Dealing with High-Frequency Noise
Dave Hall, PE
Power Quality Assurance Team
Wisconsin Electric Power Company
Compared to other electrical problems, such as harmonics, relatively little has been written about high-frequency noise. Perhaps this is because relatively little money can be made from solving high-frequency noise problems, although the symptoms can be just as baffling as those caused by more expensive problems.
High-frequency noise can cause electronic equipment to refuse to operate, to operate incorrectly or to halt an otherwise smoothly operating piece of equipment. It can cause one solid state motor drive to “trip” others (this is sometimes called “cross-talk”) or cause relatively small devices such as electronic scales or controls to become erratic.
All electronic devices operate on DC internally. This means they must convert the AC electrical power into DC in what is usually called a rectifier section. Some rectifiers, such as those used in a radio, use simple diodes. Others, such as those used in a solid-state motor drive use Silicon Controlled Rectifiers (SCRs). SCRs (also called “thyristors”), are diodes that can be switched on and off. It is this process of switching that causes the high-frequency noise.
In a three-phase rectifier, there commonly are six SCRs that conduct in pairs. Often, when one pair is being turned off and another pair is being turned on, there is a brief instant of overlap when two pairs are conducting. This causes a temporary condition similar to a short circuit. The condition is so brief that the SCRs are not damaged, but a voltage dip is caused at this “commutation point.” Rather than thinking of the condition as many individual voltage dips or notches in the waveform, it is easier to think of it as ongoing high-frequency noise in the same way as a broad expanse of many trees is considered to be a forest. A voltage waveform displaying this condition is shown below.
High-frequency noise may come and go on a particular drive, or it may not happen at all. It is so common, however, that it can be assumed to exist whenever solid-state motor drives are operated without input line reactors or isolation transformers. It is not always a problem; many devices can peacefully coexist on a noisy electrical system. Occasionally, however, the noise level is high enough or the victim device is sensitive enough that action must be taken.
When the problem does present itself, there are two options: clean up the noise or filter the power used by the victim unit. The voltage notches extend from the drive unit (or other power rectifiers using SCRs such as an induction heating unit) upstream to the first transformer. For this reason, the noise pollution can be “dammed up” near a drive by using an isolation transformer or line reactors (also called “chokes” or “inductors”) on each offending unit.
It will not work the other way. Generally, polluted voltage will not be “cleaned up” by adding a transformer or choke on the victim unit. In the case of drive “crosstalk” where drives are generating noise that trip each other, adding a choke (line reactor) or isolation transformer to each unit is the only practical solution.
In a case where the only victim is a small electronic device such as a scale or control unit, though, the noisy voltage can be economically cleaned up by installing an electronic filter. These filters often are sold as voltage transient (or “impulse” or “surge”) protectors. Since the noise actually consists of a barrage of such transients, however, the filters work well for small applications.
Line reactors or isolation transformers on all motor drives and other such devices are always a good idea for noise limiting and other reasons. Regrettably, they are not used often enough in the real world because of economics. An electronic filter is sometimes as effective and much more cost-effective when only a few small electronic devices are having trouble.