 |
|
|
EMR MPE standards compliance evaluation tool usage notes
Step by stepThere are many ways to go about calculating the applicable safe distance for a particular scenario, here is one:
Note that the results will depend on all of the parameters to some greater or lesser extent. The safe distance for an SSB voice transmitter might be quite different to the same transmitter running Morse, so you may need to model each mode, and each antenna that you use on each band. To help with Step 1, you may find my VSWR calculator helpful. Input values / formatsLimits on range of valid data are shown adjacent to the input field label, ** means a large number. The input fields may support flexible input format. In general, the formats supported include traditional floating point number (50.00), scientific notation (5.05E1), in combination with qualifiers decibels (-3dB), percentage (50%), per unit (0.5pu). On the Relative Field Strength field, 50 == 5.0E1 == -3dB == 0.5pu. Similarly, you could enter 14dBm (-16dBW or 25 mW) in the PEP field as -16dB. Not all fields support all formats. Model VariablesAntenna GainAntenna gain is the gain of the antenna in the direction of interest (usually the main lobe of the antenna pattern) above an isotropic radiator. Gain is often expressed with reference to a dipole, in which case 2.14dB needs to be added to obtain the "isotropic" gain. Aperture DiagonalAperture diagonal is the lineal measure of the diagonal of the antenna aperture in metres. It is used for modelling the near field effect, which is that power flux decreases less rapidly with distance than the inverse square law dictates for the far field. CACA is a flag for circular aperture (instead of a rectangular aperture). The choice of this factor affects the modelling of near field effects. It seems to me that it is appropriate to use for the main lobe of a parabolic dish, but probably not appropriate for the side lobes of such a dish. CommentFor documentation only, not used in the calculations. Extra MarginThis is normally 0dB, it provides for increasing the margin, for instance modelling a site with multiple transmitters. The dB figure is used to adjust the average EIRP calculated from the other arguments. FrequencyThe frequency in MHz. The model implements the reference technique for frequencies from 1MHz to 300GHz. Ground reflectionGround Reflection is a flag which causes the average EIRP to be increased by 4dB to model the effect of ground reflection on the field strength, in line with the OET recommendations for antennas mounted near the ground. Modulation FactorThe Modulation Factor is a factor for calculating the average power of a modulated signal given the PEP. Modulation TypeThe Modulation Type control allows easy selection of one of several predetermined Modulation Factors from a list of Modulation Types. Both figures for SSB were experimentally determined, and are used to build the AM figures. The Morse figure is from the OET reference. MPE StandardThis menu allows selection from a list of standards for Maximum Permissible Exposure. Implementations include:
The ICNIRP and ARPANSA implementations include the frequency dependent time average interval and frequency dependent limitation on the peak to average ratio. The ARPANSA draft seems to suggest that it is not safe to assume plane wave conditions below 10MHz and possibly in any near field situation, suggesting independent evaluation / measurement of E and H intensity is required below 10MHz. I am trying to clarify this and some other aspects of the ARPANSA draft (update - trying to obtain clarification from ARPANSA during via their public discussion venue proved to be a waste of time). PEP InputPEP Input is the Peak Envelope Power absorbed by the antenna. This is less than delivered by the transmitter and is probably less than the transmitter rated output. The practical PEP output of a transmitter may depend on the modulation type, a transmitter that delivers 100W PEP on SSB voice may have a practically maximum available power of 50W PEP on FSK (RTTY) for heat dissipation reasons, so use the PEP value that is appropriate to the mode that you are modelling. To avoid accumulating unnecessary conservative margin in the model, use the VSWR calculator to find out more accurately, the net power absorbed by the antenna. Key issue is to make sure that you have made allowance for nominal cable loss, additional cable loss due to VSWR, and reduced output from the transmitter due to mismatch. Measuring the PEP for some modulation types is a bit tricky, since you need a peak detecting wattmeter or an oscilloscope that is calibrated at the frequency of interest. If you can measure a CW output, then you can estimate the PEP from that for most scenarios.
Relative Field StrengthAllows calculation of the safe distance in different directions for the antenna. Use 1 for the main lobe, and a per unit relative factor for other directions that you wish to model. This figure is used to adjust the average EIRP in the modelled scenario. Note that power flux density is proportional to the square of this value. Safe DistanceThis is the minimum distance that a person should maintain in the modelled scenario to ensure that flux density from the modelled source does not exceed the standard. TitleTitle for documentation purposes only, appears as a sub heading on the results form. Transmitter OnTransmitter Off is the period in seconds of transmitter off time during periodic on-off-on-off... operation. The model will calculate the worst case ratio of transmitter on time to transmitter off time in a rolling window of the period specified the the selected MPE Standard and adjust the EIRP Transmitter OffTransmitter Off is the period in seconds of transmitter off time during periodic on-off-on-off... operation. The model will calculate the worst case ratio of transmitter on time to transmitter off time in a rolling window of the period specified the the selected MPE Standard and adjust the EIRP Modulation factorsWhen first reading the ACA's draft reference and the supplementary tables for evaluation of Amateur Stations, I noted the lack of consideration in the models for the effect of modulation and transmitter duty cycle that are part of the real operating scenario for many communications transmitters. (This was amended in a later draft, but the modulation factors were the same as OET's, so the following discussion is relevant.) Since the standard relates to the power absorbed averaged over a sliding window of 6 minutes, we are interested in estimating the average power (over six minutes) to something more readily measured or rated on a transmitter such as the PEP. This ratio is of interest in carrier telephone systems, where measurements have lead to practice that designs capacity around a ratio of average talker power to peak of -12dB to -13dB (a supergroup amplifier or a microwave mod/demod does not need to, and is not designed to carry all channels at 0dBmO - indeed it will cause serious problems). So, I did some measurements. For the test setup, I connected a Kenwood TS-440 transceiver to a Motorola R2009D service monitor, and connected the demod output to a PC sound card and recorded the demodulated audio as a WAV file. I then analysed the WAV file to find the the long term RMS audio level relative to the peak RMS audio level (being assumed to be 3dB below the peak wave sample) - these would correspond to the average "heating" power and the PEP power of the radio signal. I ran a 30 second test for the transceiver optimally configured (ALC action, but not exceeding the ALC range) without the speech processor activated. The ratio for 30 seconds was -15dB, and for word duration periods was around -13dB. I ran a 30 second test for the transceiver optimally configured (ALC action, but not exceeding the ALC range) with moderate speech processing (near onset of audible distortion). The ratio for 30 seconds was again -15dB, and for word periods was around -11dB. Although this is a very limited trial, it is quite consistent with the models used for carrier telephony design and suggests that the OET Supplement B figures might be conservative to the tune of at around 5dB, possibly 8dB in the case of speech processing for short time averaging (300mS), and 10dB to 12dB for averaging over 30S. I have used (conservatively) 13dB for voice modulation without speech processing and 11dB for voice modulation with speech processing pending further measurement. The 40% RMS/PEP factor for Morse suggested by OET seems to be reasonable. I have obtained figures experimentally that are close to that number. More recently, I calculated the factor using a table taken from a cryptanalysis text of frequency of occurrence of the alphabet in plain English text, and the Morse Code encoding of those characters, and obtained a figure of 46% (rounded), so I have used 46% as the modulation factor for Morse. (There is quite a deal of scope for variation of this figure considering factors like hand key generated morse, 'thinking' breaks, digits, punctuation etc.) Relativity / a balanced viewI should go and have a cup of coffee, but I just thought it through. I would ingest 250 mL of liquid which is 20 deg above my body temperature, so it contains 20 * 250 or 5KJ of energy. I will take a few minutes to drink it, so the average temporal power is 5000/180 or 28W, and I weigh 90Kg, so that is 0.3W/Kg whole of body energy absorption rate, got to be at least ten times that in some parts. Doesn't sound too healthy, I think I better get a cold beer instead. You will note a similarity between the methods and modelling techniques proposed between Australia and North America, however the MPE levels vary dramatically. Watch what the new ARPANSA standard proposes. Perhaps it indicates something of the science behind it all! Assumptions
Radiation is only from the antennaFeedline radiation or radiation from the equipment is not significant. If the configuration uses a feedline with significant common mode current (eg current on the outside of the outer of coax, or unbalanced open wire, or single wire) either driven or induced (eg asymmetry of feedline routing from the antenna, unbalanced drive, unbalanced / asymmetric load, lack of measures to prevent common mode current) then this modelling technique will not be appropriate. It might be simpler to address the factors in your configuration that render the model inappropriate than to produce a more complex model or measure the field strengths. A commonly held misbelief is that open wire feedlines must radiate and that coax feedlines cannot radiate - both are untrue. It is true that single wire feedlines (eg a Windom antenna) must radiate. Another commonly held misbelief is that feedlines cannot radiate when VSWR is low, and that feedlines must radiate when VSWR is high - both are untrue. References
ImplementationThe calculator is written in a combination of HTML, PHP, and Javascript, using Microsoft Frontpage as the development tool. PHP is a server side scripting tool (or Dynamic HTML). Most of the smarts are in PHP, and you won't see them directly. Javascript is used to a very limited extent because of the lack of support for recent Javascript features in IE, especially dynamic objects and string objects. It is mainly used for initialisation of the form and linking transactions in a session-less pseudo-conversational style. The pages are served from an Apache server with PHP, running under Linux on an Intel x86 box. The pages have been tested under IE5 and Netscape 4.7x. DisclaimerUse at your own risk, not warranted for any purpose. Do not depend on any results without independent verification. Last update: 01 April 2009 15:50 |
|
VK1OD on the 'net. I appreciate your feedback. © Copyright: Owen Duffy 1995, 2006. All rights reserved. Disclaimer. |