Fullscreen HTML5Med Res (??MB)HTML4

Measure sound and vibration way below human hearing by Allan Linton-Smith and Ross Tester Infrasound Detector Photo: Harvey McDaniel     Wikipedia Are wind turbines making you sick? Is building vibration making you nauseous? Or do you just want to measure infrasound in your environment? You don’t need to spend thousands of dollars to do it properly; just build our low-cost but accurate Infrasound Detector. T here’s been a lot of press lately about infrasound, particularly as it applies to wind turbines. But until now, you’ve needed tens of thousands of dollars worth of test equipment to detect and measure it. Our Infrasound Detector can be built for less than a hundred dollars yet will give very accurate results. You can either read the sound pressure directly or store and analyse readings on your computer! So what exactly is infrasound? It can defined as sound below the range of normal human hearing. That’s 30  Silicon Chip generally reckoned to be below about 20Hz. Below that, you can perhaps sense or even “feel” sound but you can’t actually hear it. In practice, infrasound involves frequencies from about 20Hz to 0.5Hz but some natural phenomena can cause infrasound down to the millihertz (.0001Hz!) region. When people complain about illeffects from infrasound (and there are legions of those reports), many acoustic consultants have taken the attitude that “if you can’t hear it, it can’t be doing you harm”. We disagree – and the publisher of this magazine even wrote an editorial on the subject back in February 2010. Reported human reaction to infrasound from wind turbines is varied but some of the reports associate infrasound with a general feeling of malaise, nausea, vertigo, blurred vision, memory problems, tinnitus, anxiety, uneasiness, extreme sorrow, nervous feelings of revulsion or fear, chills down the spine and feelings of pressure on the chest. Others have reported headaches and migraines, major sleep disorders and even self-harm tendencies. siliconchip.com.au Some researchers have even given it a label: wind turbine syndrome. Wind turbines are one example but you’ll also find infrasound caused by traffic noise, heavy surf, engines/motors (especially things like compressors), building vibrations being excited by wind, machinery and so on. Large animals such as whales, crocodiles, alligators, elephants and emus communicate with infrasound so if you want to record amorous crocodiles, our Infrasound Detector is a good way to go about it (from a safe location!). Other source of infrasonics are heavy artillery, the calving of icebergs from glaciers and earthquakes. In fact, there is a theory that the buildup of stresses with the earth’s crust before a major earthquake causes infrasound – which could explain why birds and some other animals appear to have some warning of an imminent quake. Want another example? The very act of opening or closing a door produces infrasound waves. But that is transitory – you don’t normally stand there for hours opening and closing doors! Whatever the infrasonic phenomenon you want to investigate, our Infrasound Detector is an effective and low-cost way to do it and it compares more than favourably with commercially available equipment. While it’s economic, it’s also acFig.1: the testing unit is based on a modified PreCHAMP preamplifier which detects sound via the electret microphone, then removes all but signals below 20Hz. This signal can then be analysed by a computer running “Fatpigdog” software, or it can be fed to a modified CHAMP amplifier which drives a multimeter in its AC range to deliver readings of sound pressure levels. 'PRE-CHAMP’ PREAMP (MODIFIED) *10k RESISTOR ADDED TO POWER ELECTRET MIC 22k 100k 10k* Q1 BC548 SHIELDED LEAD B 1000F VR1 100k C 8.2k 150k 1uF MKT 470F 16V TO PC SOUND CARD 1k VR2 10k *SEE TEXT SC siliconchip.com.au 220F 16V 4700F 8 16V 1 IC1 LM386N (SEE TEXT) 2 5 10F 10V 12V LED 100nF A 68 2.2k INFRASOUND TESTING UNIT 9V BATTERY 7 4  K 10 LED COMPONENTS IN RED ARE CHANGED/ADDED 2011 6 3 120pF 100 470F 16V POWER 100F 16V B S2* 39k S1 ANALYSER GAIN 100F Q2 BC558 E 16V C E ELECTRET MICROPHONE 0.778 'CHAMP' AMPLIFIER (MODIFIED) 2.2k 10F 16V JAYCAR QM1327 MULTIMETER ON FREQUENCY RANGE K A BC548, BC558 ELECTRET B OUTPUT EARTH E C March 2013  31 . curate and reliable – we believe it can be just as accurate and reliable as commercial gear. In fact, while our unit should cost well under $100 to build and is easy to put together, it took hundreds of hours to develop and test. That is because infrasound sweeps can take hours to settle, measure and average – and some very specialised and expensive equipment was required to design and test it. If you wanted to buy that commercial equipment yourself, you’d have little change from $30,000! We also had to develop a method for testing and calibrating high levels of infrasound without upsetting the neighbours! How it works The output from a wide-range electret microphone is fed to a verylow-frequency bandpass filter. The infrasound signal is amplified and fed to a “virtual” spectrum analyser which then plots the amplitude of the infrasound signal on the vertical (Y) scale versus frequency on the horizontal (x) scale using a principal known as Fast Fourier Transform (FFT). A computer can then be used to analyse the signal and/or a direct frequency readout can be obtained if used in the field. Our Infrasound Detector is built into a small diecast box, with an old microphone shield attached to the front. Inside this shield is a low-cost electret mic insert. The terminals at left are the output to a frequency counter (or in our case, a budget multimeter) while a socket is provided on the right side for output to a PC sound card. Suitable analysis software is quite cheap. Specifications Microphone frequency response G-weighted:........... ±2.0dB corrected (0.5Hz-26Hz) Microphone frequency response C-weighted:........... ±2.0dB (10Hz-20kHz) Microphone intermodulation distortion:......................... 0.8% <at> 100dB SPL Preamplifier frequency response: ....................................... ±0.2dB (0.5Hz-20kHz) Power amplifier frequency response: .............................. ±0.2dB (0.5Hz-20kHz) Power amplifier output (before clipping): .................... 200mW into 8Ω Frequency response of virtual instrument: ................. ±0.4dB (0.5Hz-20kHz) Overall measuring accuracy – Without calibration table:................................................. ±15dB (20Hz-20kHz) Using calibration table: ..................................................... ±1.0dB (2Hz-20kHz) THD+N preamplifier: ........................................................................ 0.102% at 1kHz (5Hz-22kHz) THD+N power amplifier:................................................................. 0.40% at 1kHz (5Hz-22kHz); 250mW Preamp input maximum: .......................................50mV Preamp input minimum: .............................................................. 1.0mV Power amp input maximum:...................................................... 500mV Power amp input minimum:....................................................... 30mV Preamp phase distortion:.......................................±6.35° (below 200Hz) Preamp intermodulation distortion:.......................0.095% (88mV output 70Hz/7kHz) Preamp S/N ratio:...................................................-107dBV (10Hz-80kHz ref 630Hz 25mV) 32  Silicon Chip Good grief: The CHAMP is back! After constructing many circuits which offered good theoretical performance we discovered that the good old PreCHAMP preamplifier, combined with the equally elderly CHAMP audio amplifier, could be easily modified to do the job admirably. Yes, we know, we said only two months ago that our new CHAMPION amplifier would kill off the PreCHAMP and CHAMP but there’s a good reason for resurrecting it here: low quiescent current. The PreCHAMP and CHAMP draw only about 4mA each on idle, so prolonged operation (which you’ll need for field checks) is quite practical using only a 9V battery. By comparison, the CHAMPION draws up to 60mA so your 9V battery wouldn’t last long at all! If you built the CHAMPION project (based on the Panasonic AN7511), you could use it for infrasound with only a few modifications but you’ll probably need to use it with an external supply. siliconchip.com.au As used here, the modified PreCHAMP now has much improved frequency response; within +/-0.2dB from 2Hz – 20kHz. The modified CHAMP also gives a flat frequency response at around 0.25-0.5W – so you can feed any oscilloscope or low frequency counter. Optional CHAMP The CHAMP is optional – it has been included so that you can take quick measurements in the field. The PreCHAMP is set up as a bandpass filter and high gain amplifier which is approximately G-weighted, ie, its centre frequency is around 10Hz with -3dB points at 500mHz (0.5Hz) and 26Hz. A selector switch is provided for switching to “C” weighting (ie, flat response) so that the unit can easily be calibrated at 1kHz. The infrasound signal from the PreCHAMP is fed to the CHAMP amplifier which has been modified to give a flat frequency response from 0.5Hz to 20kHz and is set at high gain so that the signal output to a frequency counter is over 130mV at 1Hz. Electret Microphones The electret microphone is pretty inefficient at frequencies below 25Hz, hence the very high amplification. There are lots of electret microphone inserts available but we are specifying a particular Jaycar model (Cat AM4011) because we found it to be a very good match for this project. However, you can see from the graphs below that even these specific Jaycar mics are not all the same – some are more sensitive than others due to manufacturing variations – so you may need to buy a few to experiment. Frequency counter? Whoops! Haven’t got one of those? That little problem is solved very cheaply with a multimeter – specifically the Jaycar QM1327 auto-ranging multimeter, which can read read down to 0.1Hz and sells for only $34.95. While its specs state it needs a minimum of 3V RMS AC before it will show a frequency reading, we found that it far exceeded its specification and 130mV was sufficient. Few frequency counters go below 10Hz so the Jaycar meter represents good value in this application. If you use the CHAMP together with the Jaycar multimeter, then you will be able to determine SPL (the sound pressure level in dB) by switching the DMM to AC volts. This will give an approximate SPL in dB (decibels) as described. For signals below 0.5Hz this approach will not be accurate but this will be more than sufficient for the majority of applications. Because they have flying pigtails changing them is a pretty easy soldering task. Each electret will need to be calibrated as described below in the “Calibration” section to enable you to assess sound pressure level (SPL). By the way, we averaged the frequency response of several of the Jaycar electret microphones combined with the Pre-CHAMP and compared them with an accurately calibrated Bruel & Kjaer microphone/preamp (expensive!) – and found that the Jaycar electret was actually better at infrasound frequencies! Fatpigdog again! The direct readout is very handy in the field but if you want to do some real analysis, you’ll need a computer and suitable software. Readers may remember “fatpigdog” from our feature article on measuring siliconchip.com.au March 2013  33 We used this sweep to show that 1Hz was easily detectable with a resolution of 0.5Hz. By correction, the sound source is 100dB. You need to be patient because the analyser sometimes sets the sweep time to 10 seconds automatically and you have to wait before you can make adjustments. Loudspeaker Frequency Response in the December 2011 issue. We’re using this software again but it has since been updated considerably (the latest version is 4.04) and has more usable functions than the original version. You can purchase and download the software for around $30 from www. fatpigdog.com. On their website, you will also find various dedicated bench top spectrum analysers for sale but the virtual instrument is about 99.9% cheaper! Fatpigdog is fun to use, easy to manage and includes all sorts of extras such as a waterfall display, spectrum analysis to 22kHz, BMP capture and much more. The PreCHAMP output is simply fed to the sound card input of your computer. You could feed the spectrum analyser from the ‘CHAMP’ output but we don’t recommend this because your computer soundcard is usually set up for microphone-level inputs (ie, Using the Agilent 35670A, the sweep gives the lower response for the G-weighted PreCHAMP down to 0.1 Hz… that’s 1 cycle every 10 seconds! -3dB points are 0.5Hz and 26Hz. Mains hum is not a problem at these frequencies! 34  Silicon Chip millivolts not volts). Any large voltages will usually result in clipping and consequently the spectrum analyser will show multiple peaks from the odd harmonics. The Jaycar multimeter is an option if you wish to have a hand held detector for quick infrasound detection without having to set up a computer and adjust the software. It is fed from the Pre-CHAMP output via the 10k preset pot. You can set the maximum output from the Fatpigdog spectrum of a 15-inch speaker fed with 200W. The resolution is set at 1Hz and the sweep time is one second. You could actually feel the sound – and it was not nice! siliconchip.com.au The Jaycar QM1327 Multimeter works fine as a frequency meter and also an AC voltmeter. It’s simply held in place on the back of the Infrasound Detector with self-adhesive hook’n’loop tape (usually sold under the “Velcro” brand). “CHAMP” by setting the preset fully anti-clockwise. The only other modification is the addition of a 68Ω “dummy load” resistor which prevents the output capacitor from building up a DC charge, which would otherwise result in false readings. You could attach a loudspeaker instead but you won’t hear much below 25Hz (and it will drain the battery more quickly). By changing the parameters on the analyser – such as sweep time, start and stop frequencies and resolution bandwidth, you can save and print your spectra for further analysis. Furthermore, by setting the spectrum analyser to “max hold” you will be able to observe any infrasound which occurs during an extended period of time. Using the virtual spectrum analyser requires some practice and patience (just like a real benchtop spectrum analyser) but if you experiment, you will learn to master it all fairly quickly. We’ll have much more to say on this later. This spectrum shows the maximum sound level for suburban Pacific Highway traffic. The microphone is a good 5-10 metres away from vehicles and there is significant noise at 2Hz! Note also the peak at around 20Hz – probably from engines. siliconchip.com.au Construction The “hardware” is built into an aluminium diecast box (to minimise noise) measuring 119 x 93.5 x 34mm (eg, Jaycar HB5067). Inside this are the PCBs for the modified PreCHAMP (and CHAMP if you wish to use it) and a 9V battery in suitable holder. Layout is not particularly critical but given the very high amplification of the PreCHAMP/CHAMP combina- You can set up the detector and leave it running for up to an hour. We caught a distant thunder clap at 5Hz and a calculated 84dB. The resolution was set at 1.0Hz and the sweep time was one second. The maximum hold function runs continuously and updates every second. March 2013  35 9V BATTERY 1 F 120p S2 CON1 ELECTRET MIC TO SPECTRUM ANALYSER PRE-CHAMP 9V BATTERY HOLDER OUTPUT +V IN 14970110 CS GND GND NEW 39k RESISTOR SOLDERED UNDER PCB VR1 14920110 INPUT 4700F 16V 220F OUT GND CS S1 S1 'CHAMP' AMPLIFIER (MOUNTED SIDE-ON) 12V LED (MOUNTED ON CASE) +V GND 68 OUTPUT TO FREQ COUNTER SPST switch to switch the larger capacitor in and out of circuit but the arrangement shown (using a DPDT switch) allows easy mounting of the two external capacitors: they are simply wired across the outside terminals and the wires back to the PCB are wired to the centre terminals. Assuming you want to include the “CHAMP” power amplifier, to provide sufficient voltage to the Jaycar Frequency meter (multimeter), construct it as per the kit instructions (or refer to SILICON CHIP February 1994). The modifications we have made to give a flatter frequency response involve changing two capacitors. You will find that the 4,700µF capacitor is large but fits neatly on the PCB. However, it is a little too tall and the finished amplifier will have to be put on its side so it can easily fit in the diecast box. Now you can drill and mount all the hardware on the diecast box using the picture as a template and solder all the wires up according to the diagram. Fitting a tripod adaptor K A To enable easy use in the field, we wanted to be able to attach the unit to a photographic tripod. So we fitted Fig.2 (above): component layout is not critical but this diagram should give you a guide. Both the PreCHAMP and CHAMP PCBs are held in place with double-sided foam pads. The photo at right shows the same internal view, together with the T-nut tripod adaptor we fitted to the end of the diecast case. tion (about 4000 times), outputs should be kept relatively clear of inputs, as is normal practice for an amplifier. Start by constructing the Pre-CHAMP pre-amplifier as per the instructions given with a kit (or refer to SILICON CHIP July 1994). See Figs.1&2 for the modifications required – you will only need to change the values of three capacitors and these will easily fit on the PCB. The 39kΩ resistor should be soldered to the underside of the board input or across the input pins. To the two holes on the board marked “1n5” solder two leads and connect these leads to the two central pins of DPDT switch S2.Then solder the 1uF capacitor to one side and the 120pf capacitor to the other (see photo). Then run leads to CON3 and VR2 as shown in the wiring diagram. We could have used just a simple 36  Silicon Chip siliconchip.com.au As it has a 1/4-inch Whitworth internal thread (same as most tripods) we used one of these furniture T-nuts from a hardware store, flattened out the points, and drilled the box to suit. Then we glued it in place with some 2-part epoxy. our box with a 1/4-inch threaded bush (Whitworth thread; standard on most tripods/cameras/etc). In fact, we used a “T-nut” fitting intended for furniture and shelf hardware (pictured) which has an internal 1/4-inch Whitworth thread. It had four punched points intended to help it grip timber – we simply flattened these out with a hammer, then glued it in place with epoxy inside and out, making sure no epoxy got inside the thread. T-Nuts are available from most hardware stores and they are really cheap! However, you need to ensure you do get 1/4 Whitworth – it appears that 5/16 and 3/8 are much more common. If you must use 3/8-inch, 3/8 to 1/4inch adaptors are available from better photographic stores. Finishing off It will be easier to solder the wires to the boards first, then solder the wires to all the switches and sockets before mounting them inside the box. Because the circuit boards are tiny and sometimes have no provision for normal screw mounts, you will have to use some good quality, thick double sided foam pads. Cut it to cover the bottom of the “pre-CHAMP” board then Here’s how we mounted the electret microphone, using an old dynamic mic windshield as the base. The insert is held in place with an adhesive foam tab. press it firmly in place, allowing plenty of room for everything to clear. Then fit the “CHAMP” amplifier by putting double sided tape to the side 4,700µF capacitor and the side of the board and then pressing it all into place as shown. Check again to see if any wires have come loose then mount the battery in its holder and switch on. The current drain should be about 8mA or so. If all is OK, put the lid on and plug in your computer, set up the software and start testing. The microphone For the microphone assembly, drill a hole large enough for the electret in the base of the box, solder a short length of shielded cable to the microphone with the shielding to earth (the side connected to the outer case of the electret) and the other end to the the input terminals of the PreCHAMP. We are looking at frequencies below 30Hz on the G-weighting setting so hum should not be a problem until you switch to C-weighting We cut the top off an old dynamic mic and mounted it on the box, then attached the electret to the side with double sided tape as shown. We maintained the original mic thread to allow us to attach a wind shield and also to calibrate our setup and to make quick changes to test various microphones without having to unscrew the box all the time. But this is not critical and you can just stick the electret to the inside of the box with double-sided tape or even solder it directly to the input pins and just have an appropriate hole in the diecast box. Whatever you do, you should be able to access the electret to enable Parts List – Infrasonic Detector 1 PreCHAMP Kit 1 CHAMP Kit [optional - see text] 1 Diecast box (eg, Jaycar HB5067) 1 frequency-reading multimeter (eg, Jaycar QM1327) [optional - see text] 1 SPST miniature toggle switch (S1) 1 DPDT minature toggle switch (S2) 1 3.5mm mono socket, panelmounting 1 banana socket - red 1 banana socket - black 1 short red wire fitted with banana plugs each end 1 short black wire fitted with banana plugs each end 1 electret microphone insert (eg Jaycar AM-4011) [see text] 1 microphone 1 1/4-in Whitworth T-nut for tripod mount [see text] 1 9V battery 1 U-shaped 9V battery holder 1 3.5mm to 3.5mm shielded audio cable (to connect to sound card) Short lengths hookup wire and shielded audio cable Double-sided adhesive foam pads Self-adhesive hook & loop tape, etc Epoxy glue (for tripod adaptor) 1 Fatpigdog Virtual Analyzer program (download from www.fatpigdog. com [approx. $30]). Semiconductors 1 LED, panel mounting 12V type Capacitors 1 4700µF 16V electrolytic 1 1000µF 25V electroyltic 2 470µF 16V electrolytic 1 1µF MKT 1 120pF ceramic Resistors 1 39kΩ 1 8.2kΩ 1 68Ω 1 100kΩ (or 50kΩ) log pot 1 knob to suit pot quick changes because there is significant variation between electrets as the graphs will show and having it mounted on shielded cable makes it easier to solder and unsolder. Checking it out Once everything is done connect the output from the pre-CHAMP to your computer Mic input making sure your siliconchip.com.au March 2013  37 Frequency ADD dB to   (Hz) measurement 0.5.................................. 41 1.................................... 29 2.................................... 17 3.................................... 11 4...................................... 8 5...................................... 5 6...................................... 4 7...................................... 3 8...................................... 2 9...................................... 1 10..................................... 0 11..................................... 0 12..................................... 0 13..................................... 0 14..................................... 0 15..................................... 0 16..................................... 0 17..................................... 0 18..................................... 0 19..................................... 0 20.................................. 0.5 21..................................... 1 22..................................... 1 23.................................. 1.5 24.................................. 1.8 25.................................. 2.5 26..................................... 3 27.................................. 3.2 28.................................. 3.3 29.................................. 3.4 30.................................. 3.5 Table 1: Correction table for a Jaycar AM-4011 electret mic insert. sound card mixer is set flat; ie, no bass or treble boost. Check to see if the microphone is working by switching to C-weighting and then talking or whistling. Measure the output with a DMM set on AC or plug the output into an amplifier or oscilloscope. Alternatively, you might like to plug the output of the Pre-CHAMP into the mic socket of your computer soundcard and view your “whistle” on the spectrum analyzer. Your whistle should give you a peak at around 1-2kHz, plus harmonics at 2 and 3kHz. Once all your checks are done switch it to G-weighting and observe the LED (assuming you have added the CHAMP) It should flash in time with the signal and you can open and shut a door to test it (a car door opening is approx 0.52Hz). If all goes well you will finally be 38  Silicon Chip ready to fine tune it all and try some infrasound testing. Plug in the Jaycar multimeter, switch it to the Hz range and read off the frequency. On C-weighting you will probably see something in kHz but on the G-weighting setting you should see frequencies below 20Hz. The frequencies will probably jump around a bit and you can vary the gain control to stabilise the readings. In the SILICON CHIP office, we saw 8Hz coming up consistently on the counter and also on the spectrum analyser. It disappeared when we switched off our air conditioner but it was a hot day so we put up with the 8Hz (although it was less than 75dB [SPL]). Calibration As we mentioned before, calibration is only really needed if you want to establish sound pressure level . Frequency calibration is already inbuilt in the software and multimeter and is not required for our purposes. It is fairly straightforward but it will help if you already have a sound level meter (like the Jaycar QM1591) and an audio oscillator but if you don’t have these items and you don’t calibrate, you will still get a pretty good idea from the relative dB levels indicated in the spectrum analyser. For example our leaf blower is rated at SPL 70dB at one metre. We set the detector to C-weighting and found that the fatpigdog analyser indicated -15dB at 35Hz at 1 metre, so switching to G-weighting will mean that any infrasound frequency BELOW 26Hz will also be 70dB, if you see -15dB on the analyser. Sure, it’s a rough measurement but there are many devices which have a dB rating on their label such as mowers, snippers, saws etc and you can check these out. For a more accurate calibration, feed a tone (say 1kHz) through an amplifier and loudspeaker and check your C-weighted result against your C- weighted sound level meter. Try various levels, incrementing them by 5dB. Most sound level meters have absolutely no response below 35Hz so there is no point checking the G-weighted setting. If you don’t use the fatpigdog software, don’t worry because you can switch the Jaycar multimeter to “AC volts”, making sure the gain control is fully advanced and just take note of the reading at various sound pressure levels. Our setup showed approx 0.9v AC at 94dB. For frequencies below 7Hz the accuracy falls off somewhat but if you are looking at 0.5Hz, just switch it to DC volts and watch the rise and fall! Other unique applications – vibration anaysis This instrument is very useful in checking out vibration problems as we found with our 8Hz air conditioning. Sometimes these problems go undetected for years and some have claimed that they may be responsible for nausea, headaches, sleep problems or just a general sense of unpleasantness. Additionally, traditional methods of sound level monitoring have only focussed on the audible spectrum and have not even considered infrasound effects on the human (or animal) body and the access to infrasound measuring devices has been both difficult and expensive. Any vibrating device will give off sound and our setup will detect it and/or datalog it. Not only that but for a few dollars it could be used in just about any industrial situation where vibrations may be destructive – such as engines, chassis, suspensions even buildings and bridges! Data logging with waterfall analysis The software will also enable you to do waterfall analysis and this is really a way of viewing a spectrum analysis as it varies over time. It can be used as a datalogger for infrasound and audio signals. The vertical scale shows the frequencies of the various harmonics while the horizontal scale is time so the whole chart is a record of a few minutes. TO SETUP FOR WATERFALL CHARTS The wiring setup is virtually the same as for testing spectrum analysis microphone “Pre-champ” output (for voice prints) The setup for the virtual instrument is: Click on “preset” Then “display” Then “waterfall F2” Then “rotate” Then try different sweep times and resolution bandwidths (Res. Bw…..). And try different colour schemes by clicking on “jet” Press BMP to save the image you want. siliconchip.com.au Setting up and operating the Virtual Analyzer We assume you have downloaded the spectrum analyzer software from www.fatpigdog.com/spectrumanalyzer (or updated if you’re using an older version). The originator, Spyro Gumas, is very communicative and can assist if you have any problems. To start, open and run the program. We used Windows XP but check the website first for compatibility with Vista, Windows 7, 8, etc. Initially, you will see the black and white MS-DOS screen appear. You may have to wait (perhaps two minutes or so) and the instrument will appear similar to the trace below: This sweep shows the frequency response of the modified preCHAMP: the top line is C weighted and is flat from 10Hz-20kHz. The middle line is an unmodified pre-CHAMP (not used) and bottom line (red) is the G-weighted response which joins the top line at 10Hz.Calibration can be carried out at 1kHz on the C setting and then is the same when the unit is switched to G, up to a max of 20Hz! NOTE: Our Audio Precision analyser cannot go below 10Hz. The analyser is now ready to do a ten-second sweep of your sound source from 0.5Hz to 100Hz with a resolution of 0.5Hz and will continuously update itself with the maximum signal. For example we set it going during a thunderstorm to record the sound over a period of 20 minutes You can save an image anytime by pressing “BMP” (bitmap). You can play around with the RBW (resolution bandwidth) which you can set as low as 0.1Hz! Refer to the fatpigdog manual provided if you have difficulty because some computers have different delay arrangements with the soundcard and you may need to compensate this with “tstupid”. When you are happy with a particular trace, you might like to activate the marker to examine point of interest. Click on “marker” then “ON” and then click “peak”. The marker will then indicate the dominant frequency You will see a red dot appear on the trace, then move the marker to the area you want to measure by clicking on “<” (backward) or “>” (forward) keys. The marker reading appears at the top of the page e.g “Mrk 2.558Hz, -86.2dB”. Once you have measurements of the points you are interested in, go to Table 1 and add or subtract the dB value at the frequency of interest. For example if you measured -10dB at 5Hz from the chart you have to ADD 5dB, ie -10+5=-5 Now during calibration, for our setup we found that -15dB on our spectrum analyser was 74dB SPL so we have to add 10dB (because -5dB is 10dB louder than -15dB). So SPL=74+10 SPL=84dB Accuracy The figures quoted in this article are those achieved on a PC fitted with a generic sound card (ie, nothing special!) so we have every reason to believe that you should achieve similar results. However, no guarantees can be given! SC Once the virtual instrument pops up, plug the output from the prechamp into your soundcard mic input, switch to G weighting then set up as follows.You can attach the multimeter to the CHAMP output if you wish but in this case it is redundant. On the virtual analyser: Click on “reset” to clear any previous settings. Click on frequency Click on start (F2) and type in “0.5” <enter>. Click on stop key (F3) and type “100” <enter> (This sets the range to 0.5Hz-100Hz) Click on Lin/Log key (F4) so you see lin/(log) – now the frequency range is set to a logarithmic scale. Then: Click on bandwidth Click on RBW and type in “0.5” <enter> Click on sweep and type “10000” <enter> Click on “trace” and then “max hold” The analyser will then sweep continuously and indicate the number of averages at the top of the page. siliconchip.com.au This shows the actual frequency response of the finished setup using the Jaycar electret and is usable down to 1 Hz! You need to allow for falloff by using the table provided. Eg, for 1Hz you need to add 29dB to your base figure to obtain the correct SPL. March 2013  39