1
\$\begingroup\$

I have connected a capacitor to do some power factor correction on a load. After I connected the capacitor parallel to load, I checked the zero cross detection waveform, which is getting input signal from an ACS712 current sensor, and found that the square wave looks distorted. Is this a usual phenomena on PF correction using a capacitor? Is there any recommended way to eliminate such distortion?

enter image description here

The above screenshot is obtain from logic analyzer enter image description here and from oscilloscope

Below is the schematic diagram with relevant to current ZCD, the distortion happen every cycle and not during the switch on of the relay. enter image description here

This the current sensor use in the circuit.

enter image description here

Update:I put current limiting resistor that have an effective resistance of roughly 32.55Ω in series with the capacitor of 8 μF for PF correction on a grinder machine that previously showing distorted waveform as above and now it looks ok.

photo of capacitor 8μF(black cube) for PF correction connected in series with parallel resistors of 180Ω,180Ω and 50Ω which having effective resistor of 32.55Ω while the yellow(1 μF) and brown(0.67 μF) capacitor is connected to another 50Ω series resistor.

enter image description here

Photo of load using for PF correction(grinder machine) enter image description here

photo of oscilloscope output waveform enter image description here

Is there any possible explanation why the distorted waveform disappear by just adding resistor in series with capacitor?

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Hall effect chips have some noise output at 0 current. Maybe your zero cross circuit needs some damping/filtering. Post your schematic if you want better help. \$\endgroup\$
    – Aaron
    Commented Jun 22, 2022 at 18:55

2 Answers 2

1
\$\begingroup\$

Add some hysteresis to U2:A, the comparator. This can be done by connecting a large resistor (>470k) between pins 1 & 3.

\$\endgroup\$
1
  • \$\begingroup\$ Very sorry for the late reply, I tested with a 510K ohm resistor and this method is not working for me. \$\endgroup\$
    – chuackt
    Commented Jul 1, 2022 at 6:49
0
\$\begingroup\$

A ZCS is normally done with with a diode bridge at low threshold or in logic using an XOR gate with a the same signals delayed by some RC=T delay pulse width.

Using a Linear Hall current sensor requires a linear DSO to view the effect of RLC ringing and lack of damping on the unknown input waveform.

You cannot accurately interpret an analog signal with logic analyzer resolution of 1 bit unless you incrementally change the threshold to see the amount of ringing due inductive effects with C.

Define your network with photos, load datasheets and a schematic and define your output/input spectrum expectations with a C load in terms that will define the problem and solution.

A better question is do you really want a ZVS or a ZCS pulse or both to measure PF from the phase difference? Or do you want to compute RMS PF?

When the RC impedance is too small with an RL load impedance at some f, the result is a higher frequency resonance if R is small compared to X(f).

  • revise your question with accurate details for a better answer.
  • With a schematic , spec and purpose, a better solution is easily achieved. Without that, we are in the "dark" with you .
\$\endgroup\$
2
  • \$\begingroup\$ Did you understand what I wrote? \$\endgroup\$ Commented Jun 28, 2022 at 19:41
  • \$\begingroup\$ My bad, overlook your comment. I don't understand the 1st to 4th line can explain more? I'm using a ZVS and ZCS pulse to measure PF since my school project specify have to stick to this method. \$\endgroup\$
    – chuackt
    Commented Jul 1, 2022 at 7:05

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.