A free electronic newsletter covering news and other topics for those interested in RF safety issues.

LIVE, Web-Based RF Safety Training

Understanding Specifications—An Overview

Sometimes you have to “dig deep” to figure out what some probe specifications mean or what isn’t specified.  Here are some important points to consider when looking at probe specifications.

Frequency Response

Realistically, specifications can only be guaranteed if calibration is performed at multiple frequencies.  And since one’s definition of frequency range is tied to the limits set for frequency response, this is a very important specification.   There are probes on the market with very broad frequency ranges.  But, if you only consider frequency range without looking at limits for frequency response, you may fail to understand the magnitude of the measurement uncertainty.

Consider the specifications for two brands of broadband, shaped electric field probes that are available.   

 Probe B has a reasonable frequency response at frequencies up to about 4 GHz and is of little value above that.  Even at the lower frequencies, probe B is nowhere near as accurate as probe A although it can yield useful results.  The Latin term to consider is caveat emptor (let the buyer beware).  

Dynamic Range

While probes that use thermocouple sensors always yield true RMS (Root Mean Squared) results, probes that use diode detectors can only operate in “square law” over a limited dynamic range.  A probe that is described as having a 50 to 60 dB dynamic range may be able to measure signals over this broad range but the compromise is multi-signal accuracy.  For example, probes are sold for use in the wireless communications industry with a very broad dynamic range.  This range is attractive because it appears that you will be able to both measure very weak fields far from the antennas and significant level fields close to the many antennas typical of these sites.  But, if the levels are significant, the probe may be “out of square law”.  When a probe operates outside the square law region and provides “linear” detection, the instrument overestimates the actual field strength.  This problem is most acute when there are many signals of a similar magnitude, such as at a wireless communications site.  At first glance a overestimation may sound like it is not a problem—it is just being conservative.  If you overestimate the field strength, it may result in unnecessary and costly restrictions and controls being put into place to achieve compliance with RF safety regulations.  It is not uncommon to have errors of ten to one (10 dB).

Peak Overload

Peak overload is concern when measuring pulsed signals, such as from a radar.  Only thermocouple sensors should be used to measure radar pulses—diodes can peak detect and greatly overestimate the field strength.  A thermocouple is always a true RMS detector under all conditions.  The major concern with thermocouples is burn out which can occur from exceeding their limits under either average or peak conditions. 

Peak overload concerns are greatest with very low duty cycle conditions.   For example, a typical search radar has a duty factor of 0.001.  This means the radar pulses are on 0.1% of the time and off 99.9% of the time.  If a probe has a full scale rating of 20 mW/cm², then it must have an absolute minimum peak overload rating of 20 W/cm² at this duty factor.  It really needs to be about ten times higher otherwise you run the risk of burning out the probe from a peak overload just as the meter hits full scale. Check specifications carefully.  There are probes available that burn out from peak overload with a typical radar while the meter indication is far from full scale.

RMS Detection

A peculiar characteristic of simple diode detectors used in isotropic probes is that they can become “linear”.  This means that they stop functioning as an RMS (Root Mean Squared) detector.  This behavior is shown in the diagram below. 

Some equipment manufacturers use “squaring circuits” in the meter to compensate for the diode operating in the linear region.  This design approach can greatly overestimate field strength in multi-signal environments.  The greater the number of emitters, the larger the error.  The error is greatest when the magnitudes of the signals are similar.  The error is much smaller if one signal is far stronger than all the others.   

A sensor that is operating in the square law calculates the total energy in terms of the square root of the “sum of the squares”.  A linear detector calculates based on the square root of the “square of the sums”.   For example, if there three signals (A, B, & C) present and they were all of the field strength at the point of interest (1) the two formulas and the results would be:

Therefore, linear detection would yield results that overestimate by 73% (3/1.732).  If there were ten signals, as is often the case at a wireless site, the results would be more than three times higher than the actual field strength (10/√10). 

One manufacturer’s “compensated diode” design results in diode-detector probes that give good RMS detection over about 30 dB of dynamic range, which is far better than most diode probes.

Out-of-Band Response

It is rare for a manufacturer to specify out-of-band response for a probe but it can be very important in some applications.  Ideally a probe is less sensitive out-of-band than within its rated frequency range but this is not always the case.  Consider two common problems: 

1. Most magnetic H) field probes tend to have resonance pick up at frequencies far above their rated frequency range.  Suppose that you are measuring a VHF signal where the exposure limits in most standards is 1 mW/cm².  At co-located sites, there may be a UHF signal.  And the exposure limit at the UHF frequency would be two to three times higher (2-3 mW/cm²) depending on the standard and the frequency.  If this UHF signal happens to occur at one of the H-field probes resonance frequencies, the probe may be eight to ten times more sensitive to this out-of-band signal than it is within its rated frequency band.  So a UHF signal of 0.1 mW/cm² that is less than 5% of the exposure limit might indicate 1.0 mW/cm² on the meter.  You would interpret this reading as 100% of the limit without knowing it was not a VHF signal!

2. Most diode-based electric (E) field probes tend to be oversensitive above their rated frequency range.  One manufacturer manufactures its probes with special diode that feature a sharp roll-off in sensitivity at frequencies just above the rated frequency range of the probe.

Temperature

One specification that you will seldom see is temperature deviation.  All sensors have an inherent change in output as a function of temperature.  Instrument manufacturers put in temperature correction circuitry to minimize the amount of error resulting from use at different temperatures.  But the circuitry can only reduce the error, not eliminate it.  

The typical temperature deviation (including the affect of correction circuits) for well-designed probes is: 

[ Home ] [ Up ]

Send mail to web@rfsafetysolutions.com with questions or comments about this web site.
Copyright © 2007 RF Safety Solutions LLC
Site last modified: 2/28/2007