I am creating a system for detailed measuring of power consumption in my home. For that I am using Shelly EM, which is using 50A hall sensor clamped over my main line (single phase 240V AC) inside my circuit breaker box. Then I am gathering data from Shelly EM using ESP MCU that uploads them to my server.

Shelly only provides following data:

  • Voltage
  • Watts
  • Power factor

Out of curiosity I created a graph for current as well, despite shelly is not providing it in its JSON output. I used following calculation (power is watts, pf is power factor, shelly report it negative most of time, no idea why):

current = power / (voltage * abs(pf))

My problem is that the reading I am getting are sometimes not making much sense, for example here it was reading 18A, but my circuit breakers are 16A

current graph

I suppose the problem might be that

  • My calculation is wrong (I tried also removing power factor from it entirely, the graph was showing lower current overall, especially when consumption was low, but peaks were exactly the same, in this case again 18.4A)
  • Shelly is reporting wrong values (maybe that hall sensor isn't so accurate? they report 1% accuracy though)
  • Circuit breakers have some tolerancy to allow higher current than rated for? Unlikely IMHO
  • Something else is wrong that I don't understand

EDIT: seems that circuit breakers allow more than 16A for limited amount of time, which might explain this

Graph zoomed: current meter

  • 2
    \$\begingroup\$ It would help immensely if you could state what physical loads were connected and running when you made measurements. Your calculation method is correct. Try calibrating your measurements with a known load. \$\endgroup\$
    – Andy aka
    Sep 30, 2022 at 10:57
  • \$\begingroup\$ Sure, the load was electric water kettle and a washing machine + many small home electronics (wi-fi router, switch, etc. but they have less than 10W most of the load was the big stuff). \$\endgroup\$
    – Petr
    Sep 30, 2022 at 11:40
  • 1
    \$\begingroup\$ "Circuit breakers have some tolerancy to allow higher current than rated for? Unlikely IMHO" They can usually take like 2x their rated current for several seconds. So it's not unlikely at all. \$\endgroup\$
    – Klas-Kenny
    Sep 30, 2022 at 11:41
  • \$\begingroup\$ I actually started the kettle just for experiment, I saw in graph that when my washing machine is running it's about 10A and I remembered that when the kettle ran it was also close to 10A, but also that I definitely ran both in the past and it never triggered the circuit breaker, so I wanted to understand what's going on, started both to see and to my surprise it indeed exceeded the circuit breaker rating (at least according to this calculation). \$\endgroup\$
    – Petr
    Sep 30, 2022 at 11:45
  • 1
    \$\begingroup\$ @petr ... yes, but not for very long! You'll see the curves show it tripping after approx 10 seconds at 3x rated current. \$\endgroup\$
    – jonathanjo
    Sep 30, 2022 at 12:38


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.