Charging battery from mains is series with a capacitor

[poynt99]

OK, just a quick note:

The FWBR voltage measurement is made across the AC (~ ~) terminals, not the DC side. That way you capture not only the power going into the battery, but also the power being dissipated in the FWBR.

[havuhung]

More on the subject: http://www.johnsavesenergy.com/CapacitiveBatteryCharger.html

/Hob

Hi All,
Thank a lot

havuhung

[poynt99]

Does a simple capacitor step down the grid voltage while driving a load?

The evidence is in the experiment.

Normal power in RLL without the capacitor would be: (.707 x 170)²/100 = 144W
Power in RLL with 10uF capacitor in series is: (.707 x 60)²/100 = 18W

Pure Resistive loads don't give a hoot about phase shift or power factor; the only thing that matters is the rms voltage across them.

[poynt99]

A clear demonstration of probe positioning and the impact it has on the displayed phase relationship between current and voltage, and ultimately on the power computation.

"cap_simple02.png" is showing a series-adding positioning of the scope probes; one across the capacitor, and one across a series resistor. The probes are placed + - + - as shown. Note the scope shot "cap_simple02_scope.png" clearly showing that current is leading the voltage by 90º.

"cap_simple03.png" is showing the "standard" way we all connect our scope probes when measuring power in an element, whether it be the Grid source (as in the case shown), or a dissipating element such as RLL. That is, the probes are positioned in series-opposing (+ - - +) so that the probe grounds may be commoned to a single point. What this does is flip the phase relationship between the scope channels by 180º. The result being that in "cap_simple03_scope.png" we still see the current leading the voltage, when in fact it should be the other way around. The current through a source, and the voltage across it are always in anti-phase.

So in order to restore the phase relationship between the voltage and current when making these kind of measurements, we need to invert one of the scope channels.

This also holds true when we apply the same probe-positioning technique when measuring the power dissipated in any other elements in the circuit, such as RLL. Our probes are going to be in series-opposing, so one channel needs to be inverted on the scope in order to restore the correct phase relationship, and obtain the correct polarity of power.

Having done so in both cases will not only result in displaying the correct phase relationship between the current and voltage, but it will produce the correct polarity when using the MATH function in the scope to obtain an average power measurement of the element of interest.

[poynt99]

And now to the battery charging application of this circuit.

"bat_charge01.png" illustrates the circuit with a large capacitor replacing the battery. This charges up quite quickly so it is all over in about 1 second. But as it turns out, this is a convenient number as we shall see. Notice the dropping capacitor has been reduced from 10uF to 5uF.

"bat_charge01_scope.png" shows the battery charging up from the initial 12.5V to about 14.35V in 1 second (1s).

"bat_charge02.png" shows one way to measure the power going into the charging circuitry and battery. Notice that the probes are placed in series-opposing and that the scope MUST be gnd-isolated in order to make this measurement. Note, lifting the gnd on your oscilloscope is dangerous and done at your own risk.

"bat_charge02b_scope.png" illustrates the computation involved in obtaining the average power going into the charging circuit and battery. The current probe channel was inverted to restore proper phase, and was multiplied by 10x to account for the 0.1 Ohm resistor. The average power computes to +2.825W. It so happens that this works out to 2.825J as this is over a 1s period.

"bat_charge03.png" illustrates the probe setup to measure the Grid power. Once again the probes are in series-opposition, so the current channel will be inverted to restore proper phase. The scope does not require gnd isolation if the bottom terminal is the Grid Neutral.

"bat_charge03b_scope.png" illustrates the computation involved in obtaining the average power used from the Grid. The current probe channel was inverted to restore proper phase, and was multiplied by 10x to account for the 0.1 Ohm resistor. The average power computes to -2.83W. It so happens that this works out to -2.83J as this is over a 1s period.

So despite the phase shift and PF changes resulting from using a step-down capacitor, the energy equation (used vs. sourced) remains in balance. Note: Kirchhoff's laws apply to power as well as voltage and current. i.e. the sum of the powers in a circuit must equal zero (unless of course you have an OU device in hand).