# Homework Help: Phase shift using a series capacitor

1. May 5, 2007

### JackFoligie

I have a single phase mains supply which I need to split it into two phases, then shift one of them by 90 degrees relative to the other (with peak values of both waveforms having the same magnitude).

I know this is generally done by two branches in parallel, with a series capacitor in one of the branches and a series resistor in the other. As I see it, the magnitude of the capacitive reactance (Xc) in the one branch and the resistance (R) in the other have to be the same if the magnitudes of the two waveforms are to be the same

i.e. |Xc| = | 1/(2*pi*freq*C) | = |R|

Is this a correct assumption?

I know that for this application, a paper capacitor is the best option. If my mains supply is 230V 50Hz, does anyone know how I go about finding the required series capacitance and it's rating? None of the text books I have access to explain this topic in any detail. Also, everything I've found seems to suggest that a capacitor automatically sets up a phase shift of 90 degrees (which may only be the case for an "ideal" capacitor). Any advice or suggested online resources would be greatly appreciated.

Jack

2. Jul 10, 2008

### rborowiak

I know this is an old thread, but I too want to do the same thing, the savings in reduced filter capacitance would be worth it.

it gets tricky when dealing with -j. The voltage drop across a resistor in series with the capacitor will see the phase shift voltage across the capacitor plus the supply voltage. So the voltage might be 400 volts rms across the resistor. And don't go thinking your gaining anything your'e not. The current is also phase shifted. So you can't follow P=(E*E)/R.

I use a 1mfd mylar capacitor to replace a 10k 2 watt resistor in my power supply's, only because of heat savings, in the supply. And that's at 60 Hz. Xc of a 1mfd @ 60 Hz is like -j 2,654 ohms. And I use a 630 VDC capacitor for 125 VAC mains for 240 VAC you need like 1600 VDC.

Last edited: Jul 10, 2008
3. Jul 10, 2008

### rborowiak

I know this is an old thread, but I too want to do the same thing, the savings in reduced filter capacitance would be worth it.

it gets tricky when dealing with -j. The voltage drop across a resistor in series with the capacitor will see the phase shift voltage across the capacitor plus the supply voltage. So the voltage might be 400 volts rms across the resistor. And don't go thinking your gaining anything your'e not. The current is also phase shifted. So you can't follow P=(E*E)/R.

I use a 1mfd mylar capacitor to replace a 10k 2 watt resistor in my power supply's, only because of heat savings, in the supply. And that's at 60 Hz. Xc of a 1mfd @ 60 Hz is like -j 2,654 ohms

I think it would be cost effective to get the electric company to run 3 phase to your'e location and run 3 supplies off of the each of the phase of 3 phase.