I'm building an energy metering circuit with discrete parts and precision is not a big deal in this case. One friend had the idea of rectifying both the voltage and current waveforms before sampling them with a MCU. His idea is the following: we put big enough capacitors after the rectifiers so the ripple is neglectable; since the voltage "V" (ideally) will be sqrt(2)*Vrms and the current "I" will, for a constant linear load, (ideally be) sqrt(2)*Irms, then the RMS power will be Prms=V*I/2 and the energy will be the sum of Prms on a given interval. By rectifying the inputs and sampling them directly with the ADC he hopes we'll no longer need the op-amps for sensing and also simplify the firmware (as multiple samplings will no longer be needed for calculating the RMS values of voltage and current in a given interval). The phase shift between current and voltage would then be accomplished by two zero cross detectors (one for voltage and one for current) that in turn would trigger timer interrupts on the MCU, and the time differences would be used to calculate the power factor. I can't see the shortcomings of this (for approximately constant and linear loads, at least), but obviously they exist because every metering circuit I've came across don't rectify the voltages/currents. So I'd like to know your opinions on this idea...

  • \$\begingroup\$ Have you tried running simulations of this arrangement in a suitable simulator such as LTSpice? You may get interesting findings if you try different kinds of loads such as triac-dimmers, switching power supplies, etc. \$\endgroup\$
    – nanofarad
    Jan 13, 2022 at 18:15
  • 4
    \$\begingroup\$ Your proposal is significantly flawed and not well thought out. That's not what any half decent power meter does and neither is power factor computed that way. This and this and this may be of some interest. \$\endgroup\$
    – Andy aka
    Jan 13, 2022 at 18:36
  • 1
    \$\begingroup\$ Does this answer your question? How to sense the zero crossing of current? \$\endgroup\$
    – Andy aka
    Jan 14, 2022 at 15:16
  • \$\begingroup\$ @Andyaka thanks for the information. I really see now that this was a very bad idea haha. Yes, it does answers my question about the zero cross. \$\endgroup\$ Jan 17, 2022 at 13:43

1 Answer 1


For linear loads and no distortion of the supply voltage, the described scheme seems fine. With any distortion, the harmonic content will contribute little or nothing to real power, but will likely register harmonic voltage and current as delivering real power.


Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.