Capgo datalogger, data logger, datalogging and data logging.

  

Magnetic Noise

 
 

Magnetically Induced noise

Magnetically induction is the most frequent method by which noise is introduced to a measurement setup. The principal of magnetic induction is simple - a voltage is induced into any conductor that is, either passed through a magnetic field, or morel commonly is exposed to a changing magnetic field. The magnitude of the voltage is proportional to the rate of change of the magnetic field and the area of the conduct loop. It can measure may volts - even hundreds of volts.

The current that can flow as a result of the magnetic field is generally quiet small (<1 mA) and is dependent on the degree of coupling between the field source and sensor wiring. Generally this coupling is quite weak. An interesting result of this is the the inducted noise tends to act like an current source.

Sources of changing magnetic field are many: Some may be intense but localized, while others are milder but cover a large area:

 

current carrying wires
transformers
electric motors
radio transmitters
vibrating or moving wires

computer monitors
arc welding equipment
rotating machinery
nuclear magnetic pulse

static discharge
earth's magnetic field
lightning strikes
solar storms

There are four broad methods that can be deployed to reduce magnetic pickup and its impact. The choice depends on severity, budget and practicality.

Avoidance

Prevention is better than cure. The old adage certainly applies to magnetic pickup because removing the noise once induced can be very difficult and sometimes impossible. Rule number one is simple: place wiring away from the source of the fields. Especially, void placing signal cables in proximity to power cables and heavy electrical equipment.

Avoid ground loops that may allow common mode magnetic pickup.

Loop Area Minimization

   Reducing the area of the antenna loop is an very effect method method of reducing magnetic pickup and there are many ways in which it may be achieved. The simplest and most effective loop area minimization strategy is to use twisted wire pairs. If the field is localized the pitch of the twist need to be fine (say one twist per centimeter) and conversely for wide area fields the pitch can be coarse (one twist per meter). Twisted wires works by local cancellation. The wires form a local antenna loop, but after the next twist the loop is inverted so a voltage with the opposite polarity is generated. This will cancel the voltage from the first loop.

The use of shielded or coaxial cable is not as effective as one may think. In theory the loop area of coaxial cable could be considered as zero, but this ignores the asymmetrical way in which the induced current tends to flow in the shield. The best solution is to use shielded twisted pairs.

If twisted pairs are not possible then the signal conductors should be located physically as close together as possible.

Loop area minimization is a good general principal to follow. It reduces a systems susceptibility to electromagnetic fields and also reduces possible emissions of electromagnetic energy.

Shielding

Shielding may seem like a simple solution - but it is not! There are two methods of shielding - using a ferromagnetic shielding or eddy current shielding using a conductive shield material.

Magnetic shielding at low frequencies is very expensive. Cables can be placed in steel conduit. For small scale problems, the use of mu-metal shielding may be an option. Mu-metal is a special alloy that is particularly effective in shielding, but it is not readily available and it is expensive.

At higher frequencies, eddy current shielding is more practical. The effectiveness of eddy current shielding depends on the "skin depth". Skin depth is the depth of conductive material required to establish field opposing eddy currents.

Filtering magnetic pickup

As mentioned above, magnetically induced noise tends to have a current source characteristic. The noise current will tend to remain constant regardless of the circuit impedance. As a result the noise voltage will be proportional to the circuit impedance. Now typically the input impedance of a measuring device is deliberately kept high to avoid loading the sensor's output and thereby introducing errors. This is exactly the opposite to what is needed to reduce the impact of magnetic-pickup!

Another problem is the frequency band of the magnetic pickup. Most commonly 50 Hz or 60 Hz and related harmonics are the main noise content. These relatively low frequencies make the filtering difficult, requiring large valued and sometime impractical components. If the signal has frequency components that are in the same frequency band we are faced with a near impossible situation. Sometimes the character of the noise is known and can be compensated - either by notch filters or by simulation and subtraction.

See the section on Filtering for more details.