Sound Thinking
Since I wrote this article Tandy UK have folded. However the Realistic Meters may be still available through Radio Shack dealers in the UK. Alternatively Maplin now do a digital meter for £29.95 (price as of 12.12.2005) and I think I saw the old Realistic Meters in my local Maplin shop.
SOUND THINKING
Or why a £25 sound level meter is as good as a £400 one when it comes to keeping your site.
We all know the importance of keeping model aircraft noise down to acceptable levels if we are to keep our precious flying sites. It is in our own self interest to ensure that we do so, apart from the fact that only the truly anti-social would knowingly create a nuisance for other people.
In order to ensure that models meet the DoE Code of Practice and do not produce more than 82 dBA at 7 metres it is necessary to use a Noise Meter. However equipment that compiles with the British Standard BS5969:1980 as described in the Code of Practice is very expensive, at circa £350 for a ‘basic’ meter, and up to £600 if you get a calibrator and accessories to go with it.
This is a big expense, and it is my contention that it is an unnecessary expense. Let me explain my reasoning |
|
Expensive meters
|
The main reason meters to BS5969, such as the CEL-231 (see illustration) and others, are so expensive is that they are accurately calibrated and give precise readings. They often have other features such as overload indication, max.hold function, impulse and peak detectors, outputs to recorders etc.. All very nice but not relevant to our use. Also in order to verify and maintain these high levels of accuracy over a period of time they should be regularly calibrated. This service can be provided by a third part at about £60 per calibration, or a calibrator can be purchased at circa £180. All this can make the purchase and owning of a noise meter a major expense. |
|
)
If the DoE Code of Practice had the force of law, and the survival of a club when a noise dispute arose depended upon proving with scientific indisputability that the models achieved exactly 82.0 dBA at exactly 7.00 metres there would be some merit in spending large sums on accurate and precise meters.
In fact even the more expensive meters general only have a quoted accuracy of ‘±1 dB under reference conditions’. Also ‘reference conditions’ normally correspond to measurements taking in an anechoic wind free chamber or similar, far removed from the field measurements we require. The fact that even expensive instruments can only achieve ±1% (a quite poor figure compared to other areas of electronic measurement) serves to illustrate how hard it is to pin down noise measurements. For further discussion on the difference between accuracy and precision see bottom of page.
Unnecessary precision
This simply illustrates that in practice worrying about the precise value of the noise reading is unnecessary. While limits have to be set and adhered to, the exact conditions with regard to wind, humidity, ground surface, etc. will all effect the actual reading. A model that passes at 81.5 dB today could well fail at 82.5 dB tomorrow.
In addition to the technical issues the Code of Practice is only a guidline, and if there are objections, the local authorities will take a view on what noise controls have been implemented, they are not bound by the precise levels. The Code of Practice is not the law and there are no direct legally enforceable noise requirements for model aircraft.
This is not to say that the Code of Practice is irrelevant, far from it. In the case of a dispute or planning enquiry the Code of Practice is the document the authorities will refer to. As the BMFA say in their Members Handbook:- "If there is a noise complaint against your flying site, the Local Authority will probably send an Environmental Health Inspector to investigate. He will arrive armed with his noise meter and a copy of the DoE Code of Practice."
Once this happens your expensive noise meter is about as much use to your club as a chocolate teapot. The Environmental Health Inspector will use his own, probably very expensive, BS5969 meter to take measurements. He may take into account your club’s methods for checking noise, and your accurate and meticulous records of model testing and your documented methods of calibration, etc. Or he may not. My limited experience in small clubs without any financial muscle is that you will be warned off as soon as a noise compliant arises.
Why bother?
So why bother with a noise meter at all? Because they are a very useful and objective method of measuring your efforts at reducing noise. Because they will show that the club is serious about reducing its impact on its neighbours and because it will enable none co-operative club members to be ‘weeded out’ without any subjective arguments. In fact I think a noise meter is an essential piece of equipment for a club, and even a serious individual, to own.
However the use of low cost but less accurate meters (used with sensible precautions) is just as valid as the more expensive units.
Setting limits
|
For example the Realistic Analogue Sound Level meter available from Tandy at £24.99 (yes £24.99!) is an excellent choice. It has a clear analogue meter with a resolution of better than 0.1 dB. It covers the range 60-120 dB in 10dB steps, it has A & C weighting and Slow and Fast response options. In short it has every thing you need to carry out noise measurements. The price even includes a smart and practical carrying case. (for an explaination of Decibels and weighting Scales see end of article) |
|
)
It is quoted by the manufacture has having a ±2 dB accuracy. As a responsible manufacture this can be taken with confidence, and most of the product they sell will be better than this. If we want to ensure that our models just squeak under the 82dB barrier, then we set our pass level using this meter at 80dB. Then even if the meter is reading 2 dB low, we know that the true reading cannot exceed 82 dB.
It seems that many clubs actually set a pass level less than 82 dB, and very sensible they are to do so. Even at 82 dB model noise can still be annoying, and we should all be striving to reduce the levels of annoyance. Again in this instance the use of an expensive meter to enforce an ‘arbitrary’ limit does not make good sense.
Build your own calibrator
If you are worried that your low cost meter is going to drift with time then you can construct your own reference noise for a few pounds. While this will not be a calibration source in the ‘scientific’ sense it will give you confidence that all is OK with your meter.
You can easily buy small piezoelectric buzzers that run directly of a 9V battery and will typically generate 85dB at a distance of 10cm. You mount one of these in a cardboard or plastic tube that matches the diameter of the microphone on your sound level meter and run it from a PP3 battery.
Experiment with the distance between the buzzer and the microphone until you get a reading around 80dB. Fix the distance and make a note of the reading. Then if you suspect that your meter has changed (perhaps you’ve dropped it or it hasn’t been checked for several months) it can be tested by applying your calibrator and ensuring you get the same reading as the day you made it.
Just make sure you have a fresh battery or check the voltage before you start.
If you don’t think this works all I can say is that I once purchased an expensive Sound Meter, similar to the CEL-231, as a club unit. (This was before I saw the light!). As a precaution I constructed a calibrator as described. Over a period of two years I never recorded a change greater than 0.1dB between readings.
No excuses
As I said at the beginning of this article I am 100% in favour of ensuring models are quiet enough to be acceptable to our neighbours. Our club has a noise rule and we would not accept anybody who does not appreciate the need for noise reduction.
However it is not necessary to spend large sums of money on equipment. At £25 a meter no club has an excuse for not having a meter and not using it to test regularly.
******
The orginal article finished with examples of meters that were avaiable for purchase. As this informataion is now well out of date and most if not all the meters are obsolete I have removed it.
Accuracy and Precision
These two terms are often confused and used interchangeably. They have distinct and separate meanings and an instrument can be precise but inaccurate and visa versa.
To illustrate the difference:
- If you to say that a metre is 39 inches you are accurate but imprecise.
- If you say that a metre is 39.4592 inches you are precise but inaccurate.
- If you say a metre is 39.3701 inches you are both accurate and precise.
Digital meters give the impression of being accurate because they are precise, ie they display the value of the parameter they are measuring to a large number of digits. It does not mean that they are accurate to that number of digits. If you purchase a digital meter with a 3½ digit display you could quote the result as 82.0 dBA. However the actual true value could lie anywhere between 81.18 dBA (a pass) and 82.82 dBA (a failure).
Decibels and Weighting ScalesSound pressure levels (or loudness) are measured in decibels (dB) with the threshold of hearing or 0 dB assigned a value of 2 x 10-4 µbar. The scale is logarithmic so that each 3 dB increase corresponds to a doubling of sound pressure levels. |
click to increase size |
)
Sound pressure becomes a ‘noise’ or ‘loudness’ measurement when the frequency response of the human ear is taken into account. The human ear is more sensitive at certain frequencies, most notably those employed for speech. Several approximations to the frequency response of the ear have been developed, one of the most widely used is the DIN 45633 A wieghting curve (see diagram).
Hence when a sound is quoted as being 82 dBA this says that the total pressure level between 10Hz and 2kHz was 82 dB and the frequency weighting was in accordance with the A scale of DIN 45633. In this context A does not stand for acoustic as I once read in an ‘authoritative’ article about model noise limits!