What I Learned this Month

An example of a ground loop which causes voltage noise between two instruments due to magnetic field pick up. Courtesy of P. M. Bellan, inventor of the Loop Slooth.

If you have a pesky 60 Hz hum in your low noise electronics or some high frequency noise on a signal line which appears when a motor near by turns off or on, you may have a ground loop problem. A simple device from Paul Bellan, called the Loop Slooth, may help you find and eliminate the ground loops in your product.

Prof. Bellan is Professor of Applied Physics specializing in plasma physics at Caltech by day (and many nights), and moonlights as the inventor of the Loop Slooth. He was often faced with the problem of spurious noise on very low level signals from plasma diagnostic sensors close to very noisy plasma chambers. The problem was more often than not ground loops connecting instruments and sensors which picked up the large transient magnetic fields in the lab which turned them into induced voltages.

He developed a simple device to “tone out” which cables were part of the loop which was usually enough information need to fix the problem. When he realized other researchers probably had similar ground loop problems and could benefit from his sniffer, he turned it into the Loop Slooth.

It has two units. The “exciter” generates a 100 kHz “tone” on part of the ground loop using an induction coil that clamps around an identified ground line. This works on the same principle as a current probe, in reverse. The AC current in the exciter coil generates a changing magnetic field around the cable that’s inside the coil, which induces a current in the cable- the ground line and the other wires.

The second unit is the detector with a sharply tuned circuit for 100 kHz. The detector uses as a Rogowski coil which is a clever way of making a toroid with a gap so that suspect cables can be slipped right into the toroid.

The Rogowski coil, shown in the figure from Wikipedia, is an open toroid with both ends of the coil coming out the same terminal. The voltage between the ends of the coil is very sensitive to the changing flux through the inside of the toroid which is due to the current passing through the center of the toroid.

The detector box has a meter and an LED that lights when it picks up significant 100 kHz tone. This indicates that one of the cables that’s been slipped inside the detector coil is part of the ground loop. With the clamp on the exciter and the open toroid, Bellan says, a cabling system can be debugged as it stands, with no changes or disconnections necessary.

While you may be careful in laying out the wiring in your system and think you have minimized ground loops, as soon as you have multiple instruments plugged into multiple power plugs, with cable shields connected together, you have ground loops.

Bellan suggests a good habit is to always plug instruments that will share cable returns, into adjacent plugs, to minimize the loop area of their returns, and to twist the coax cables together to reduce the area through which stray magnetic fields can pass and induce voltages. As a final practice, ferrites around the power cables will increase the impedance of the ground loop and reduce the induced current.