I am using 2 CR8401-1000-A current transformers in opposite polarity to sense current on either side of a load. The Idea is that when the load is not in failure mode that the sensed current waveforms will be exact opposites and when "added together" will cancel each other out. In failure mode, or ground fault, there will be an alternate path for current. The wave currents will no longer cancel and yield a net current. This signal is amplified by an op amp circuit, and rectified by a precision full wave rectifier. The now DC signal is sent to an ADC. A micro controller coupled to the ADC will throw a breaker at programmed current thresholds. My question is this, can the 2 current transformers be connected in reverse polarity in parallel directly, or should there be a buffer circuit between them?

I know that I should be able to run both lines from the load through one transformer, and their signals should cancel in no fault mode, but it doesn't seem to be working out. See the following post.

 

My current design with a single wire through one transformer yields 250 mv/ma after amplification. The micro controller runs a digital display working just like an ammeter, so you can easily see the output. My next baby step was powering a light bulb on 12VAC (approx .42 amps) running both wires through one transformer. The resultant output after amplification was 5 VAC, far from canceling out.This threw me for a loop. I am amplifying something that shouldn't be there.

I studied real GFI breakers to arrive at the dual transformer idea.

Any idea why the single transformer design does not work?

Any idea on the best way to interface 2 polarity opposing transformers together? I was thinking about connecting the output of each transformer to a unity gain op amp circuit, then connecting the output of those together.

 

Any idea why the single transformer design does not work?

Do you have the current in the two wires going in opposite directions through the transformer?

 

Do you have the current in the two wires going in opposite directions through the transformer?

It is a light bulb operating on 12 VAC with both wires coming from the bulb going directly through the transformer. I would have to say yes.

 

250mV/mA with a current of 0.42A yields 105V so 5V says that considerable cancelling has occurred.

 

250mV/mA with a current of 0.42A yields 105V so 5V says that considerable cancelling has occurred.

Please allow me to clarify. If one ma flows through a single wire inside the transformer, and the output of the transformer is amplified by a factor of 220, then that output goes through the precision rectifier with a gain of one, the output of 250mv will be seen by the adc. The programmed microcontroller will interpret this voltage and display 1 ma on the display.

With the bulb and the 2 wires, I am seeing from the transformer, after amplification by 220, a perfect waveform of 5VAC which will get doubled by the precision rectifier to 10vac. This translates to 40ma. Not a total cancellation, and for a true GFI to be safe it will need to trigger at currents of 4ma or smaller, so this is not acceptable. I also assume that this error will be proportional as currents go up from .42amps.

The only possible mistake I could be making is that I am not putting a resistor across the transformer as recommended , which would take the current signal from the transformer and yield a voltage signal. Instead I have been relying on the 43 ohms of internal resistance in the trasformer winding which has worked so far with one wire samples......maybe until now.

 

You need to use a load resistor on a current transformer.
They try to output a current, and without the load resistance, the voltage can be very high- and more or less meaningless.

 

You need to use a load resistor on a current transformer.
They try to output a current, and without the load resistance, the voltage can be very high- and more or less meaningless.

Thanks, I will try a load resistor and see what happens.