When Silicon Controlled Rectifiers (SCR's) are used in electrical controls, we frequently experience line voltage distortion in the form of "notches" in the waveform. The types of equipment that frequently utilize SCR control schemes and thus experience notching include DC motor speed controls and induction heating equipment.

Line notches are just that - an irregularity in the voltage waveform that appears as a notch as illustrated in Figure 1. They are typically present in the waveform during SCR commutation or at the time when one phase SCR is being turned off and the next one turned on. For this very small duration of time, we actually experience a short circuit between the two phases. Of course when we have a short circuit, the current goes high and the voltage goes very low. This is exactly what is experienced during a notch.

Figure 1

The notch appears at the moment when current is actually rising very quickly but due to the shortage of the phases, the voltage is shorted and approaches zero. In the most severe cases, the notch does in fact touch the zero voltage axis. This causes the biggest problems.

Zero Voltage Crossing

During a normal cycle of sinusoidal voltage, the voltage crosses the "x" axis, or zero, at 0 degrees and again at 180 degrees. During normal conditions, there are two zero crossings in each cycle. Some electronic equipment is designed to be triggered on the zero crossing or when the voltage is zero. This allows equipment to be activated without the surge currents or inrush currents that would be present if switched while voltage was present. Some equipment uses the zero crossings for an internal timing signal. Some digital clocks operate this way.

When notches are present, particularly in three phase equipment, we can experience extra zero crossings. Instead of two zero crossings in each cycle of voltage, we can actually experience four notches. Think about it. There are now four signals in each cycle which tell other equipment to "turn on". That means the equipment will turn on twice as fast, run twice as fast or turn on at very wrong times resulting in damage.

The September/October, 1992 issue of Power Quality Magazine contained an article called "Clean Air, Dirty Power: The Two Timing Clock", written by Dave Hall. It described how a digital clock was running twice its normal speed due to the presence of line voltage notches. Reducing the notch depth solves this problem because zero - voltage crossings are limited to two per cycle as in a sinusoidal waveform.

Solution

The solution is quite economical and easy to install. There is no magic involved, it just follows good engineering sense. It is very easy to understand if we consider the DC drive or other SCR controller as a source of "notch voltage" and then analyze how that voltage is impressed onto the impedances in the circuit back to the source.

It is of utmost concern that we understand and consider where other sensitive equipment connects to that same voltage source. In order to protect the sensitive equipment, we must reduce the notches before they get to that equipment. We will do this through the creation of a simple voltage divider network.

Figure 2

If we add impedance, in the form of inductive reactance, in series with the SCR controller, and between the controller and the point of other equipment connection, (point B) then the notch voltage will distribute itself across the new impedance (reactance) and the pre-existing line to source impedance. If the added impedance is one-half as much as was already present, then 1/3 of the notch voltage is dropped across the new impedance and two thirds still remains at the point of common connection with the other equipment.

Figure 3

If the new impedance is equal to the existing input impedance, then the notch distributes equally across the two impedances. One-half the original notch voltage is now present at the point of common connection (B).

Notice that if the new impedance is added anywhere else but between the SCR controller and the sensitive equipment, it will have minimal impact on the notch voltage. Placing the reactance on the opposite side of point B offers no improvement to the notching problem.

Use a 3% impedance reactor to solve three phase voltage notching problems. Experience shows that it is normally sufficient to reduce the notch voltage, at the point of common connection with other sensitive equipment, to about 50% or less of it's initial value (depth). Of course this eliminates the multiple zero crossings and typically solves the interference problems with neighboring equipment. It is typically not recommended to use a 5% impedance reactor with SCR circuits because the reactor not only reduces the depth of the notch, but it also increases the notch width. Excess impedance could increase the notch width (time) too much causing problems in the controller itself.