better solution?

Mike

[bktemp]

It depends on the current transformer: There are accurate current transformers using a core having a high permeability for lowest phase error and there are saturation resistant current transformers using a lower permeability core.
If there is a dc offset, you won't measure the true rms current, because the dc value will be lost. Only the ac component will be available at the output.

[Jeroen3]

What are you looking to measure? You talk about increasing range, do you want to measure 100A with mA precision?

[e100]

What are you looking to measure? You talk about increasing range, do you want to measure 100A with mA precision?

The load is a salt water chlorinator which is resistive and non-linear.
The "average" current (20-30A) is varied by phase angle triggering of a SCR rectifier via a manually set potentiometer.
Under fault conditions (too much salt added too quickly) the current could jump by a factor of 3.

I would like to monitor the current phase angle and instantaneous current to a resolution of 1A with a view to adding some kind of automatic shutdown mechanism.

[e100]

Thinking of multiple Hall sensors, presumably if they are reasonably matched I could sum the analog outputs into one ADC. Would that work?

[bktemp]

So, you are trying to measure the pulsating DC current?
You can easily get current transducers up to many 100A. They work from DC up to many kHz.
Depending on the isolation requirements and accuracy you can either use hall effect sensors (e.g. Allegro ACS7xx) or current sensors with integrated magnetics (LEM has some nice but expensive current transducers, but there are also many others).

[Jeroen3]

Just use a LEM sensor indeed. You can get them for PCB or wire-trough.
They'll withstand overloading, and will give AC and DC. LEM series CAS - CASR - CKSR are nice. Many choices in primary value, and you can choose if you want to series/parallel the sensing wires. You can even feed through them, as bugfix, if you really need to.

Note: LEM sensors are vulnerable to near* magnetic fields. From transformers or magnets!
*(~20cm, 8 or 0.6 freedom units)