I'm working on a project to safely monitor a 3-phase, 220V AC line with a microcontroller for power metering. My goal is to measure RMS Voltage, Real Power, Reactive Power, Apparent Power, and Power Factor, as well as analyze signal harmonics. I initially tried using a ZMPT101B voltage transformer, but I've run into a significant issue: it introduces a noticeable phase delay (around 60-62.5 Hz at the fundamental frequency). I'm concerned this delay will drastically affect the accuracy of my power calculations, and I'm also unsure how it might interfere with the measurement of higher-order harmonics. Could anyone suggest alternative methods or components for safely sampling the AC waveform with minimal distortion and negligible phase shift? I need a solution suitable for accurate power metering and harmonic analysis. Thanks in advance for your help!
-
\$\begingroup\$ It seems you are asking a shopping question. \$\endgroup\$Andy aka– Andy aka2025-07-07 20:44:02 +00:00Commented Jul 7 at 20:44
-
\$\begingroup\$ Phase delay can be "compensated" with a "good" acquisition and DFT techniques ... \$\endgroup\$Antonio51– Antonio512025-07-08 06:55:40 +00:00Commented Jul 8 at 6:55
-
\$\begingroup\$ An analog front end (AFE) with digital isolators, is the good way. EV-ADE9153ASHIELDZ Arduino Shield \$\endgroup\$Marko Buršič– Marko Buršič2025-07-08 08:08:51 +00:00Commented Jul 8 at 8:08
2 Answers
Not sure what "phase delay" is. Phase shift is measured in degrees, and group delay is measured in seconds. In either case, you can measure it in a test setup. Then it can be compensated for in software.
The transformer definitely will affect the amplitudes of the high-order harmonics. Consider a high-voltage-rated audio transformer. Something designed for a 10 kHz output will be pretty flat down at 1 kHz, and that is the 17th -ish harmonic. Only you can say how many harmonics matter in your system.
Current transformers have a pretty wide bandwidth, and again you can compensate for any phase delay in software.
One alternative is to make the microcontroller live and then use a serial link through an optocoupler to read the data (or just insulate the display and the controls)
Back to your circuit, be sure to use a non-inductive resistor in the source circuit (R' in the datasheet)
