Model 70 Series Application Overview
Application Details, FAQ’s and Special features
Sonomètres Pulsar 
Lärmmessgeräte Pulsar Model 70 Series Application Overview
The 70 series are essentially series 50 with Leq and Peak added, plus octave filtration on two models, so all the features of the 50 series are maintained, including the ease of use and the excellent performance.
All four models in the 70 series have Leq, True Peak, ‘S’, ‘F’ & ‘I’ Time-weightings as well as ‘A’ & ‘C’ Frequency-weightings plus backlight and tripod mount.
The table below shows the differing features of the four units and it clear that model 74 a Type 2 unit with octave analysis built in represents remarkable value, encompassing as it does the H & S requirements of all EU countries and most of those outside Europe, except of course the USA where OSHA still uses their own obsolete rules (5dB doubling instead of true Leq).
| Model | Accuracy Class (IEC 60651/60804) |
Filters |
| 71 | Type 1 | No filters |
| 72 | Type 2 | No filters |
| 73 | Type 1 | Octave filters |
| 74 | Type 2 | Octave filters |
One of the main advantages of the 70 series over computer-based units is simply “ease of use”. The controls are clearly labelled, the function names are the same as used in everyday speech and the ‘learning time’ needed to operated the 70 series is truly minimal. Once used, it is rare for even an untrained user to need the manual to make the legally required measurements.
Usage
The manually operated 70 series does EVEYTHING needed for industrial safety measurements for hearing damage risk assessment; where computer storage is not needed. Most people expected that by 2004 computer based data logging meters, such as the 60 series and 30, would become the ‘only way’ – they have not! The reason appears to be the simplicity and ease of use of the 70 series and “computer phobia” among some users.
The advantage of octave filters is that it allows the exact frequency spectrum of the noise to be checked against various manufacturers published EAR PROTECTOR data.
Many users in industry use their meter perhaps four times a year and do not want to have to re-learn it every time, as they believe they would need to with complex computer based units. For these users the 70 series is ideal and thus most 70 series units are used by Safety Officers in industry, all round the world. Safety Officers are usually not noise specialists; they just want something easy to use, low cost, accurate and rugged – in other words, the 70 series, or form 2004, the Assessor.
Application Details and FAQ’s
What are ‘Peak’ and ‘Max’
True peak is the maximum value of the un-rectified acoustic signal and is a required metric for many hearing damage loss legislative requirements such as the EU Directive or the UK Health and Safety Executive regulations. Peak is always implemented as ‘PEAK HOLD’, where the peak level is held until cancelled either automatically or more usually manually. PEAK HOLD cannot ever be the same as MAX HOLD. The FORMAL definition of Peak Level is: twenty times the logarithm of the greatest absolute instantaneous sound pressure during a stated time interval. Peak can be expressed in true pressure units, i.e Pascal or in decibels when it is called Peak Level. To give an idea of scale, 140dB peak level is identical to a peak pressure of 200Pa.
Max is an exponential or rms metric, with an averaging time set by one of the three ‘S’, ‘F’, or ‘I’ time-weightings and is simply the maximum reading that the instrument gives using these time weightings. The formal definition is: the greatest time-weighted level within a stated time interval.
For a sine wave, such as is given by an acoustic calibrator, reading 94 dB on ‘S’, ‘F’ and ‘I’ should read 97 dB on PEAK. This is because the peak value of a sine wave is 1,414 times the root mean square value and 1.414 times is 2v2 or in decibels it is 3dB. Naturally, when the signal is NOT a sine wave, the PEAK value will usually be more than 3 dB higher and for a single impact noise such as a single gun shot the peak value can easily be 40 dB or 10,000 times higher than the rms max value.
The peak value of the C-frequency-weighted signal – in symbols LCpk – is a required metric in the European Union for hearing damage risk and MAX – even the A-frequency-weighted I-time weighted data – in symbols LAImax - should never be used in it’s place. In fact unless there is some legal reason, such as in obsolete German regulations, the I-time-weighting should never be used as it has a VERY poor correlation with the ’impulsivity’ of a sound. This is because it has a fast (35ms) rise time and a very slow fall time (2,9dB/s) thus it ‘ratchets’ up when it has several impulses to measure. I-time-weighting is not in the body of the new standard IEC 61672, but German representatives on the IEC working group insisted it be added as an annex as “Our laws demand it and they are difficult to change”; comment is superfluous!
As a historical note, when the Impulse weighting was being determined, the International Working group could not agree on the ‘up’ Time constant for Impulse. Some wanted 50mS and some 25mS. Two of the most senior members met in a ‘smoke filled room’ and came out with the announcement that it was to be 35mS. We have had to live with this ‘fudge’ for a quarter century.
Octave analysis
Models 73 (Precision type 1) and 74 (General Purpose Type 2) have octave frequency filters incorporated. There are many uses for this, but for hearing damage risk assessment, the most common is ‘classifying’ the noise to select a suitable ear defender for the particular noise.
What is an octave?
In acoustical terms, it is simply a band of noise with 2:1 frequency
limits. Thus, the primary band is centred on 1kHz – the reference
frequency – and this 1 kHz band passes all frequencies between 707Hz
and 1,414 kHz, that is everything between v2 kHz and 2v2 kHz; giving the
2:1 frequency ratio. In music, the frequency ratio of an octave is also
2:1, but as there are 8 notes between the two limits, this ratio was called
an octave.
Above this reference frequency of 1kHz, the band centres are 2kHz, 4kHz,
8kHz and 16kHz, while below the 1khz reference they are 500Hz, 250Hz,
125Hz, 63Hz and 31,5Hz. These 10 octaves encompass all possible needs
for hearing damage risk assessment; indeed the 16 kHz band is not strictly
needed, as it is above the normal IEC 60651 Type 2 frequency limits of
31,5Hz to 8 kHz, but many users demand it and it is useful for some precision
work. The ‘top’ of the 16 kHz band is obviously 16 X 2v2,
about 22,6 kHz and this is often outside the maximum frequency of the
“Type 1” microphone used by some manufacturers. This is so
in the model 74 Type 2 unit where the type 2 microphone has a maximum
response of about 16kHz. Thus the 16kHz band on model 74 should not be
used for accurate measurements.
In theory all frequencies between the upper and lower band limits are passed with equal sensitivity, but in reality there are ‘ripples’ in the response and the theoretical vertical sides of the response cannot be met in a practical realisation and so the IEC standard 60642 defines the limits permitted for each octave.
Note: Although 1,414 is normally used as the value of 2v2, when calculated to 30 decimal places of resolution, as every maths student knows, it is actually 1.414213562373095048801688724210. Such a resolution is pointless in acoustics, where 0,1 decibel or about 1% is the usual best resolution needed, but to save angry mail, precision in all things!
Serial and ‘real time’ or parallel octaves
The traditional octave analyser such as model 74 measures each octave independently, one at a time. Thus to measure a single noise source, up to ten measurements need to be made. Any other method, such as making all the ten measurements together is beyond the scope of a manual device as effectively ten separate circuits have to be incorporated. Today, now that computers have become an essential part of acoustics, all the ten octaves can be measured in parallel, or as it is more usually quoted “in real time”. Obviously Pulsar Instruments Plc. also have such a REAL Time Analyser unit – normally called an RTA – in models 30 and 30-2, qv.
Many users now decry the use of serial octaves, but since about 1960, such serial analysers have formed the backbone of octave measurements and have provided almost all the data we have available today. If the user is a ‘computerphobe’ these traditional methods are still very useful and in daily use and for ear defender work they are perfectly valuable. Indeed, some modern computer based units still use serial octaves for ease of use.
The obvious problem with serial octaves is that the noise spectrum may change between each octave measurement and of course this is true. However, the intent for Health and Safety is to measure in Leq, an averaging metric and so if a complete machine cycle is measured on each octave, the measurement is totally valid; it simply takes a little longer.
Special features and applications
Ear defender selection
The method of doing this is to measure the ‘noisy’ machinery. Usually, a machine cycle is fairly short and if a complete cycle is measured, using Leq on each octave filter in turn, a simple plot can be drawn on squared paper of the SPECTRUM of the noise, or the actual data can be used in tabular form. This data is then compared with the ear defender data and if the two spectra more or less match, the resulting level in any band AFTER the ear defender is worn can be used as the resulting level in that band. For Pulsar users, an even simpler way is to request the optional software that does this calculation for you if you are happy using conventional MS-Windows software.
Single octave Leq
Because the 70 series are manually selected serial filters, these units are very useful to measure the level of a single frequency inside an octave. This application is not uncommon where a production process error occurs and the resultant change of noise can be picked up only at one or a small band of frequencies. Obvious applications are gear noise where the overall level is high masking the error signal, but an octave filter may well pick up the error. In fact any system where a signal is masked by other noise may be made useful by removing the masking signal and effectively lowering the noise floor.