Noise Debug 101: Electromagnetic Conduction

Electromagnetic Conduction: The process where a conductor placed in a changing magnetic field (AC) causes the generation of a voltage across the conductor.

This coupling type is classified by a changing magnetic field across a wire or wire loop(s). A voltage and/or current are produced in the wire cut by the changing magnetic field. Note: Only changing magnetic fields can induce a voltage; therefore this is only applicable to AC sources.

The classic model of magnetic coupling is a transformer: the AC source flows through the primary winding (coiled wire loops), creating a changing magnetic field in the “core” of the transformer. The changing magnetic flux in the core then passes across the secondary winding (coiled wire loops) and creates a voltage on, and usually current flow through, the secondary coil.

Vout =Vin*(# windings in secondary / # windings in primary)
Vout = Vin*(Lsec / Lpri)

Voltage on the secondary winding is based on the ratio of the inductance (L) on primary and secondary coils - usually the ratio of the turns. Voltage is fairly independent of the impedance of the secondary coils so lowering the secondary load impedance won’t make much difference.
The transformer example uses an iron core because iron is an excellent magnetic conductor. High frequency transformers will use an “air core” for better linearity versus frequency. Notice in the transformer illustration that the magnetic flux flows perpendicular to the current flow in each wire of the transformer windings. Note: This will be key in the magnetic coupling test experiments. The magnetic field is perpendicular to the current flow in each turn, and the magnetic flux flows parallel to the wire-coil winding axis.
In Noise Debug 101 - Part 2, we saw how a grounded shield was very effective at stopping the electrostatic noise conduction. However, an electrostatic shield is very ineffective at preventing magnetic conduction because the thin metallic shield does very little to stop the magnetic flux.
Electrical Engineering Concepts

• The letter “L” is used as the symbol in equations in EE. Inductance is measured in “Henries.” An inductor will impede a change in current flow.
• The formula for impedance of an inductor is: XL = 2π f L
• Inductance results in an AC signal impedance. An inductor will “impede” AC current signal flow, the same way that resistance “resists” DC current flow.
• An inductor is (almost) always made by a coil of wire, with 1 or more turns. However, even a straight wire has a very small bit of inductance.
• The inductance of a coil is related to the number of turns and the area of the coil. (The area of the coil will be key later in this part.)
• To induce a voltage signal in a coil, the magnetic field direction must be parallel with the axis of the coil (see above).
• The magnetic field generated is dependent on the inductance of a coil and the current flowing through the coil.
• Inductance = loop area * the number of turns
• More turns or larger area makes more inductance and thus more noise conduction.
• Magnetic fields cannot be shielded or destroyed, only diverted.

Lab Setup Example

In this example, the magnetic noise source may be a transformer in a power supply or the ballast in a florescent overhead light, denoted by the coil. Note: Most 12v DC LED lights have a high frequency circuit containing an inductor in the LED bulb to control LED current. Consider these as a possible noise source.

The “receiving inductor” is formed by the loop of: Signal source, Signal wire, Amplifier input, and the return Ground wire. The area of the loop is based on the length of the cable and the separation distance between the two wires. Twisting the wires will match and cancel the induced voltage signal at each twist.

Solution: To reduce magnetically conducted noise coupling,

• Turn the loop or turn transformer/power supply 90 degrees;
• Move away, move out of field – move tether extensions away from magnetic fields (older CRTs will have major magnetic fields);
• Reduce the loop area - twist exposed wires to keep them tight;
• Magnetic shielding – for high frequencies use Aluminum, Steel, or Mu Metal:
• Aluminum (low cost) – about 24db (magnitude of signal reduced to 1/16 of the original) at F <10Khz
• Steel (low cost) – about 20 dB (reduced to 1/14 of the original) at F <10kHz, heavy
• Mu Metal (very expensive) – about 10 db at higher frequencies; (Fe, Cr, Ni)

For situations where you have two or more long tether cables (ie., from a ceiling mounted commutator), you may see a reduction in noise by simply twisting the two tether extension cables together. By twisting the cables together, the amount of noise induced into both cables will either cancel, as above, or be closer to the same on the electrode and reference wires so the Common Mode Rejection of the Neuralynx Data Acquisition Systems will cancel the common mode noise.
Tests for Determining Magnetic Coupling
Note: Remember to use the audio monitoring of a channel in Neuralynx’s Cheetah software to listen for changes in the coupled noise during these tests.

• Use a headstage with a shorted input; this will avoid electrostatic coupling, and move to the suspected magnetic sources. When the source is found, rotate the headstage in all 3 axes to minimize and maximize the coupling. If the coupled noise changes when rotated 90 degrees, the coupling is probably magnetic.
• Twist or tie long cables together to minimize the “loop area” between the cables/wires. If this has an impact on shielded cables, the coupling is probably magnetic. Keep the twists in place and move the cables away from the source.

A good webpage on inductance can be found at: www.allaboutcircuits.com/textbook/alternating-current/chpt-3/ac-inductor-circuits/