Measurement artifacts are the technical reasons why the results that you observe may not be accurate. Some of these artifacts are fairly well known in the industry. Others are more obscure. The artifacts described are not related to any single probe, probe type, or manufacturer. They are not all defects in design since they may or may not contribute to any significant error. Some artifacts are due to design and are present in only some probes. In other cases, the artifacts may be the result of a conscious decision and tradeoff made during the design process. The following are some of the more common measurement artifacts.
Most instruments today have digital readouts. Zero drift is easier to understand when you think of an analog meter movement. When no signal is present, the needle should be on zero. When it is off a little, you correct the problem by adjusting a small screw or knob on the meter. You perform a similar task when you use a scale to weigh yourself. For some instruments, you must ensure that the instrument is not measuring anything while you “zero” it in the same way that you set the scale to zero before you get on it.
Digital meters have the same requirements, but it is a bit more difficult to visualize and to detect zero drift. And most digital meters do not automatically display negative numbers. You can see when the needle goes below zero on an analog meter. A digital meter may show a negative sign, a negative sign with a value, or simply 0.0 units.
Zero drift is largely caused by temperature changes. This is due to both the changes in temperature that occur as the instrument warms up and from changes in the ambient temperature over time. Equipment design impacts the amount of drift. Probes with high-gain amplifiers magnify the amount of drift. Some equipment is designed with the diodes close to saturation so that there will be little drift. But this design approach gives up RMS detection, which can lead to far bigger measurement errors than zero drift.
To minimize zero drift:
Let the equipment acclimate to the ambient temperature for at least 10 minutes. Do not take an instrument out of a heated car and immediately try to make measurements in the cold or vice versa.
Turn the equipment on, set it up, and then zero it. Wait 2 to 5 minutes and zero it again. This allows the equipment to warm up and stabilize.
To check for zero drift:
Shield the probe occasionally and make sure that you get a zero reading. A positive value indicates positive zero drift.
Check for a negative indication or all zeroes. Most sites have some energy present. If the meter is always indicating 0.0, then the system may have drifted in the negative direction.
Survey instruments can be oversensitive to signals below their rated frequency range for a variety of reasons. The most common problem occurs because a meter and probe, especially when separated by a cable, can be at different voltage potentials when in close proximity to a radiator. The problem is acute near AM radio antennas (560 kHz to 1600 kHz), near industrial process equipment that operates at 50 kHz to 400 kHz, and near power frequency sources, such as power transmission lines. This artifact occurs because the impedance of the sensor antenna becomes as high as or even higher than the impedance of the transmission line. The result can be some abnormally high meter readings.
There is a simple way to tell whether the indicated value represents a true RF field or a voltage pickup. Simply shield the probe with a metal can, aluminum foil, or some of the fabric used in RF protective garments. If the meter indication drops significantly, then you were reading a true RF field. But if the shielding has little impact on the indicated value, then you are looking at an indication of a voltage field. The shielding will block RF energy from the sensor but will have little impact on the survey instrument’s tendency to act as a voltmeter. Note that it is important that you do not touch this shield or let it touch a conductive object.
To minimize or eliminate the problem:
If the meter and probe are isolated electrically, then the problem cannot occur. The only way to do that is to connect the probe to the meter with a fiber optic cable.
If the meter and probe are at the same or almost the same electric potential, then the problem is minimized. If possible, eliminate the cable and connect the probe directly to the meter. While this is often an awkward way to work, this is the one time when it is the right thing to do. If you cannot do that, coil the connecting cable up and move the handle of the probe alongside the meter.
Another related problem occurs at these low frequencies if you are holding the meter. Your body can act as an antenna or a ground when you hold the survey instrument. The best approach is to set the instrument down on a nonconductive stand of some type—even a cardboard box will work. Stand back and read the meter. Alternately, attach an insulated handle to the meter. Hold the meter away from your body by holding the insulated handle.
Probes that have diode sensors are generally not suitable for measuring radar pulses. They can be used, but only when the field levels are very low. And since it is difficult to know the transition point for a particular instrument, it is safer to always use a thermocouple probe to measure radar pulses. The diodes tend to peak detect, and you can get readings that are 10 to 100 times higher than the actual RMS field level.
Diodes have a square-law detection characteristic at low levels and a linear characteristic at high levels. This is because the video resistance of the diode changes with the level of RF current through the diode. The resistance decreases with the amplitude of the pulse, which increases the efficiency or sensitivity of the diode as a rectifier. The low resistance allows rapid charging of the circuit capacitance. The diode becomes a peak detector rather than providing an integrated, averaged DC output.
Probes can be oversensitive to signals at frequencies above their rated frequency range. Unfortunately, this parameter is rarely specified. In extreme cases, a probe can be 5 to 10 times more sensitive to the higher frequency signals than they are within their rated band. The result is that a weak higher frequency signal, where the exposure limits are likely to be higher, may result in an indication of a problem that does not exist.
Diode probes can be operated out of the “square law” region of the diode and result in “linear detection.” When this happens, the survey equipment indicates field strengths higher than it should. The worst case is when the signals from multiple sources are of a similar magnitude. Under these conditions, it is possible that the field strength could be overestimated by as much as 10 dB (10:1). Read specifications carefully, and ask questions before you invest in survey equipment.
Rapid movement of a probe and/or the presence of a flag or an article of clothing that can hold a charge, such as a nylon windbreaker, can result in very high, momentary false readings.