Low Power Factor


I have setup an esp32 and emonlib to monitor my home power consumption. I have a 70A/1V ct clamp and zmpt101b transformer on a custom board to measure current and voltage. It seems however that at low current draw the power factor is way off. Currently sitting at 1.8A and 0.55 power factor. Would this seem correct? I have a computer running and a few LED lights as well as things always plugged in such as the fridge. I have calibrated the phase angle with my kettle and got a calibration value of 1.68. Any insight would be much appreciated.


Welcome, Tom, to the OEM forum.

I assume you got a p.f. of around 0.99 or better after calibration with your kettle?

A p.f. of 0.55 doesn’t sound too unreasonable for computer, LED lights and an induction motor in the fridge, though I think would have expected a little better.

The phase error of your c.t. might be changing a bit down at 1.8 A (2.5% of rated current) so that might well be worsening the power factor a little. I’ve only tested and measured the c.t’s from our shop, so I can’t say for certain how well or otherwise yours is likely to perform at low currents. One thing is certain - the way the magnetics work means that errors, and particularly phase errors, increase at very low currents, and it’s inherent, so you can’t avoid it.

Thanks for the quick reply :smiley:

Yeah i got around 1 p.f. to around 4 decimal places on average with the kettle.

I was wondering if the CT clamp was the problem. I did hook it up to a oscilloscope and with the low currents it shows a more square wave than a nice sine wave so i guess it isn’t reading accurately at these low currents and as you say there is probably phase errors too. Could the ones from your shop that are say 100A/50mA be better than the 70A/1V as i can choose the burden resistor to get a better reading?

How did you get your low current - was it a ‘well-behaved’ load, one that is resistive and linear? Because it doesn’t sound right that the c.t’s output becoming square.

We do at least know the phase error of the YHDC SCT-013-000 (see the ‘Learn’ article), and you can indeed choose your burden, noting that c.t’s like nothing better than a short circuit to work into (so the lower the voltage you need, the lower the power it needs to transfer). The other question - do you actually need 70 A? If this is the rating of your supply, then it’s likely your maximum current could be a lot lower.

It helps us to know a bit of background - where you are in the world helps quite a lot. Can you add this to your profile please, because this will tell me what your electricity supply is like.

The load was what i described earlier so defiantly not a ideal resistive load. Sorry if I’m asking obvious questions, but would say a TV and computer UPS powering a PC cause a less sine wave looking current draw? I imagine they would using cause an odd waveform from the SMPS so i guess the p.f. is effected by this. I’m in the UK btw.

In that case, I wouldn’t worry about your c.t.

To show you the sorts of waveshapes that are possible, here’s a very spiky current waveform (Washing machine brushless d.c. motor on spin by @dBC Phase measurement and correction in IoTaWatt - #17 by dBC)

And there’s a picture of a most distorted channel made up of a “boat load” of CFLs here
The importance of continuous sampling - #5 by dBC

Ah okay. I’ve posted a oscilloscope output of the voltage(red) and current(blue) sensors below. I guess this could be right, probably just me over thinking things.

No, with the sort of load you have, I wouldn’t raise too many eyebrows at that.

It’s reasonably symmetrical, discounting a few spikes, there’s a noticeable peak at peak voltage, which you’d expect if there are bridge rectifiers and reservoir capacitors providing a d.c. supply to something. Current appears to be leading voltage by quite a bit, so your 0.55 power factor doesn’t look silly for those waveforms.