CT saturation question


What are the implications of driving the CT connected to the high sensitivity input of an emonTX V3 to saturation? Will the measurement just be inaccurate at high current usage? Will the CT burst in to flames? Will the emonTX V3 be damaged? Or some combination of the above?

The reason I am interested is that for most of any 24 hour period our house is running at less than 300W, (peak daily usage I have seen is 20kW at Christmas when just about everything is on). The maximum load I have seen at any instant during a normal day is about 6.5kW (cooker, kettle, washing machine all on) and that would clearly drive the high sensitivity CT in to saturation.

Now I believe from reading the forum that the normal CT inputs to the emonTX V3 will be at the lower end of their accuracy at 300W and was wanting to connect both a normal CT and the high sensitivity CT to the whole house supply but don't want anything nasty to happen to the house or the kit!

I am currently using a CurrentCost Envi connected to emonCMS running on a Raspberry Pi with hard disk and want to move to a full OEM system to be able to monitor individual circuits. I guess one option is to change the burden resistor to a value suitable for the maximum load, something I may have to consider when I come to monitor individual circuits.



Robert Wall's picture

Re: CT saturation question

I have never seen a Spice model of the CT that we normally use, so it is difficult to be specific. You can see in the test report that when the CT saturates, the output waveform ceases to represent the input wave and the peak to rms value increases significantly. Saturation will also increase the core losses, but during tests I noticed that it only got moderately warm. I don't think anyone would guarantee (without rigorous testing) that no harm would come to it or its surroundings if it was abused badly enough.

Coming back to the output waveform, the rapid rate of change of flux as the core switches from saturation in one direction to saturation in the other results in a high induced voltage in the secondary, this is clipped by internal zener diodes at around 23 V (i.e. a pair of 22 V zener diodes connected in inverse series) and the input circuit of the emonTx has a 1 kΩ series resistor that limits the current to around 20 mA under those conditions, this current of course flowing to the supply or the 0 V rails via the internal protection diodes. It's clearly not a good idea to operate under these conditions continuously, and, assuming you have a 3.3 V emonTx, you might like to consider a pair of (say) 4.7 V zeners in addition, which would limit the voltage to around 5.3 V and the current to around 2 mA.

You can of course add a second resistor in parallel with the 120 Ω burden to lower the sensitivity to your normal maximum of 6.5 kW, and that is probably the pragmatic approach whatever else you do.

There's another suggestion that you might like to try: use a relay driven by the software to short the CT that's otherwise being overloaded. That way, the CT is happy as it's no longer overloaded, equally the input as it sees almost nothing (the burden will be the contact resistance of the relay!).

BraileTrail's picture

Re: CT saturation question


Thanks for the detailed response. The kit is now on order and I think I will try adding a resistor in parallel as you suggest to lower the sensitivity of one channel to start with. I assume that somewhere I will have to adjust a parameter in the software to match the new sensitivity?



Robert Wall's picture

Re: CT saturation question

>i>" I assume that somewhere I will have to adjust a parameter in the software to match the new sensitivity?"
Indeed you will. You'll find the formula in Building Blocks > CT and AC power adaptor installation and calibration theory.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.