" 5 ae ita eS so " ees ly ia eee mobi al sb a7Programmable High Energy Ignition Kit This advanced and versatile inion system can be used on both two & four stroke engines fo intercept and modiy the factory timing of asthe basis for 2 stand-alone igniton system wit varable gnton Hing, ‘ection col eonol and ane-knock soning KI inclses PCB wih over, programmed mero, al eleetont: $6995 PCB: 102 x 81mm (C5442 ‘Tempmaster Fridge Controller Mk Il Kit ‘Tuim an od chest eez0" no an energy-aficientfige or beer keg tye Grconver a standard fndge ita a wine cooler, These are just wo of he obs ths low-cost and easy-to-buld electron thomostat kt can do without he noec {o modi intemal wieng! Shot-ferm ki contains PCB, sensor and al spoctieg ‘components, You need a3 your wn 240VDC GPO, eutohod IEC socket and case $3295 * 240V 10A Deluxe Motor Speed Controller Kit — = PCB: 08 x67mm KC5476 “The dene motor speed controler kt allows the speed of @240VAG motor be Contoed smoothiy fom near ero ful sped. The advanced design provos Improved spoed regulation & low speed operation, Also features sofia, Iterorences suppression, fuse protection and overcurrent protect + Kit supled wh all pars incuding prescut etal case Oe pet Bat = Capacitor Discharge Ignition Kit for Motor Bikes Many modem motor kes use a Capactr Discharge litian (CD) to improve perlrmarce and enhance relay However. the CD nition module fal, Eoplacement can be very expensive, This kt wil replace many fled factory unt and seultable for engines hat provide a postive Capactor volage and have a separate wager cal. Supplied ‘ith solder masked PCB and overlay, case and a hardware requrea = POB: 45x 4mm Kc-546, 433MHz Remote Switch Sutabiofor emote conta of practically anything up toa range of 200m. The receiver has momentary or topale output an the Momentary period can be fojusted: Up fo five receivers can bo used inthe same vc. Shor form it Shine we BOD sa pthed conse $4qes + Requires case & 9 12VDC power = POBS.TX: 85x 63mm FK.793 48mm Ke-5a73, Entra vansmiter kt: KO5474 $22.95, ‘An extemely use and versal kt that enables you to use a tiny gger ‘utrent =a low as 400) st 12VDC to sete upto SOA at SOVOC. h hh an soisted np, and ie suitabe fora vanity of tragering ‘components. ‘options. The tines PCS with $4995 peed 16 DC Relay Switch Kit ‘very ang al electron a = Voeme= Oe Theremin Synthesiser Kit MkIl Kest7a $9935 Ces? was Marine growtnelecronc atfoulng systems can cost thousands. "Tha projet uses the same ultrasonic wavolorms and vtualy inca utlrasoni transducers mounted in sturdy polyurethane housing. Standard Uni consist ofa contol eecronc kt and case, preultuasonc transducer, lung components and housings, The single Wansducer design of hi kt slab for Boats Up to 10m (32) oats longer ran about “tam wit need two transdcers aed crvers. Bascaly a pars re supplied in he project Kt ineuding wiring, = s2qgm + Stale for power o sll * Could be powered by & solar panel vind generator = POB: 104 Tame Kc-5496. $74q35 Jacob's Ladder High Voltage Display Kit MK2 {Win hs tana the purchase oa Y2VDC ition cl avaaie rom auto stores and parts rocylers, rea an awesome ring laser of oy sar hat nis the asc ‘rot of drone Kitincades PCB, po-ct wiroiaddor anda cron components + 12VDC automate jgiton col and case ot incudes 1 12V ca altry, TAR SLA or >5Amp OC power Create your own eerie sconce fiction ‘Sound affects by simply moug your ane near te antenna. Easy fo setup and bul Complete kt coniar tase and all spect ~ FCB with overly, pre-machined components “guts '2YBC poe si comeanel = esta siry = 9 = =) = Piste enn Senn == veers ccses RED (ss Improved Low Voltage Adaptor ALITA 1.3V to 22VDC IA Voltage Regulator Kit Prowde upto 1,000mA.at any voltage trom 1.3 22VOC Ideal for expermantal projets 0” as mri banc Dower supply et Kit supped wah PCE Bnd al eectone components $1655 OCU eed Seemeenemiinlll uns avant of doves such a CD or MPS players rom your car ciara lighter sake or even powered speakers fom the power suply nse your PC. vnllspply ites 8V. 5 6V, 2 12 or 1SV and (hen used wih an appropriate Input wotage and hetenk) geiver opto 48 lhe saced eutput votage. +t ncides sreen printed PCB and all spoctiod $4795 comoenents (heatsink not included) OB 108% 37mm Kc-5463, = PCB: 38 x 35mm KostasFeatures 14 Argo: Drones Of The Deep Oceans Thousands of drones are floating deep in the oceans of the world, monitoring temperatures and other data. They are fully autonomous and can change depth or rise to the surface to send data to satelites - by Dr David Maddison 22 Review: AmScope Stereo Microscope Even people with 2020 eyesight have difculty soldering fine-ptch MDs to PCBS. Tis stereo microscope Is made forthe task and even has @ camera ‘eyepiece to record your work ~ by Andrew Levido 36 Eye-Fi Mobi SD Wireless Camera Cards Eye-Fi have gone mobile with their new range of "Mobi camera SD cards. Now ‘you can very easily send your pictures fo any Android or iOS tablet or phone and share them with friends! ~ by Ross Tester Projects To Build 26 Threshold Voltage Switch This versatile unit switches a relay when an input voltage crosses a preset threshold. Use it to switch power to a fan, warning light or similar or use it to prevent a lead-acid battery from being over-charged - by John Clarke 60 Micromite, Pt.3: Build An ASCII Video Display Terminal Its VT100-compatible and lets you add a video display, keyboard and USB to ‘your next microcontroller project. I's the perfect companion for microcontrollers with serial input/output such as the Micromite and PICAXE- by Geoff Graham 70 Touch-Screen Digital Audio Recorder, Pt.2 ‘Second article gives the assembly details, provides some performance graphs and describes how it's used — by Andrew Levido 80 L-o-0-o-n-g Gating Times For The 12-Digit Counter This module enables even higher resolution measurements with our 12-Digit Frequency/Period Counter. adds a decade divider for the external timebase input to allow measurements using a gating time of 10,000s - by Jim Rowe Special Columns 40 Serviceman’s Log (Oh goodie, a valve radio to fx ~ by Dave Thompson 57 Salvage It! Wrecking The Computer Itself (Once you've pulled the main stuff out of your old PC, what do you do with the rest? Don't bin it yet: there's still lots of goodies waiting to be recycled ~ by Bruce Pierson 87 Circuit Notebook (1) Signal injector & Tracer Uses TDA2822M Dual Power Amplifier; (2) Head- light Circuit For A 36V Electric Bike; (3) RS-232 Input For Non-Standard Signals; (4) Simple DIY Gizmos For SMD Desoldering 92 Vintage Radio The upmarket 1950 HMV RSSA radiogram — by Rodney Champness Departments 2 Publisher's Letter 90 Online Shop 4 Mailbag 98 Ask Silicon Chip 47 Subscriptions 48 Product Showcase 103 Market Centre 104 Notes & Errata SILICON > Threshold Voltage Switch - Page 26. ASCII Video Terminal With VGA, ‘Composite Video, USB Port & PS/2 Keyboard Input - Page 60. Building The Touch-Screen™ Digital Audio Recorder ~ Page 70. Long Gating Times For The 12-Digit High-Resolution Counter ~ Page 80. Jury 20141SIMEON eee eee Leo Simpson, B.Bus., FAICD Cer Greg Swain, B.Sc. (Hons.) Technical Editor POU aa) Dr Ross Tester ey Cees Td Cesc Rey ‘Ann Morris Advertising Enquiries Glyn Smith CE kad ‘Mobile 0431 792 293 Nic ee Regular Contributors Ce Cray Kevin Poulter Stan Swan Pend Pete eed a yearby Silicon Chip Publications MeO) Cee OR nat oe a ten consent ofthe publisher. Printing: Hannanprint, Noble Park Pits Pea ad oy REC Rad MM ogr eect) Pies Tee std rs Ps Tee CEA} ee eee Ce EL Ee Er) E-mail: sillcon@siliconchip.com.au {ssn 1030-2662 eee uae aLoea SILICON CrP Publisher's Letter Argo drones gathering deep sea data This month, we havea most intriguing story, written by Dr David Maddison. Did you know that there a thousands of drones drifting deep in all the world’s ‘oceans, quietly gathering temperature and other data? Most people are aware of airborne drones and their amazing capabilities in surveillance and remote wai fare but few would know that there are thousands of drones in the oceans, including in the Antarctic and often right under the sea ice This program started in 1999 so there has been a gradually increasing fleet of these drones over the last 15 years. They have since collected a mass of deep ocean data and will continue to do so at an ever increasing rate. Some 30 nations are involved in the Argo project, the USA being the biggest, followed by Australia, How can these drones communicate and deliver their collected data? They are programmed fora 10-day cycle which continues for many years, until their batteries are exhausted. For most of that 10-day cycle, they float at around 1000 metres, drifting in the deep ocean currents. Then they sink to 2000 metres and then slowly rise to the surface, logging temperature and other data as they They beam their data to satellites and then submerge to begin the cycle ag I is most important that all this data is collected and carefully analysed for it will tell us much about the world’s climate and how it is changing. Un doubtedly, it will tell us a lot more about the deep ocean currents and how they contribute to ocean temperature cycles which can run over many decades. Some climate scientists believe that the current pause in “global warming may be due to the excess heat being stored in the deep oceans. Maybe they are right but it could be many decades before the Argo drones demonstrate the truth of that notion. Arguably, the drone program may have to be considerably extended because great areas of the world’s oceans are considerably deeper than 2000 metres. This fact has been emphasised by the loss of Malaysian Airlines Flight MH370 which may be in very deep water indeed. So deep that very little of today’s technology has any chance of finding it. Maybe the Argo drones might have found it if they had been equipped for such monitoring, Actually, the Argo program must be regarded as an initial stage in exploring the deep oceans. While virtually all of the Earth’s land surface has been fairly thoroughly explored, much of the world’s oceans are about as well known as the far side of the Moon. For example, while we know that there are many thousands of undersea volcanoes, relatively little is known about their activity and how they might be affecting the undersea environment. Perhaps in the future we will see undersea drones which don't merely drift with the ocean currents but which are actually self-propelled, skimmi: vast areas of the ocean floor while they collect video and other data, returning to the surface to be recharged and then to be programmed again for other mis. sions. Imagine how these might have expedited the search for Flight MH370 as pods of these things were launched, like robot dolphins, to search the seas. Or perhaps such drones could be programm or monitor the life of large fish shoals as they roamed the oceans. It is an ex. citing prospect, is it not? It is gratifying to know that Australia is taking such a major part in the Argo project and it is to be hoped that we can similarly be involved in future exploration of the world’s oceans. d to follow whale migrations. Leo Simpson siliconchip.com.auValue Instruments: The quality you expect at an unexpected price. Ret High quality = high price? Not with our Value Instruments. Value instruments are versatile T&M instruments for everyday lab use. 1 Quality T&M solutions engineered by Rohde & Schwarz | Accurate, reliable, easy to use | Comprehensive support thanks to the extensive service and technical ‘support network www-rohde-schwarz.com.au sales.australia@rohde-schwarz.com ind the right tool here: www.rohde-schwarz.com/value ROHDE &SCHWARZ siliconchip.com.au Juv 2043Suggestion for a simplified head-up display Thave been monitoring the interest in head-up displays (HUDs) in your ‘magazine. [believe the main purpose of a HUD is to provide basic informa- tion to the driver without the need to take his/her eyes off the road or even focus on a nearer display. Ibelieve this could be economi and simply achieved with a single tricolour LED. My idea is that this tri- colour LED would be positioned on top of the dash, reflecting in the screen, It would be green at the ‘set” speed, blue below that speed and red above that speed, with the green coveringa couple tkm/heitherside ofthe sotspeod. An enhancement could be flashing of the red about 5km/h above the set speed. Setting could be done by pressing a button at the set speed or by turning a knob until the green LED lights up at the desired speed using the speedo (or tachometer) to provide the necessary information but only once for a given setting, This information would be retained until aneed tochange it arose, ly RIGOL DS-1000E Series. » SOMH2 & 100MHz, 2 Ch » 1GSis Real Time Sampling » USB Device, USB Host & PictBridge FROM S339 csr Letters and emails should contain complete name, address and daytime phone number. Letters to the Editor are submitted on the condition that Silicon Chip Publications Pty Lid may edit andinas the Fight to reproduce in electronic form and communicate these letters. This also applies to submissions to "Ask SILICON CHIP" and “Circuit Notebook’. possibly at the next town, or retained long-term fora given situation such as highway speed. Input of speed data could ideally be as versatile as possible to allow fitting to all vehicles independent of age or planned end use. [would envisage this unit being useful as an indication of more than speed, including optimum revs independently of gear and pos- sibly to indicate the tacho red line. ‘The system does not even need the driver to focus on the light and it could easily be positioned wherever suits the driver or even directly viewed. Dim- ming at night would be required and in conjunction with this light there could be others for temperature and oil pressure which could be easily organised with existing systems based on frequency or voltage. Jeff Montague, Toora, Vi Comment: this project would require @ microprocessor and not much other circuitry. We are not sure that most readers would want to settle for such a simple display, given that the micro “Rigol Offer Australia’s Best Value Test Instruments” NEW RIGOL DS-10002 Series » 7OMH2 & 100MHz, 4.Ch » GS/s Real Time Sampling 12Mpts Standard Memory Depth FROM SG DA cs: Buy on-line at wwW.emona.com.au/rigol could easily drive a 4-digit 7-segment display with little extra circuitry being required, Belief in warming effect, of carbon dioxide While John McDonald (Mailbag, page 8, June 2014) is “absolutely amazed... people . . . are still be- ing sucked into believing this global ‘warming nonsense”, | am absolutely amazed thal, given the credible scien- tific evidence of the heating effect of carbon dioxide released by burning fossil fuels, there are people who still do not believe in anthropological- generated climate change. Perhaps the common ground is that we both believe the climate is changing. Therefore, | propose a risk management approach toan issue that cannot be confirmed one way or the other before any necessary mitigation is too late to be effective. NEW RIGOL DS-2000 Series » TOM, 100MHz & 200MHz, 2.Ch » 2GS's Real Time Sampling > 14Mpis Standard Memory Depth HOM 93M 4 SmIcoN CP siliconchip.com.au_. Design, Develop, Manufacture - - with the Latest Solutions! - ‘ In the fast paced world of electronics you need to see, test and compare the latest equipment, products and solutions for manufacture and systems development. CT ae Seas Sib ‘The Electronics Design’and Manufacture Conference delivers the latest information on a host of topics... + New product, system & component technology The last conference in Sydney attracted over eee ere PEP ner aurea ner EUs Ot ‘ 3 rors ees i Sec Ce ura For conference details please visit Pte acuity fate Oo Smee ec oo Cee eee a : Hi : a Fate Ph: (03) 9676 2133 fall Oa Se ees Sd LY Lu erat www.electronex.com.auMailbag: continued Concerns about wiring in Chinese electronic equipment Recently, our test section took delivery of an inexpensive frequency counter directly from China. It was fitted with a flat US-style 2-pin mains plug but before I lopped it off to fit a 3-pin plug, I opened the unit to have a curious look around, The first thing that caught my at- tention was that the flimsy primary leads to the small mains transformer were bundled and cable-tied to the equally poorly insulated second- ary leads. I doubt they would have withstood the standard one-minute KV isolation test. The leads were duly separated but it soon became obvious that the fuse was wired into the Neutral lead of the mains cable, Given that this unit would only be earthed via the test lead plugged into the front panel, any. ‘o-chassis internal fault would be effectively unfused, exposing the operator to both high voltages and/or high currents; a very unsatisfactory situation indeed. However, before correcting this my attention was drawn to the 5V regulator heatsink that was sitting crooked on the board. The corner of this bent aluminium shape was now less than 2mm from the solder pad that brought the mains lead up to the PC-mounted on-off switch. Here was exactly the type of fault that could put 240VAC on the test leads! The heatsink was fabricated with oversized mounting holes that read- ily allowed it to slop over towards the mains switch. It had apparently been this way since assembly, as all the screws were tight. Closer inspection of the PCB showed that in places, the 240VAC. tracks came within 2.5mm of the chassis connected ground plane. ‘This is well inside the 4mm stand- ard practice for double-insulated equipment such as this and would cortainly fail the isolation test. A little work put all this right but poor design and poor assembly prac- tice created a potentially dangerous situation that took the shine off the bargain price. Incidentally, afew months earlier had encountered aChinese-manufac- tured power supply (this time from ‘an Australian distributor) with no continuity from the chassis to earth. ‘The mains earth lead was terminated in an eye-lug that was held against the chassis with a bolt and nut. The problem was that the painted chassis, insulated the lug and the bolt. Obvi- ously, their final factory tests do not include earth integrity. End users of electrical equipment rely on the competency of design- ers and manufacturers for our very safety but in the case of Chinese origin gear, it pays to doa few simple checks yourself. Plainly, the mains wiring practices of some Chinese manufacturers are still dangerously inadequate. Keep up the good work SILICON Cup magazine. Your efforts are consistently world-class and well appreciated. Mark Hallinan, Sunnybank Hills, Qld. Faced with a potential catastrophic ‘show stopper’ risk, best project man- agement practice is to mitigate the risk. An astute project manager will do everything possible to reduce the risk no matter what the chance of it occurring. After all, if the risk comes to pass, all other effort is for naught. There are two possible outcomes in this approach. First, mitigation action of reducing greenhouse gasses and preparing for environmental change is taken in a considered, orderly and af- fordable way (although Idon’t dispute that there is an actual cost in the same ‘way that payingan insurance premium reduces disposable income for more enjoyable pursuits) and it turns out climate change is natural. In this case many of the measures taken will be necessary to help man- kind cope with the altered environ- ment; a small “w” win as society and the economy cope. Second, mitigation action is taken as above and the climate is being changed by mankind; a capital “W" win as society and the economy suc- 6 SILICON CuP cessfully adapt with the least cost and inconvenience overall. For completeness of options, let’s consider the alternative business-as- usual scenarios: nothing is done and climate change is natural. Money will still need to be spent in the future to somewhat ameliorate the effects of the changing climate on agriculture, dwellings, workplaces, transportation, infrastructure, medicine, defence, etc. As any project manager will tell you, rushed ‘fix the problem now’ ‘expenditure is always very expensive. Society and the economy will suffer a severe negative impact. Finally, climate change is man-made, It’s business-as-usual and, having ig- nored the warnings and continuing to exacerbate the effect, the worst case predictions are the future reality. De- spite massive expenditure to finally deal with the issue, it’s too little, too late as by this time climate change is unstoppable for over 1000 years. The massive expenditure and nega- tive impacts on production drive the world’s economies into severe depres- sion. Over the millennium, society is turned upside-down trying to cope with the environmental, economic and dislocation effects of sustained climatic changes. Considering the very likely out comes of each scenario, I think the only sensible and safe option is to immediately discuss what is the ap- propriate balance of GDP to invest in ‘greenhouse gas reduction technologies and climate change amelioration pro- jects. Then implement the mitigation strategy without delay. Despite the low risk and high premi- ums, no-one regrets having insurance when their house burns down. Ifthere is any disappointment, it usually is wish I'd paid more to make sure I was fully covered. ‘Tim Herne, Calwell, ACT. In defence of climate modelling George Ramsay's statement (Mail- bag, May 2014) that scientists, using computer models, “extrapolate past siliconchip.com.auEmbedded 3D Gesture and Position Tracking © with Mictoghip ‘ 's GestIC® Technology i Microchip's patented GestIC® 3D Technology provides a single-chip, real-time gesture recognition solution. The MGC3130 enables the next breakthrough in user interface design for any product at a very low cost point. MGC3130 _ The MGC3130 runs the Colibri © The MGC3130 delivers the eee Mca oleh mam TN Cece Retin eoeeuienkaKesuis el) para RR GCM oct eee acee NE Uiema cure keg eee] Perenci (er Le Microchip Technology Australia Tel: (02) 9868 6733 E-mail: aust_nz.inquiry@microchip.com microchip _EIRECT MICROCHIP microchip.com/gestic TRSSSRIEn atest ant mit tipMailbag:| continued A while back, I wrote a criticism of your GPS-disciplined 10MHz standard (S1LicoN Cp, March-May 2007 & September 2011). At the time my ideas for improvement ‘were theoretical and as we all know “in theory, theory and practice are the same, in practice they aren't” To cut a long story short, the ideas have been incorporated in a design that bears little relationship to the original SILICON CutP circuit. The end result is built on strip- board as a ‘proof of concept’ rather than a finished product. Given that many readers would have more design skills, more test equipment and more use for the end product, I believe they should have no trouble adapting the design for their own uses. Itis written up at www.cashin, net/projecti/index.html The design uses a commercial 10MHz OCXO (oven controlled crystal oscillator). Older used ones are available on a well-known site for less than $30. Apart from this and a GPS unit, the components are their sot of data points” to “predict the future” reveals a misunderstand- 1g of how current computer models forecast weather and simulate future climate. Such models are based on known laws of physics such as the conservation of mass, Newton's laws of motion, the first and second laws of thermodynamics, the laws governing radiative transfer ete. ‘These models integrate differential equations that describe the physical laws obeyed by the atmosphere, using the current state of the atmosphere (current weather) as a starting point. The equations are time dependent and so are predictive. Such models do not “extrapolate” from a prior sequence of data points! ‘The accuracy ofthe forecast depends on the accuracy of the input data, the resolution of the model and the inclusion in the model of all relevant hysical processes. Clearly, unless the input data has infinite resolution and 8 SILICON CHIP inexpensive and readily available. A PIC16F628A is used to discipline the OCXO and after some ‘burnin’ time, ‘compares the GPS and OCXO over 38-hour periods. In a day or two, the OGXO settles to much better than one part in 10%°, Original tests were done using a poor antenna location in the house, Where the temperature is relatively constant. The worst case result after two days was two parts in 10", and better than one part in 10" for most 3-hour periods. The rig is currently in a carport with the antenna in a favourable position (on the roof) but the rigis subject to more temperature variation. lexpect to have the results for this location by the time this let- ter is published. ‘These results are in the same Jeague as rubidium standards with- outany set-up procedure or age limit, costing the same or less and with the satisfaction of ‘rolling your own’. If anyone is interested and wishes to correspond, there is an obfuscated email address at the link above. Alan Cashin, lington, NSW. is perfectly accurate, unless all physi- cal processes are included and unles: the computer has unlimited speed and storage capacity, the forecast weather ‘will diverge from the observed weather as the forecast period increases. Most weather models would not claim to be accurate past about a week. So it is reasonable to ask how climate models can claim to “predict” the climate far into the future, ‘There are important inputs to mod- els other than the current weather. Called climate forcings, these include boundary conditions such as surface type (soil, sand, water, snow, sea ice, vegetation, topography, albedo etc.), the composition of the atmosphere (particulates and gases including water vapour and greenhouse gases because they strongly influence radia- tive transfer) and the source of nearly all atmospheric energy, solar radiation, They are especially important in the long term; there are many more, Suppose a model is run to simulate the weather over a long period, say decades. Now we are not interested in specific weather events but rather in averages. We have a climate model. ‘The simulated climate (average weath- er patterns, temperature and rainfall distributions etc) during the period can be calculated and compared with real climate statistics. If there is good agreement for past periods when climate forcings were known, then the model should be capable of simulating how the cl mate might change under differing forcing scenarios. For instance, what changes of climate might be expected if the concentration of a particular greenhouse gas were to double? What if particulates increase due to an im- mense volcanic eruption? What ifsea ce disappears from the Arctic Ocean? What if Brazil is deforested? Ifyou are seeking additional infor- mation about Global Climate Models (GCMs), a good place to start is the Geophysical Fluid Dynamics Labora tory website at Princeton University, http:/Avww.gfdlnoaa.gov/climate- modeling Bob Lile, PhD, Princes Hill, Vic. Scepticism should be the basis of climate science Just wanted to let you know I loved the Publisher's Letter in the June 2014 issue, on reducing clutter and organ- ising stuff to make things easier for others later on etc. We've just been clearing out all areas of our house and the effect is brilliant. It seems to help clear the mind as well. My workshop so much more functional. I used to foe] overwhelmed by the sheer quan- tity of junk and unfinished projects, to the point of paralysis and stagnation, ‘Also, I wanted to express my appre- ciation for Leo Simpson's continued sceptical stance on climate change, despite the volume of critical mail you seem to be getting. I don't pretend to beatall knowledgeable in that area but to me the whole thing stinks, just on the basis of how anyone expressing the slightest bit of scepticism is labelled a “non-believer” and assumed to be in collusion with the evil polluters! ‘miners etc or be politically motivated and have extreme right-wing views, —____| siliconchip.com.aupe lifetet | SSG TraceME modules can save your business a lot of money because: Re eee eT nec Crue Renn g ee ead ea eee at ees They have countless onboard sensors and advanced software to reduce battery eo eRe gee cel aes eae eae tran eea aCe tee Ete (Mey olteCu cL ee MAIN FEATURES [GESTANOEGIONASS) GF GM RANEs) [MULTIEUEZSENSORS) 6 1-(0)733-197-302 nsland, AustraliaMailbag: continued volume levels with good audi {with the odd exception, as sor Favourable experience with digital hearing aids Over the years I have followed with considerable interest the Siti CON GitP articles on the Australian- manufactured hearing aids (ini- tially Australia Hears, now Blamey Saunders Hears). I would like to add my comments to writer G. H. “Hearing Aids Are Too Expensive” (Ask SiLiooN Cin, June 2014) and Ross ‘Tester's response, the points of which are confirmed by my ex- perience. For some years, | had become increasingly dissatisfied with the sound from my fairly expensive hifi system, eventually thinking that the speakers mustn't be up to scratch. I also like listening to FM radio but I found the announcers’ voices to be muffled and the music reproduction very ordinary sounding, My wife, who suffers from several problems which affect her hearing, had an audiology test several years ago and also saw a specialist. But the downsides ofa possible surgical procedure outwoighod any potential benefits and she did not proceed down that path. At that point we invested in two pairs of wireless headphones for watching those TV programs which we had difficulty hearing (strong English dialects, background music seemingly dominatingactors’ voices etc). On the other hand, with locally produced programs, we tended to use the set’s loudspeakers with the volume fairly well advanced. Several months ago we did the on-line Blamey Saunders hearing test and as a result my wife ordered a pair of their SIE-64 hearing aids with the programmer box. These aids are not cheap but are half the cost of what you would pay for those from an audiologist supplier. After a few days use, she advised me to try her lefthand aid while ‘we watched TV. ‘The change was remarkable — the volume could be turned down and I could hear the voices with a crispness which [ didn’t realise existed! After several more days use, she fine-adjusted the aids using the programmer box, tweaking them to her satisfaction. Next, Iwas encouraged to book an audiology test at our local National Hearing Care clinic, the results of which showed my high-frequency hearing to be poor, especially on the left side. So my next move was to or- dora pair of the SIE-64 from Blamey Saunders, attachingan emailed copy ofthese results to them so they could perform an initial set-up of the aids for me. Ihave now been using the aids for several weeks and their performance is absolutely impressive. TV can be watched at considerably lower DVDs require the use of the head- phones as do some TV programs, due to accents and varying sound mixes). Listening to music via the hifi system is now a most enjoyable experience, with great treble and stereo performance from our CDs. There was nothing wrong with my speakers, after all! Asa further comment, the hearing aids do not restore one’s hearing to what it may have been some years before but assist in smoothing the defects which have occurred over time, thus enablingamore enjoyable life. “As the name implies, they are an aid, just as spectaclos aro an aid to deteriorating vision. Yos, aids all have their down sides. After all, they are objects we have to wear and hence we notice them physically. But without them we can become less connected to friends and society. They are a crutch of sorts but where would the person with a broken leg be without crutches? So as we age or suffer from a dis- ability we employ the technology that aids us to enjoy life more fully Itcan take abit of adapting to but the end result is tomy mind extremely beneficial. ‘Thanks for the magazine; some- times controversial, always inter esting. Richard Kerr, Millfield, NSW. ‘This is not what science is supposed to be about - scepticism is the basis of science. The IPCC seems to be a highly questionable organisation and as far as [ can tell it’s only them saying that 97% of scientists agree etc. There are also still problems with the whole theory, the biggest of which that I know of is the fact that over the last several thousand years, increas- ing temperature has driven increasing CO, (ie, atmospheric temperature rose bofore CO, levels), which is tho exact opposite to what the “eco-mentalists” are claiming is happening now. ‘The climate may be changing but as far as [can tell the links that “prove” it to be due to human activity (I have 10 SILICON CHP read amounting to an & production over previous levels) are far from solid. Tguess it comes down to the whole “greenie” mentality that somehow humans are a blight on nature and all ofour incredible achievements are un- natural, After all, what we are mostly doing that is supposedly wrecking the planet is simply striving to make our lives better and easier, which is surely as natural as it gets. Tove the parts of our planet that are unspoilt by humans and absolutely agree that they should be kept that way. But would I go back to living as a cave man? Absolutely not! Anyone cerned that what humans do is ‘unnatural should go back to living as a hunter-gatherer. Keep up the great work; i's a fa tastic and unique magazine, Ionwyn Buckland, Hornsby Heights, NSW. Problems with isolation transformer article Trofer to an article in the May 2014 issue of SILICON CrP, where a mains isolation transformer is assembled by ‘employing two identical step down transformers with the low-voltage windings connected togother, the idea being that the power mains is stepped down to 12VACand then stepped back up to mains potential, in this case siliconchip.com.auSS oVAC 920 watts So far so good except that the 12V windings on both transformers need to be equipped with windings capa- ble of delivering or accepting about 78A. Judging from the photograph on page 86, this would appear to be quite unlikely. Furthermore, the secondary wiring from the transformers certainly appears fartoo thin to carry such a high current. The power rating of 920W is assumed because of the 4A maximum output current stated at 230V Fig.3 on page 86 of that issue shows, 3A fuse wired in series with primary winding ofthe input transformer while the picture on page 87 shows the output GPOs labelled to deliver up to 4A. One begins to wonder how many people were involved in the article being described because the input current will exceed the output cur- rent for a number of reasons. They are primarily dueto the need for the input transformer not only requiring its own magnetising current but also needing to supply the output transformer’s magnetising current. In other words, the input current will be measurably higher than the urrent delivered to the output load. Itappears that no reduction in output voltage has been taken into account due to PR losses. The losses will also vary according to output current. It would appear that the writer of the article has taken the wattage of the transformers to be the same as the VA with an available power of Damn you, Micromite! T'm a retired engineer living in Tasmania and my ima, sparked by your article on the Micro: mite in the May 2014 issue of SILICON Cup. My discovery of the Arduino MCU development platform last yoar prompted me to take up elec tronics again as a hobby (much to the delight of parts suppliers). am more familiar with the BASIC pro- gramming language, in its various permutations but thought that the C’ language, being compiled rather than interpretive, was the way to go. Having bought a few books about the Arduino, learnt all about the IDE and taught myself to program in ‘C’, along comes the Micromite and rating because the rating of the UPS from which they were removed claims a VA rating of 920. If this is the case, the writer is wrong, VA is derived by dividing the power in wattsby the load power factor A typical monitor-computer com- bination that forms the load for a UPS exhibits a power factor of 0.6 s0 by manipulating the formula to find power, we need to multiply the VA by the power factor; for a VA of 920 and a load power factor of 0.6 we obtain a power of 522W or a current of 2.274. not 4A as suggested. I would suggest the input fuse for such an arrangement be of the order of 4A slow-blow. It should be mentioned that the lec RDP eee oa rues ed Pore eer Software Selection of Heat ce en Maun Ode Cea ceo einen eee Cec Pract EAT ETD wuvw.wittronics.com.au MMBasic. So damn you Micromite! guess the PIC32MX MCU is that fast and powerful, who cares if it’s running BASIC! However, I do believe there is ar error in the said article on page 36, where the text description of the NMEA date and time fields for the GPRMC message type are the wrong way around. ‘Date’ should be the tenth field and ‘Time’ the second. The program is correct. As an avid reader, in the past, of Practical Electron ics Australia” and “ETI”, 1am now a happy subscriber to SILICON Cit. Keep up the good work! Tony Barrett, Queenstown, Tas. Electronics” transformers employed the article are designed to operate for a relatively short time, say between five and 20 minutes, depending on the load cur rent when usinga typical 12V SLA 7A battery in a UPS. During standby, the input current only has to be sufficient tomaintain the battery voltage near full charge. Thus heating of the transformer will be minimal, whereas during the power supply mode the heating of the transformer is considerable and is not designed for continuous use Finally, I still have reservations about the current ratings of the 12V windings on the transformers (14V CT in the text!). Even with a power rating of 522W the 12V windings will needMailbag: continued Isolation transformer’s output socket must not be earthed I love your articles in “Salvage It’, Lwas brought up in the days of “make do and mend” and “waste not, want not”, Nothing useful was ever thrown away. I still adopt that attitude today although it can make access to my shed a trifle crowded. Thave a concern over the safety aspect of your design for the isola- tion transformer. In your bid to make the most use of the salvaged components you have included the double GPO. In so doing, this has completely defeated the whole point of such a device, which is to isolate from earth and all other live equipment. If two operators were to use both outlets at the same time for con- necting two pieces of equipment is no longer isolating anything, The idea of isolating transformers is to totally isolate any appliance from all connections which may accidentally become earthed. This allows a person to handle a “live” chassis without forming a return ‘This small isolation transformer the SILICON CUP laboratory has a label on top stating that the secondary must not be earthed. path to earth through their body. If two appliances are connected at the same time then a person handling one of these immediately forms a return path to the other appliance. Each appliance then becomes “live” with respect to the other and would electrocute both operators. similarly if on applionce with an earth fault were plugged into an carthed socket then the isolation is, immediately lost. For these reasons Australian Standard AS3000 clause 3.9.1 states “An isolating trans- former shall supply only one item of equipment at low voltage”. I think that it may be worth pro- ducing a safety information article for your readers who may have a need to use an isolation transformer as they may be tempted to plug in double adaptors etc. Koop the recycling ideas coming; the world needs your help. Keith Brown, Robe, SA. Comment: you are certainly correct that the output socket of an isolation ingle 3-p not be earthed. ly, we have two isolation transformers in our laboratory, one rated at 240W and one at 1kVA (a very heavy beast). We have included photos of the smaller unit to show its warning that the secondary must not be earthed. Furthermore, the earth pin of its 3-pin output socket is not connected, However, with respect to Austral- ian standards, while the reference clause 5.3.9.1 is in the 1986 edition of AS3000 it is not in AS3000:2007. However, there is mention in clause 4.14.3.2 with respect to earthing where it says that: “Exception: where the low voltage transformer output, complies with the requirements of the AS/NZS 61556 series, earthing is not required on the secondary side”. Typical commercially-made isolation transformers comply with AS/NZS 61556 2.6. The old standard was AS3108. Only one output socket should be fitted. We have modified the article pub- lished in the May 2014 issue and itis now in the on-line edition available from our website. The changes also reflect the comments made in Victor Barker's letter on pages 11 and 12. outlet and includes a warning on the to be rated at 43.5A and the wiring will need to accommodate this value. Talso make the point again that the output voltage will always be less than the input voltage. I therefore suggest that using back-to-back transformers for isolation purposes, despite the cost advantage, may not be such a good idea, Vietor G. Barker, VK2BTV, Gorokan NSW. Comment: we agree that the transform- ers featured in the May 2014 issue would not be able to provide an output of 920W. Judging from the size of the transformers’ cores, a more realistic rating would be no more than about Such an appliance would be very ‘handy when doing repairs or tests on equipment with live chassis. 12 SiicoN Crp siliconchip.com.auToo much signal from cheap video cameras Tnoticed in the Ask St.icoN Cx pages of the May 2014 issue that P.M. complained of a weak and washed out video signal from a reversing camera. I had a similai problem with both a “sugar cube’ black and white camera and a small colour camera used for security ete. At first, I thought that it was due to the fact that both cameras were cheap. But in attempting to solve the problem, Thad been referring to two “Electronics Australia” projects: the Frame Grabber in August 1989 and Slow Scan TV in July & August 1990. ‘What caught my eye was the input ofthe latter. The designer, Leon Wil- Jiams, had fed the video signal into a 1k0 potentiometer and labelled it as “contrast” So I simply placed a grounded 1kQ pot between the camera and my display, which was an Acor computer monitor with multiple inputs, including composite video. On adjustment, I was able to get a crisp, well-defined picture from either camera. Contrary to what 1 thought, they were both capable of producing good pictures. ‘The washed-out video from my cameras was duet the signals being too strong, not too weak. Perhaps that is the case for P. M. as well. George Ramsay, Holland Park, Qld. Isolation transformer should not have a double outlet ‘What a good idea to make an isola- tion transformer from the two UPS transformers, as depicted in the May 2014 issue. However, think that your readers and users of the isolation trans- former should be aware that the orig nal isolation transformers sold in this country (eg, Saf-T-Pac) had a shield around the outlet socket to prevent double adaptors being inserted and a warning that more than one item/tool/ appliance should not be used with the transformer at one time. The reason is that if one of the devices connected to it has an earth fault, itmeans that the secondary ofthe output transformer would no longer be isolated from earth, thus defeating the purpose of the device. Your version has a double outlet and users should be warned to replace that with a sin- gle general purpose outlet and also warmed against using two devices on it concurrently. Peter Chalmers, Clear Mountain, Qld. Ethernet over mains may not work across phases Thave just read Leo Simpson's arti- cle in the June 2014 issue. Itreminded me of the issues I dealt with in retro- fitting gigabit Ethernet tomy home. We now have eight cabled points and two wireloss access points to cater for all ‘manner of desktop and mobile devices. During the planning stages, I con- siliconchip.com.au sidered the Ethernet via mains option but ultimately opted for cable because I could not get a straight answer on how the ‘via mains’ system would func- tion in a house whose mains power is 3-phaso. I suspected that the ‘via mains system might only operate within one phase and not across phases. as, the three phases are separate circuits. ‘Am I correct on this and if this is worth drawing it to the attention of readers who might be considering installing this system? Robert Allan, Hunters Hill, NSW, Leo Simpson comments: Ido not have 3-phase wiring in my home so Ido not know the answer. I think your doubts are justified. Ultimately, your wired system is likely to he superior which is fine if it can be installed reasonably easily. ‘ms pulse interface information available on Microchip website With respect to the question con- cerning a ims interface in the Ask SILICON Cup pages of the May 2013 issue, the Microchip application note ANS47 covers this insome detail, Also ANS05 (Brushed DG Motors) isa good reference. Both are available at no cost from the Microchip website. Koop up the good work with Si con Crt. Mike Abrams, Capalaba, Qld. SIGNAL HOUND USB-based spectrum analyzers and RF recorders. ‘SA44B: $1,320 inc GST + Up to4.4GHz + Preamp for improved sensitivity and reduced LO leakage. + Thermometer for temperature correction and improved accuracy + _ AM/FM/SSB/CW demod + USB 2.0 interface ‘SA12B: $2,948 inc GST + Upto 12.4GHz plus all the advanced features of the SA44B + AM/FM/SSB/CW demod + USB 2.0 interface BB60A: $3,663 inc GST + Upto 6GHz + 1PPS input for GPS time- ‘stamping of recorded RF streams + Simultaneously monitor ‘two stations or stream the entire FM radio band to disc + USB 3.0 interface Vendor and Third-Party Software Available. Ideal tool for lab and test, bench use, engineering students, ham radio enthusiasts and hobbyists. Tracking generators also available, Silvertone Electronics 1/8 Fitzhardinge St Wagga Wagga NSW 2650 Ph: (02) 6931 6252 contact@silvertone.com.au Jury 2014 13This photo shows an Argo float being d@ployed into the ocea St eee nee ene shown here. The usual method is to lower them gently into the ocean in a cardboard box to protect them hitting the side of the ship. Thabox is in a sling with quick release-on the bottom. RNa eee een a es re i d the biodegradable box floats away, releasing the Right now, thousands of drones are floating deep in the oceans of the world, monitoring temperatures and other data. They are fully autonomous and they can change their buoyancy to sink deeper or rise to the surface to send data to satellites. ‘OST PEOPLE know about drone aircraft and their many types and capabilities but did you know that there are thousands of drones in the deep seas? Over 3600 such drones are quietly floating at around 1000 metr deep in the oceans of the world, mon toring temperatures, salinity and other parameters. Not only that, they also egularly descend to 2000 metres, then slowly float up to the surface, taking measurements as they goand then they beam their collected data to satellites. Afier transmitting their data they submerge again, endlessly repeating the cycle, unseen and autonomous for 14 SILICON CHIP many years, until they reach the end of their lives due to misadventure or battery failure. ‘Thisis Argo, an international project involving 30 countries including Aust- ralia. It consists of thousands of free: ranging ocean floats that monitor the temperature, salinity, currents and other parameters of the ocean, Data from the floats is used in the study of ‘oceanography and climatology Important data The data obtained from Argo is im portant because it is acqui in nearreal-time and cana: and long-term weather forecasting, monitoring of long-term trends in the ocean, monitoring of ocean currents and for other weather, climate and oceanographic research. Until recent times, ocean tempera: ture and other measurements have been made by research ships or com- mercial or military ships participat- ing in the Voluntary Observing Ship scheme. But such measurements are limited in scope and follow the main shipping routes. In addition, beca of the greater volume of shipping in the Norther Hemisphere, there was far more data from there than from the siliconchip.com.auSouthern Hemisphere, where there also happens to be a greater volume of ocean, How does it work? So how do the Argo floats sink to 2000 metres deep or rise to the surface? They do it by controlling their buoy- ancy. Fig.1 shows a cross-section of a typical Argo float; they are essentially a cylinder which is more than 1.1 metres long and they float vertically. ‘At depth, the buoyancy is controlled by anexternal hydraulic bladderat the bottom. To make it rise, a geared motor drivesa rod which pushes down a pis- ton in a cylinder filled with hydraulic fluid (oil). The hydraulic fluid inflates the bladder and the float then displaces more water, increasing its buoyancy and up it goes. ‘To reverse the process, the motor retracts the piston and the fluid from the bladder is forced back into the siliconchip.com.au cylinder, reducing the buoyancy and accordingly, the Argo float sinks. The process is quite precise as the pressure is monitored by a sensor adjacent to the bladder. As we shall see, the water depth is directly proportional to the pressure, and vice versa Extra buoyancy is required when the float reaches the surface to ensure that the antenna is clear of the water. ‘This is provided by a pneumatic blad- der which can be inflated by another pump. A typical float weighs 20-30kg, Sensors at the top of the float mor tor temperature, salinity and other parameters, depending on the particu- lar model of Argo float. An antenna at the top of the unit sends the data to a satellite. That broadly describes an Argo float but there are many varia~ tions, as described later in this article, ‘Argo was conceived in 1999 when international organisations met to discuss creating a more coordinated approach to the gathering and distribu. tion of oceanographic data. Follow this meeting, a group of scientists developed a plan to have a 3000-float array in place by 2007 and this objec tive was achieved, the first floats hav ing been deployed in late 1999. The figure of 3000 floats was arrived at by a requirement for each float to sample a roughly 3° x 3° latitude by longitude area between 60° north and 60° south. Higher latitudes were ini tially excluded because ofthe problem of the floats becoming entangled with sea ice and polar ice-sheets. There is, now a program to deploy polar floats which will be discussed later. In 2009, suggestions for further improvements to the array were made such as providing extra coverage in certain areas and adding chemical and biological sensors to the floats By November 2012, the one millionth “profile” (data set) of temperature and salinity had been gathered which represented twice as much data as had been collected by research vessels over the entire 20th century. AL the time of this one millionth profile, 120,000 profiles were being collec x every year orabout one every each profile consisting of up to 1000 temperature and salinity measurements, d jon that can be ga from the study of Argo data inclu ‘urement of ocean tempera over a range of depths. a pomp sear mato Figat: cross-section diagram of typical Argo float. Note the pneumatic bladder in this model. This is inflated near the surface to ensure the float rides high enough so that the satellite antenna is clear of the water, ‘© Moasurement of ocean salinity over arange of depths that can reveal where the ocean has become less salty due to rainfall or river outflows and more salty due to evaporation or by the flow of ocean currents with various levels of salinity. This leads to insights into the hydrological cycle. ‘© Measurement of ocean circulation and temperature characteristics which lead to phenomena such as El Nino (an abnormal band of warm water of greater than 0.5°C above average that periodically develops off the coast of South America causing adverse ‘weather events in Australia and many other countries); Decadal Oscillation ~ sea surface temperature anomalies which affect climate in ‘western North America, Siberia, In: dia and Australia; and other similar phenomena. © Accurate mapping of ocean circ lation © Seasonal variations in the ocean and long-term variations. July 204183: these two plots show data from a flo Paneer peer along the WA coastline. Fig, PURER bottom shows the correspond Of special interest to some research- ers is the heat content of the oceans. A 3-metre column of ocean water contains as much thermal energy as the entire height of the atmosphere of the same column diameter. Knowing the temperature the ocean and how heat is exchanged between the acean and the atmosphere s important for understanding the climate system nd other parameters of A typical Argo mission A typical Argo float mission is 10 days. It involves sinking from the sur- face to a depth of about 1000 metres and parking at that location for around nine anda half days while it takes tem: perature, salinity, pressure (equivalent to depth) and other measurements the float is equipped to take; see Fig.8. 16 ‘ON CHIP Peri ae Sores eens A depth of around 1000 metres is typically chosen as it is usually a region with minimal current and the float will not drift away too far from its desired location. Following the park- ing period, the float drops toa depth of about 2000 metres and then proceeds to rise to the surface over a period of eight hours during which it takes, further temperature, salinity, pressure and possibly other measurements along the approximately 2000-metre water column, depending on which sensors the float is equipped with. Pressure equals depth Note that in oceanography, water pressure, measured in decibars (dbar or db), is used as a measure of depth {inmetres), Onebar roughly equals one atmosphere and a decibar is roughly Fig.a: this is the path of an Argo the Antarctic Circumpolar joat had been deployed for six years, repoi 2000 metre profile every 10 days while drifting at a depth of 1000 metres between reports. in Current that runs Pern eo tery 0.1 atmospheres. The pressure in deci- bars is for most practical purposes the sameas the depth in metres, so that an increase in depth of one metre equates (0 increase in pressure of one decibar 100 decibars is 100 metres. While pres sure in the ocean would comprise the depth of water plus the atmosphere, the relatively small contribution of the atmospher surface, the pressure is considered to be 0 decibars. The precise conversion formula between decibars and metres of depth later in this article. is ignored so at the Argo floats can phone home Argo floats communicate by one of two methods. Older floats typically communicate to the Argos satellite which is a general-purpose environ- mental data receiving satellite specifically associated with the Argo program despite the similar name. Newer floats use the Iridium satellite phone network. Essentially, they make a phone call to the relevant Argo data centre Older floats which communicate with the Argos satellite have to sit on the surface for 12-26 hours in order to transmit their 78 data points to the satellite. They can only store one pro- file ata time. These long surface times mean that wind and surface currents can move the floats away from their intended location and they can even Another risk of long surface times is that they will be spotted by fishermen and picked up when they should be left alone. This is a major reason for Argo floats, particularly in the tropics siliconchip.com.auFig.5: this is a Belt’ Oceans where i if any changes take place. sometimes ending up in remote fish- ing villages in the middle of nowhere! Since the older floats don't have GPS, their location is determined by calculations involving Doppler shift of the radio signal, Newer floats which communic via the Iridium network only requirea surface time of around 15 minutes and can store up to 1000 data points per profile and 60 profiles. Their location is determined by GPS. One might wonder if the flo constitute a shipping hazard but ther have been no incidents. Their tim at the surface is relatively short and since they are generally far away from shore they are not likely to be hit by small speedboats. In any case, there is vastly more natural and man-made debris floating in the ocean, much of it larger than the floats. Australia is a big player ‘The USA has the largest number of Argo floats while Australia has the second largest, representing about 11 percent of the total number (see Fig.6). Argo in Australia is operated by CSIRO Marine and Atmospheric Research in Hobart, with support from the Bureau of Meteorology, IMOS (Australia’s Integrated Marine Observing System), the Antarctic Climate and Ecosystem siliconchip.com.au general model of oceanic circulation, also known as thermohaline circulation or the “Global Conveyor Its driven by differences in water temperature and salinity which affect the density of seawater. In general, warm shallow water cools and sinks in the North Atlantic and deep cold water returns to the surface i again warms. Argo can help monitor these currents, measuring temperature and sali the Indi Jason & jhe Argonauts ‘The name Argo derives from Greek mythology and is the name of the vessel in which Jason and the Argonauts went looking for the Golden Fleece. Argo also works in a complementary manner with the NASA Jason satellites to measure sea levels, Jason provides extremely accurate measurements of the sea level (to a few centimetres with complimentary gravity data from the NASA GRACE mission), while Argo provides measurement of salinity and temperature. This gives the contribution of water density (derived from temperature and salinity) to sea level which helps both validate Jason satellite data and also helps determine the contribution of sea level due to changes in the density of the water as opposed to extra water mass being added to (or removed from) the oceans such as that due to melting (or formation) of land-based ice. Cooperative Research Contre, the Royal Australian Navy and the Depart- ment of Climate Change and Energy Efficiency. Worldwide, the Argo program is sponsored by the World Climate Research Programme's Climate Vari- ability and Predictability project (CLI- VAR) and by the Global Ocean Data Assimilation Experiment (GODAE). It is a pilot project of the Global Ocean Observing System (GOOS). ‘There are about six major manufac- turers of floats plus some minor ones. The Argo program does not specify the exact design of each float but does specify required performance data such as accuracy, type of sensors and float and battery lif ince the exact specifications are not defined it allows manufacturers to come up with better, more efficient and more capable designs and also allows float costs to be reduced. A typical float costs around $21,400 although the total deployed cost including the cost of the float, a ship to deploy the float and staff is around $35,000. ‘Argo floats used in Australia are disassembled and undergo a thor- ough check before deployment and older models had their alkaline battery packs replaced with lithium ones. Asa result ofthese pre-deployment checks, JuLy 201417—— nme Fig.6: this diagram shows the global distribution of floats and country of origin. The US has the highest number of floats (2000) w! le Australia has the second highest with 386. France has the third highest number of floats, with 256. Note that these are representative locations for a certain point in time only as the floats do drift around. Argo floats in the Australian fleet have very good longevity. ‘The lifetime of older floats was manufacturerrated at 3.5-4.5 years Fig.7: Argo data in its most basic form, showing a plot of temperature and salinity versus pressure (depth in metres) for a given position in the 18 Siicon Cu but due to the battery upgrade they have lasted up to 10 years. Currently deployed floats have a typical life- time of 7-8 years because of a more complex mission profile and more measurements being taken, resulting ina reduced battery life. This lifetime refers to time out in the ocean before the batteries go flat as the floats are not usually retrieved, ‘When the battery fails, the float is usually unable to rise from its ap- proximate 2000 metre depth as there is insufficient battery capacity to re- inflate the buoyancy bladder. There it will remain indefinitely, never to be retrieved. Note that while failed Argo floats are not usually retrieved due to the diffi- culty and expense of doing so (which would exceed the value of the float), if it comes to the attention of the Argo organisation that one has come ashore in an inhabited area or has actually fallen into someone’s possession, it is important that it is recovered. This is because the large lithium-ion bat- tery pack could be a safety hazard in the wrong hands, especially if treated inappropriately. In addition, if traces of water have You can make your own world globe in the form of an icosahedron showing the location of Argo drifters for a given day, The image can be found at: http:// projects/argo/images/icosa.tif and assembly instructions can be found at: http:/Awww.pac.dfo-mpo.ge.cal science/oceans/Argo/documents/ Argo icos.pdt entered the float, there could be an explosive and toxic mix of hydrogen, oxygen and chlorine gas, due to el trolysis of seawater by the battery. Sensor accuracy Since large amounts of scientific data are derived from the floats and that data is further incorporated into climate, oceanographic and other models, itis extremely important that the float sensor data be as accurate as possible. Temperature accuracy is +£0,002°C, salinity is within 0.02 parts per thousand and pressure is wit 2.4 decibars. ‘This is a very high level of accuracy siliconchip.com.auso scientists can have great confidence in the results. In regards to the missing Malaysian Airways flight MH370, the floats obvi ously have no capability to directly locaie the wreckage. However, data from the floats feeds into and is the major contributor to the ocean currer models that were used to track and predict the possible location of the crash del ‘The most energy consuming process in the floats is changing the buoyancy to make the float rise or fall. Forcing hydraulic fluid into an external blad- der at a depth equivalent to 2000 dec- ibars, oraround 1975 metres (ie, where Argo descends), requires a significant amount of energy. At that depth the pressure is around 200 atmospheres (20 megapascals or 2900 psi). Note that the exact depth where 2000 decibars occurs varies slightly according to the latitude. It equates to 1971.7 metresat 60° degrees latitude, 1976.1 metres at #35° degrees latit and 1979.55 metres at the equator. Accessing the data Anyone, including Siticon Cui readers, can access the Argo data for free and make their own discoveries. A website at http://wo,jcommops.org/ gicbin/WebObjects/Argo has gateways to the two global data centres and also other information. The US Global Data Center (the other is French) can be accessed at http:/Avww.usgodae.org/ argo/argo.html Data for the Australian Argo array can also be seen at: www.cmar.csiro. awargo/tech/index.html Occasionally, Argo floats wash up on beaches or are otherwise found and it can involve some real detective work to track them down. The main reason for the ‘needto recover such floatsis thatthe large lithium battery inside them can bea safety hazard in the wrong hands. In one case, a float was deployed by the San Diego-based Scripps Institution ‘of Oceanography near New Caledonia. It {alledto surlace after about 18 months and its last known location was off the coast of Mooloolaba in Queensland. {twas then actually rawled by a fisher- ‘man who thought it would be a good idea all data is Float tak Once the data is recelved itis sent tothe closest research station for analysis ett toa satelite Se \ 1 once at the surface, @ Ship drops he floatinthe ocean. transmitted we. dings eet) re Fig.8: a typical Argo float mission. The float descends first to 1000m and then to. 2000m, switches on its sensors and then floats to the surface so that the collected data can be transmitted to a satellite. The satellite data is then downloaded to a ground station. Interestingly, in March 2013 the data centres were hit with a huge number of downloads involving computers from all over the world and hundreds of gigabytes of data. The reason was a ‘mystery until it was discovered that it corresponded to the film “Argo” being given three Academy Awards and peo- ple were looking for free downloads to tum itinto a letterbox but it was spared that fate due tothe intervention of CSIRO scientist Dr Ann Thresher who isin charge ‘of Argo Operations. (Once the float was brought to the sur- face by the fisherman, itstarted broadcast- ing its location again. The precise location couldbe determined to only about.a block and Dr Thresher travelled from Hobart to find it. She intially failed to do so and decided to return home but then changed her mind, more determined than ever to recover it She went tothe yacht club and then the fishing boats and after showing a picture of the device eventually found of the movie. Naturally, they would only have downloaded raw Argo data, not video. Recent developments ‘The Argo platform is very flexible and as noted above, is not strictly defined in terms of shape etc. Thi allows floats to be developed with a This photo shows the sensor head of the float recovered off Queensland. After spending 18 months on the sea floor, it was fouled with barnacles. the fishing boat crew that had retrieved it. The device was then collected and returned to the CSIRO for examination. siliconchip.com.au Jury 201419Fig: ow the Argo floats cope with surface ice. The float will only rise to the surface to transmit data if the surface is ice-free, otherwise the data is stored until a break in the ice is detected. Contrary to what is shown in this diagram, in the ‘current operational scheme, if there is overhead ice detected the float descends again to about 1000 metres and coi its 10-day mission eycl wide variety of sensors tosuit different applications. Newer floats may contain oxygen sensors, transmissometers to measure water turbidity (a measure of the biological productivity of wa: ter), an FLBB device (fluorometer! backscatter combination sensor) for chlorophyll measurement and meas- urement of nitrates, and a variety of other sensors. In Australia, these new “Bio Argo” floats will be deployed this year in places such as the Bay of Bengal, as part of an Australia-India collabora- tion and off the north-west coast of Western Australia, These floats will mainly work at a depth of 300 metres. Incidentally, some of the more restric- tive countries of the world will not allow Argo Moats that collect biologi cal data into their oceanic territories, presumably since it has implic for fishing policies ete. A particularly interesting sensor has been developed that measures the elec- tricfield produced when (conducting) seawater currents move through the Earth’s magnetic field. This is usually called motional induction. It allows the direction and speed of ocean cur rents to be determined. The specific type of Argo float that carries this sensor package is called the EM-APEX or ElectroMagnetic Autonomous Pro- filing Explorer. ‘The float contains a compass, ac- celerometers, magnetometers and a processing system to convert voltage differences measured by sensor elec- trodes to velocity components of the ‘ocean current. This float also measures salinity, temperature and pressure, as do the other floats. Coping with ice Looking into the future,a number of Converting Pressure)to Depth Based on UNESCO Technical Papersin Marine Science No. 44, gravity ata specific latitude and pressure is given by the following empirical, computationally-fiendly equations: ‘a(tw/sec?) = 9.780318 [1.04+(5.2788 + 103+2.36+ 105*x)+x]+1.092+ 108*p where in (latitude ~ 57.29578) F° and p = pressure (decibars) Depth is calculated from pressure as follows: depth (metres) = [((-4.82 * 105 « p+ 2.279 + 10") + p- 2.2512 + 105) + p+ 9.72659) + p] +9 where p = pressure (decibars) and g = gravity (msec?) ‘These formulae assume a certain water temperature and salinity In reality, the aitference between depth in metres and decbars isso small as to be of litte practical significance. 20 SILICON CHIP ‘As the floats used in such areas can store a large number of profiles, they can make many attempts to surface (at intervals of 10 days) until success is achieved. new varieties of Argo are envisaged. Bio Argo has already been mentioned. Asstated, the initial deployment mod el for Argo excluded the high-latitude regions because of the possibility of entanglement and destruction in the sea ice. These issues have now been resolved with new ice-hardened floats features such as antennas that are resistant to ice and also methods of detecting overhead ice. Overhead ice can be avoided by the float sensing a temperature close to the surface consistent with sea ice and then descending again if ice is expected. The float can stay submerged for a long time if necessary as numer- ‘ous data sets can be stored and then transmitted to the Iridium satellites. Australia has deployed 29 floats in the seasonal ice region of Australia’s section of the Southern Ocean. Other future planned developments include a total fleet of 4500 floats and deep-profiling floats that go to 4500 or 6000 metres. Argo is providing unprecedented amounts of information about the ocean environment, It is a major part ofthe world’s ocean observing system. Among many other things, it should increase the power of predictive mod- tls of short-term and long-term climate forecasting, patterns of ocean currents and present and future trends in the global climate, as well as provide i formation on the interaction of both the shallow and deeper ocean with the atmosphere. ‘New developments also allow moni- toring of the biological productivity siliconchip.com.auThe Argos System Satellites LDER AUSTRALIAN Argo floats transmit their data via the Argos System satelites. While the names are similar, theres no direct relationship be- tween the two programs, apart from the fact that Argo uses the general purpose ‘Argos satelite system. These satelites ‘are designed to receive and disseminate data of a primarily environmental nature from both fixed and mobile platforms ‘around the world, Applications include but are not lim ited to: Tracking land and marine wildlife such as sea turtles, fish, birds and land animals fitted with miniature transmitters, Receiving environmental data from fixed and floating marine platforms (manned and unmanned); Monitoring of disease outbreaks, food shortages, therapeutic drug availabilty ‘and humanitarian aid resource utilisation in Third World countries (via aid-worker ‘mobile data terminals). This data is rel ‘evant to public health and aid authorties, and the system can even monitor schoo! attendance rates; Monitoring the climate via Argo and ‘many other floats and buoys; Monitoring of global water resources such as river levels, snow fall, dams and the status of water distribution infrastructure; Monitoring fishing vessels via trans- rmitters installed on them to ensure com- pliance with national and international fishing agreements; ‘Tracking of adventurers in extreme environments and international yacht races; Improving maritime security by alow- ing shipping operators to keep constant track of their fleets, with all ships of over ‘500 tonnes gross being required by the of the ocean which might lead to new sources of sustainable fishing and other marine food sources (and may also indicate where these resources siliconchip.com.au eee Nes Gere Leta Satelite Fig.10: di eee eS Cea heme Arey ss COE ne \gram showing the direction of Doppler shift as an Argos satellite approaches and then retreats from a transmitter. International Maritime Organisation (IMO) to have a Ship Security Alert System (SSAS) installed ‘Argos satellites are able to receive location data trom GPS equipped trans- mitters but can also compute position data from platforms not equipped with GPS by utilising the Doppler shift of several received signals over a period of time. This is the same technique by \hich the rough locaton ofthe missing Malaysian Airlines. Fight MH370 was determined. In practice, locations can be determined with an accuracy of 150 ‘metres using Doppler shift as opposed toa few metres with GPS. In Doppler location, the Argos satel- lite records the precise frequency of the received signal for every message received. Several messages need to be received in order to obtain a positional fix in order to generate a Doppler shift ‘rofl’ of how the frequency changes asthe satelite frst approaches andthen recedes from the transmitter. are being depleted, to give fisheries a rest), Other benefits of Argo are that it fosters international collaboration and helps in the development of global ‘At some point in the frequency versus time profile theres an inflection point rep- resenting the true transmitter frequency. ‘The orbit of the satelite is known, as is the altitude of the transmitter, leaving the latitude, longitude and the true transmis- sion frequency ofthe signal unknown for ‘each transmission. These unknowns can be determined with two or three mes- ‘sages but a fourth message is required to completely solve the equations and determine the positional accuracy. The solution to the equations provides ‘wo possible locations and then plausibil- ity tests are usedto determine the actual location as one solution will most ikely represent an unrealistic position of the platform. ‘The latest Argos-3 satelites represent a significant improvement over previous versions and have 2-way communication, better transmission management (eg, acknowledgement that data was correct- ly received) and the possibilty of platform remote control and programming. environmental information databases. Itis widely supported internationally, Australia is a major player and the future looks very bright. sc Jury 2014 2aa boon for working with surface-mount devices How you do cope with assembling PCBs with surface-mount devices? You can always inspect the solder joints with a magnifying glass after they have been made. Or perhaps you use a large illuminated magnifying glass while you solder. But the “deluxe” way is to use a stereo microscope. hese days it is more or less im- I possible to avoid surface-mount technology if you are involved in building or repairing electronics. Almost all the really interesting de- vices seem to be available only in fine- pitch or leadless packages. Rather than fight this trend, I have been working quite happily with SMT technology 22 SiLicoN CHIP for many years and have developed an armoury of tools and techniques to cope with most of the packages those devious chip designers can dream up. One of these tools was a very cheap, hand-held USB microscope, ‘This was handy for inspec orlooking for solder bridges but it was not practical to use while soldering since the working distance (the tance between the work piece and the lens) was only around 20mm and there was a considerable lag in the video. Thus, [tended to do most of my SMT work without magnification. joints siliconchip.com.auViewing your work through the eyepie the same place takes a become second nat Recently, after spending an after noon hunched over the bench with my face only centimetres from a PCB, the rick in my neck told me something had to be done. I hit the ‘net to see what the options were T steered away from video micro scopes because I was wary ofthe video lag effect. I therefore narrowed. my search down to optical stereo micro scopes with a wide field of view and a good working distance. Incidentally, I should point out the difference between a compound microscope (the type you m used in science at school) and a stereo microscope. A compound microscope may have one or two eyepieces but hasa single objective lens close to the object being examined. Often there are several objective lenses mounted on a turret but only one is in the optical path any time. Compound microscopes offer high magnifications (up to 1600x) but the image is two-dimensional. A stereo microscope, on the other hand, has two completely separate ht have siliconchip.com.au it of getting used to - but after a couple of hours, it does ie. The microscope is perfect for surface-mount devices, optical paths offering slightly different views, resulting ina three dimensional image. Magnification is usually lower than for compound microscopes (less than 100%) What's available? There is a wide variety of options out there, ranging from the top-of the-line Mantis Elite at nearly $3,000 microscopes under to low cost “toy $100. ‘After much searching I settled for a microscope from the US-based online retailer AmScope (www.amscope. com). They offer a truly bewilderin, array of microscope types and configu: rations, Ultimately I chose their SM- 4TX-144A microscope package. See www.amscope.com/sm-atx-144a, html This is a trinocular microscope: as well as two eyepieces, it has a third optical port suitable for mounting a camera, It has a magnification ranging from 3.5x to 45x, continuously vari able via a zoom knob, a wide field of m and a working distance view at 65 of 200mm. It comes with a boom stand that al- lows the microscope to be moved in and out of the workspace and a LED ring light to illuminate the work sur face, At about $US550 this seemed to be a very reasonable package. I went onto the website and placed my order, only to discover the shipping would be around $US360, bringing the total to almost $1000 Aussie dollars right at the top end of what I was prepared to spend. Given that the package weighs over 30kg, there were no cheaper shipping options. Iswallowed hard, entered my credit card details, hit the go button and sat back to wait. But a few days later I got an email from AmScope explaining that they did not accept foreign credit cards, aski account. No problem here but it seems an inefficient way to operate. They also stated: “all of our items come standard 110V, if you need 220V please let us know as there is a $5 foo for the up grade”. Again, no big issue but the website could be a bit more friendly to overseas purchasers. A few weeks later two parcels duly arrived, one containing the micro scope head, the eye pieces, Barlow lens, eye guards, a plastic dust cover: the ring lig second box {which I could barely lift) contained the parts for the boom stand ‘This is truly a massive (30kg) piece of hardware. zme to pay via their PayPal t and its control box. The Caveat emptor! By the way, there is an enormous price range for the AmScope SM-4TX- 144A on the internet — and we're not sure why Some sources offer this microscope for about the same net price as Am Scope or even less: after ordering (of The eyepieces have a rubber shroud which can be extended, as shown collapsed for those of us who lly aided (ie, wear glasses!) Jury 201823Included in the package is this ring LED worklight, absolutely essential for illuminating the subject, along with its power supply/controller. We have just a few reservations about the quality ~ but course!) we found ot for SUS520 with SUS® Australia (-SAU600), Butthere are others selling for much more (eg, >$AU1350 ine shipping) with no apparent “extras”, You can order this scope through amazon.com (“sold and shipped by AmScope”) which suggests a 'SUS499.98 (SAU597) including post- ‘age. But when you go through the cart process to order one, you're likely to find that they either “will not supply to your geographic area” or the “free postage” suddenly becomes not quite- so-fre ‘That same site adds a “per ship- ment” charge of $US29.00 and a “per weight” charge of SUS8.99 per pound — and the AmScope website states a shipping weight of 68 pounds. So that free postage could equate to SUS611 or more than $AU650! So if you order on line, shop around, check the fine print and watch forany emails from the supplier. In the end, I was happy that I or- dered it from AmScope direct (and the security that provides) even if I did pay a premium. site selling shipping to Assembly . . . instructions? The assombly instructions are pretty rudimentary and Ifound it easier to use photograph from the website rather than the provided booklot to assemble the stand, Otherwise, assembling the micro- scope head is fairly foolproof with 24 SiicON Gur works! the two eyepieces dropping in place and the Barlow lens screwing onto the objective lens at the bottom. The ring light is attached by three radial thumbscrews toa supplied collar that is screwed on to the Barlow lens. Incidentally, the Barlow lens is used to increase the working distance at the expense of magnification. Inthe case of my microscope, the Barlow lens dou- bles the working distance and halves the magnification. If itis removed, the maximum magnification increase to 90x but the working distance is only 100m ‘The ring light is worthy of comment as it contains no less than 144 LEDs arranged in concentricrings. The con- lows the LEDs to be di o for separate control of quadrant. This is handy if you want to light from one side only (for example to read those pesky black-on-black IG markings). ‘The ring light and its control box are shown above. Note the (mis)use ‘ofa USB connector and the US mains cable. The photo opposite shows two of the four segments illuminated. Ono surprise was that my five-dollar -220V upgrade” consisted of an unap- proved Chinese power adaptor 80 1 could plugthe supplied US-style mains plug into an Auss Pretty dumb really, since the mains cable connects to the light control box via a shaverstyle cord. It would have been better (and safer) if they had simply packed an Australian cable, ‘The ring light control box is also a bit disappointing and has a bit of ‘a home-made feel. It works fine but the buttons feel cheap and the ring light connector is a USB mini B type. I don’t like the idea of using these connectors for non-USB purposes in commercial gear. One day some “valued customer” will pluga USB device into the control box or the ring light into a USB device with ugly consequences. The control box does havea CE mark label (although no US or Australian approval marks), so I assume it is reasonably well designed and made. Using It Apart from abit of playing about, the first intensive use of the microscope was to construct the prototypes for the Audio Recorder published in the May and June 2014 issues of Siticon Ci ‘This required five or six hours of in- tensive SMT work including soldering 0.5mm pitch TQFP microcontrollers and a.3mm x 3mm leadless pack, also with 0.5mm pitch, Although it took a few minutes to get used to the idea of looking more orless straight ahead, while my hands were working down at bench level, | quickly became a complete convert to working with a microscope. Formost work, [kept the magnifica- tion quite modest, with the greatest benefit being improved posture and zero eyestrain. I think the consistency ofmy solder joints wasalso improved, ‘The microscope really came into its ‘own when checking for solder bridges between pins orto check that a pin had properly reflowed By increasing the magnification, those tiny 0.25mm gaps between pins became yawning chasms and any sol- doring faults were rendered blinding! obvious. I was even able to tilt the ope 45° to look at the edges of the leadless packages to see if the solder had reflowed properly. As mentioned above, the ring light plenty of modes but in practice Tused it with all segments on and at close to full brightness almost all the time. The ring light is essential since without it the microscope image almost unusable, despite my having a very bright lab. I found the working distance to be completely adequate and never felt constrained by the presence of the microscope while using the soldering siliconchip.com.auHere's that same LED. iron or any of my usual tools. The field of view was also fine but there was an unusual side effect of viewing the work through what is effectively a 65mm diameter hole. I would tend not to lift my eyes from the eyepiec iron, for example, and then fail to be able to bring it to the “hole The answerof course, isto only look through the microscope once your tools are nearly in position, but this took a little practice. One other pleasing feature com- pared to the old USB microscope used to use was the true three-dimensional nature of the image. You can really see the way the solder fillet runs up the end of a SMT capacitor, for exam- ple. The depth of field is adequate at about 5 or 6mm for typical levels of magnification. Once you start using the micro: scope, you really begin to appreciate that heavy stand that cost so much to ship. The horizontal movementis very smooth and the solid construction and heavy counterweights mean that the microscope remains rock solid even at the fullest extension. Thismeans you can push the micro- scope right out of the way when you don’t need it, but pull it into place in a couple of seconds I wear spectacles all the time and I was a little concerned how these would work with the eyepieces since Thave had trouble with microscopes and telescopes before. I need not to grab the soldering siliconchip.com.au nglight in action. It simply clips to the bottom of the Barlow lens (also supplied) and can be varied in br segments illuminated. We tend to use it with all four on and flat out! 198 as well as in the have worried, since the rubber eye guards (in their folded-down position) worked perfectly with my glasses. These eye guards also fold up for those lucky enough to have perfect vision. Conclusion On balance, the AmScope SM-4TX- 144A is a great tool for anyone regu- larly using surface mount components —and that’s most of us these days. It is easy to use and will in all probability significantly improve the quality of the work you can do, and give you the ability to tackle the very fine pitch, and miniscule components that seem to be the norm these days. I think that the manufacturer could easily do a little more to make things easier for non-US purchasers, by im: proving the website, fixing the “220V upgrade” and most importantly opti mising the freight If T have a criticism it is the poor documentation and the barely ad- equate quality of the ring light power supply. At around $1000 delivered, it is an investment on par with an entry-level, oscilloscope ora top quality soldering station, so not one to be made lightly Nevertheless, I am a complete con- vert, and would recommend a stereo microscope if you are doing a lot of SMT work. For further information, go to the Amscope websit ‘om se leadless, at_www.amscope OSEAN, Tonnes eee N480D PID Controller Low-cost PID controfer thet ACEeDIS T/C RIRTD pt 3 i foreriving SSRs. Auto-tuning Pib funtion, dual display & Famps goak fnction. 240 VAC KU: CET-052 Price:$79+6ST Wireless Freezer Thermometer FT0076 fe an 8 chonne! Wireless thermometer Sultabie fo" mentoring Femperature of freezers Tehas fem accuracy of 18C. Comes with 2 remotes & up to 8 Sensors cam be added, plus alarm configu Fatlon_ 100 rm transmistion range. Batteries ob include Price:$85+GST Limit Switch Roller adjustable lever type limit Switch comes in a solid fPs7 die fast cast mith LED operation Indicator. eis designed to for high mechanical intensity envi- fonments where itis exposed to heat oll & dust. 24 VDC powered, SKU: HNR-402 Brice:$49.95+GST that tures on fignt when motion ie detected, it features @ Selocta bie metody tones that ae ated b {0.90 dB. 24 VOC/AC powerea, SKU: HNL-101 "Price:$89,5046ST Se 4 Channel Temp Logger Serial port powered, tongs” tbat as 9 Continuous eta seam overan Roose Serial port Supports up to wr DetaE26 Sensor ony 1 iBincluaed re sorvare Beovided for logging ata to a Gey hte SKU! KTA-145 ice:$25+GST Ethernet Serial Server Gonnects an 85-222, 8S. 422 oF RSWA85 serial signal tan ethemet network 20/100 Nps ethernet au Ported. SeS001 offers full Guplex and bi-directional Gata that's transmission Eransparent betwoen the serial port and the Ethernet network. DE-9 serial port. 5 0 30, VBC powered. SKU! ATO-161 Price:$129+6ST GSM Wireless Home Alarm Kit alarm kitwith base unt, m Wireless PIR onc door sén- oP Sor strobe siren, 12 VOC Plugpack aug ovo keyfob \ ge fom notifications 0 sen Lal Gresereuie stone See keRa00”°price:s2494csr gee Aeon ea Ce Ree Ph: (03) 9782 5882 Ceara) Jury 201425A sitiple but wersatile device to smiten a rela NPN woltAgEe Crosses A Preset threshold This versatile Threshold Voltage Switch takes the output of an analog sensor, battery voltage or other varying voltage and switches power to a fan, warning light or similar when a preset threshold voltage is reached. It can be set up for use with a 5V, 12V or 24V supply. It can also be used to prevent a lead-acid battery from being over-charged. can handle a relatively high current. switch one or the other, depending on value or wk value. This is PCB. ed by links on the Circuit description Fig.2 shows the complete circutt for It comprises two ICs, a 3-terminal regulator, the relay and a few other ‘components. mp [Cla is wired as voltag ator to monitor the input fect 1s tho samo but it may bo more convenient or even a requirement to circuit or whatever you fancyBy JOHN CLARKE inverting input, while the threshold comparison voltage is fed to pin 3. If the required voltage threshold is above 3.3V, you will need to attenu- ate the input voltage and this is done by inserting link LK1. The amount of attenuation is then adjusted with ‘multi-turn trimpot VR1 A3.3V reference voltage is provided by REGI, an LM2936 low quiescent current, low-dropout regulator which is fed from the V+ supply rail. It feeds trimpot VR3 and in tur, its wiper voltage is fed to IC1b which acts asa low impedance buffer to provide the reference voltage to pin 3 of ICla. ‘Trimpot VR2 adjusts the hysteresis, of comparator ICia, Hysteresis can be regarded as positive feedback and it reduces the sensitivity of the com- parator to short term variations in the input voltage. ‘Toexplain further, say the threshold voltage at pin 3 is 3V and the sensed voltage at pin 2 goes slightly above 3V, resulting in the comparator’s output going low. The feedback connectior from output pin 1 to pin 3 means that siliconchip.com.au Main Features © Operates from 5-24V DC (nominal, SOV maximum) ‘© Adjustable trigger threshold ‘© Tigger on high or iow vottage ‘© Output state indicator LEDs ‘© Mutiple relay options, up to OA SPDT or 104 DPDT Power Supply: 5-30V, Current Drain: <1mA with indicator LEDS off (LK out), relay off and VR2 set to >100K0. With the relay on, the current is dependent on the coil resistance, Signal Input Impedance: 470k0 minimum. Trigger Threshold: adjustable. Input Divider: divide by 1 (LKt out) or divide by greater than 5.7 (LK in). Hysteresis For No Input Attenuation: ~SmV-2.5V for SV supply; ~12mV.6V for 12V supply. Hysteresis For 10:1 Input Attenuation: ~50mV minimum for SV supply; ~120mV minimum for 12V supply. ‘Maximum Switching Voltage: Contact ratings for off-board relay. the voltage at pin 3 is pulled slightly lower than it was before pin 1 flicked low. That means that the sensed volt- age at pin 2 will have to drop some- what below 3V to cause the comparator to change state again. So the output will not switch again immediately if there is only aslight drop in the voltage at pin 2 immediately after the output switches, Conversely, when IC1a’s output goes high (near V+) in response to a drop- ping voltage at pin 2 of ICta, pin 3 is instead pulled higher than before and pin 2 will have to rise by an increased amount to switch the comparator’s output low again. So the threshold voltage for ICla varies depending on the output of ICta, In practical terms, hysteresis pro- vents the relay from ‘chattering’ on and off when the sensed voltage is close to the voltage threshold. It also stops the + top Fig. 30V DC/40VAC for on-board relay; limited by circuit from switching on and offevery few seconds. Say for example, you want fan to cool a heatsink whenever the temperature reaches 60°C. As the temperature sensor reaches 60°C, the fan will run and almost immediately the temperature will drop by a small amount, ‘This means that, without hysteresis, the fan might run for a less than a second before switching off and then a second or so later, it will be on again as the 60°C threshold is reached. By adding hysteresis, the fan can be sot to start running at 60°C but only switch off at say 55°C. That way, the fan will run for longer, preventing rapid on and off cycling. When setting the threshold voltage for IC1a, we monitor test point TP2 This actually allows us to set the two switching thresholds: one when IC1a’s output is high and the second when its he relay can switch either the positive or negative supply lead to the load. Fig.1A shows the load’s ground connection being switched while Fig-1B shows the relay contacts in the positive supply lead. Joly 204271 double-sided PCB with plated-through holes, code 99106141, 107 x 61mm 1 UBS plastic utility case, 130 x 68 x 44mm (optional) 412V DPDT relay (RELAY1) (Altronios 8A $4190D or low- profile $4270A, Jaycar 5A SY-4052)" 2 2-way PCB-mount screw term- inals, 5.08mm spacing (CON1) 2.3-way PCB-mount screw term- inals, 5.08mm spacing (CON2)# 2 8-pin DIL IC sockets (optional) 5 2-way SIL pin headers with 2.4mm pin spacings (LK1, Lk2, LK4, LKSa & LKSb) 1 3-way SIL pin header with 2.54mm pin spacing (LK3) 66 jumper shunts (shorting blocks) 3 PC stakes (TP GND,TP1,TP2) 41M x 6mm machine screw & nut Semiconductors 1 LMC6482AIN dual CMOS op amp (IC1) 1 7555 CMOS timer (IC2) 1.BC337 NPN transistor (Q1) 4BC327 PNP transistor (Q2) 41 IRF540 N-channel Mosfet (Q3) 4 LM2936-3.3 3.3V regulator (REG!) (Jaycar ZV1650) 2 1N4148 small signal diodes (01,02) 1 1N4004 1A diode (D3) 1 1N5819 Schottky diode (D4) 1 1N4744 15V zener diode (2D1) (two required for 24V supply) 2.3mm red LEDs (LED1,LED2) 1.3mm green LEDs (LED3) 2 100k02 25-turn trimpots. (VR1,VR3) 1 1MQ.25-turn trimpot (VR2) Capacitors 1 100uF 16V radial electrolytic 1 22uF 16V radial electrolytic 1 1nF 16V radial electrotytic, 5 100nF MKT Resistors (0.25W, 1%) 1470k0 3.3.3k0 (0.5W) 2 1ko. 2 1000" Plus Rit (5W) if required (see text) Notes * see text and Table 1 for other relay options. # not required if an off-board relay is used; two PCB-mount vertical spade connectors plus matching crimp connectors are required instead. For 5V supply, delete 1 x 1002 resistor and add 1 x 100; for 24V supply use 2082 0.5W. output is low. The threshold measure- ‘ment is made between test points TP2 and TP GND. IC2, a CMOS 7555 timer, is used as an inverter and its pin 3 output goes to one side of 3-way header LK3. De- pending on how link LK3 is set, the gate drive for Mosfet Q3 can come from either pin 1 of ICta or pin 3 of IC2, This means that the relay can be turned on when the input voltage ex- ceeds the threshold (LK3 set to HIGH) or when the input voltage goes below the threshold (LK3 to LOW). As shown, the HIGH setting selects the output from IC2 while the low set- ting selects IG1a’s output. The selected output then drives Mosfet Q3 via a 10022 gate resistor, When Q3's gate goes, high, Q3 turns on and powers the relay coil. LEDS (green) is also lit whenever Q3 is switched on. Note that although Q3 isn’t a logi level Mosfet, it’s suitable for use with 28 SILICON CHIP ‘ 5V supply which results in the Mos- fot gate drive being less than 5V. We have specified an IRF540 Mosfet for this reason ~ it doesn’t need to fully saturate as it’s only switching a small current (the relay coil current). Output indication LED1 and LED2 are used to indicate IC2’s output level and are selected by links LK5a & LKSb. They simply indi- cate whether the input signal is above or below the threshold voltage. LED1 is driven by NPN emitter-follower transistor Qi while LED? is driven by PNP emitter-follower Q2. In operation, LED1 lights when the input is greater than the threshold, while LED? lights when the input is less than the threshold. After setting up the threshold adjustments, the two LK5 jumper shunts can be removed so that these LEDs no longer light ‘This reduces the current drain of the circuit which can be useful in situ- ations where current drain must be ‘minimised, Supply voltage The circuit can be operated from supply voltages ranging from 5-30V. Most of the circuit is fed via Schottky diode D4 while the relay is directly powered from the input supply. D4 is included for reverse polarity protection. It’s followed by a 1002 resistor (R2), while zener diode ZD1 is included to clamp the supply to 15V. ZD2 is used to drop the supply by 15V when a 2aV supply is connected while LK2 is used to short ZD2 out if the supply voltage is below 15V. A 100uF electrolytic capacitor filters the resultant supply. Note that the 1002 resistor (R2) in series with D4 should be reduced to 102 if a 5V supply is used Ifthe supply voltage is significantly ‘more than the voltage rating of th lay, it will need aresistor in series with the coil. This is shown on the circuit as R1. As previously stated, the relay is driven by Mosfet Q3. If the voltage rating of the relay coil is close to the supply voltage, resistor R1 is omitted and link LK@ inserted instead. Do not be concerned about the nor- mal voltage variation which can be expected from 12V or 24V lead-acid batteries. A 12V battery may goas high as 14.8V while being charged while a 2AV battery can go to 29.6V. Both 12V ‘and 24V relays can cope with this vari- ation and there is no need for a series dropping resistor. Diode D3 and its associated 100nF capacitor suppress the back-EMF transient when the relay switches off. Construction The Threshold Voltage Switch is built on a double-sided PCB coded 99106141 and measuring 107 x 61mm. This is designed to clip into the side channels of a plastic UBS box (130 x 68 x 44mm), with the external leads exiting via a cable gland. ‘The UBS box is optional, howover. Depending on the application, it may be more convenient to house the PCB inside existing equipment. ig.3 shows the parts layout on the PCB. Begin by inspecting the PCB for any defects (rare these days) and checking that the hole sizes for the larger parts are correct. If this checks ut, the next step is to select the relay siliconchip.com.au29 JuLy 2014 siliconchip.com.auTonos Fig.3: follow this diagram to build the TVS with an on-board relay. Install LK1 to divide the input signal, remove LK2 and install ZD2 for a 24V supply and install LKa if the supply voltage doesn't exceed the relay rating (see text). LK3 selects high or low threshold triggering. to be used from Table 3 (near the end of this article) Choosing the relay Basically, there are several different rolays that can be used with the TVS ‘The overlay shows a standard 12V DPDT relay set up. It’s just a matter of selecting a relay that suits your application, Note that LK4 is fitted for most re- lays. However, if the supply voliege exceeds the voltage rating of the to be used, then LK¢ is let out and SW resistor R1 is fitted instead. R1 is wired in series with the relay coil to drop the voltage. The value required for R1 is easily calculated. For example, if the relay coil is rated at half the supply voltage (og, a 12V relay with a nominal 24V supply), then the resistorneeds to have about the same resistance as the relay coil. In other cases, you can calculate the required value for R1 as follows (1) subtract the relay coil voltage from the powersupply voltageand multiply the result by the coil resist (2) divide the result obtained in step 1 by the relay coil voltage to obtain the resistor value required. For example, torun a 12V relay with a coil resistance of 1200 from an 18V supply, you will need a 60225W series resistor. This is calculated as ((18- 12) x 1200) = 12. If the calculated value is not a standard 5W resistor value, choose the next highest available value, As stated earlier, fora 5V sup- ply, resistor R2 must be 100. Regulating the supply By carefully choosing the values for ZD1 & ZD2, the supply for IC1 can be regulated. However, this is only required if the threshold voltage must have a very high precision, ie, the swing in the input voltage being monitored is helow 100mV. The 3.3V reference is quite stable but it will vary by about 1mV for each 1V variation in the V+ rail Another reason for a regulated sup- ply is that it makes for a more consist- ent hysteresis voltage For example, if a 12V lead-acid battery is used to power the TVS, the supply can vary from 11.5-14.8V. In that case, changing ZD1 to 10V will ‘minimise any change in the threshold or hysteresis as the supply vari Similarly, for a 24V battery, both ZD1 and ZD2 can be 10V types. The point is to ensure that the supply ISIStOr 4-Band Code (1%) yellow violet yellow brown Colourcode ‘orange orange red brown brown black red brown brown black brown brown voltage always exceeds the sum of the values of 2D1 and ZD2, but ZD1 must be between 5.1V and 15V. Resistor R2 can remain at 1002 0.5W for a 12V supply but should be changed to 2202 0.5W for a 24V supply. Installing the parts Once you've decided on the relay and supply regulation option, you ca begin installing the parts on the PCB. ‘The resistors, diodes and zener diodes can go in first. Table 1 shows the resis- tor colour codes but a digital multi ter should also be used to check each resistor before soldering it into place. Make sure that the diodes and zener diodes are installed with the correct polarity, ie, with the striped end of each device orientated as shown on Fig.3. Note that ZD2 is not required if you intend using a supply of 12V or less (LK2 is fitted later instead). ‘The three PC stakes can go in next, one at TP GND and the others at TP1 & TP2. Follow these with Mosfet Qi — nounted horizontally and secured to the PCB using an M3 x 6mm screw and nut. Bend its leads at right angles before mounting it into position and be sure to fasten its tab to the PCB before soldering the leads. ‘5-Band Code (1%) yellow violet black orange brown ‘orange orange black brown brown brown black black brown brown brawn black black black brown 30 SILICON CHIP siliconchip.com.auVoltages For The TVS Although its contacts may be rated higher, the maximum switching volt- age for the on-board relay is 60V DC ‘0r 40VAC. Do not try to switch mains voltages using an on-board relay, as the tracks on the PCB are too close together. I you do want to switch mains, you willneed to use an off-board relay that has contacts rated for 230VAC. Many will be rated for 230VAC but those designed for automotive applications (eg, horn relays) will not be. REGA, Qt & Q2 are next on the list Be sure to use the correct device at each location and note that Q1 isa BC337 while Q: (don’t get them mixed up). IC: & IC: can then go in, again taking care to get them mixed up and making sure that they are orientated as shown (io. pin 1 at top left). They can either be soldered directly to the PCB or you can use IC sockets Now for th ‘The elec trolytic types must be installed with the polarity shown (the longer lead being positive), while the MKT ca- pacitors can be mounted either way around. Once these parts are in, you can fit the various pin headers for the jumper links. LK1, LK2, LK4, LK5a & LKSb all require 2-way pin headers. Note that the LK4 header must not be installed if resistor R1 is to be fitted. 3-pin header LK3 should also be fitted now. LEDs & trimpots The three LEDs can be pushed all the way down onto the PCB or they can be mounted a fow millimetres proud capacitors. siliconchip.com.au of the PCB. Make sure that each LED is orientated correctly, with its anode lead (the longer of the two) going to the pad marked ‘A’, A cardboard spacer slid between the leads of each LED when soldering can be used to ensure consistent lead lengths. Alternatively, if you want the LEDs to later protrude through the lid of the case, then it will be necessary to extend their leads and sleeve them in heatshrink tubing. You could also glue them to the lid and connect themto the PCB via flying leads; you could even fit pin headers in their place and use fly ing leads terminated in header plugs. Trimpots VR1-VR3 are straightfor ward to install, Use the 1M@ trimpot (code 105) for VR2 and be sure to install them with the adjusting screws to the left Now for the screw terminal blocks CON1 consists of two 2-way terminal blocks and these must be dovetailed together before fitting them to the PCB, Push them all the way down onto the board and check that the wire entry holes are facing outwards before soi- dering the pins. CON2 is required if you intend ing a PCB-mounted relay. It consists of three 2-way (or two 3-way] terminal. blocks and again check that it sits lush against the PCB and is orientated cor rectly before soldering the pins. Alternatively, if an’ external relay with quick connectors it to be used. then the two 6.35mm PCB-mount male spade connectors will need to be in- stalled. These are located just above Q3 and provide the relay coil connections. Configuration Once the PCB assembly has been completed, go back over your work and check it carefully, In particular, look for incorrectly orientated parts, parts in the wrong position and missed solder joints. If all is correct, follow this step-by-step procedure to config- ure the unit: Step 1: if you are using a 12V or 5V supply, install the jumper shunt for LK2, Alternatively, for a 24V supply, install zener diode ZD2 and leave jumper shunt LK2 out. Step 2: fit jumpers on LKSa and LKSb so that LED1 & LED2 will work. Step 3: fit a jumper on LK4 if Ri has not been fitted Step 4: adjust trimpots VR1, VR2 & VR3 clockwise until the end stop clicks can be heard (note: these are 20-turn or 25-tumn trimpots) Step 5: apply power and check that voltage is present between pins 8 & 4 of IC1. The actual voltage will depend on the supply, zener diodes ZD1 and ‘ZD2 and whether ZD2 is bypassed. Ifyou are using a 12V supply and a 15V zener for ZD1 (LK2 in), IC1 should have around 11.7V between pins 8 & 4. For a 5V supply, you should get a reading of about 4.7V. And for a 24V supply (ZD2 in and LK2 out), you should get a reading of about 8.7V. Input signal level adjustment LK1 can be installed to allow the input signal to bo reduced if the volt- age to be monitored is going to exceed 3.3V. To set VR1, apply a voltage simi- Jarto that you require forthe threshold (say 10V) to the input, switch on and measure the voltage between TP1 and TPG. Adjust VR1 to obtain less than s.3V at TPA Threshold adjustment The threshold voltage adjustment is done as follows. Apply a voltage at the level you want the TVS to switch ‘The PCB clips neatly into the slotsTVS Supply Voltage: 5V (LK2 in) 12 (LK2 in) 24V (LK2 out, 202 installed**) On-Board Relays (Maximum ‘Switched Voltage = GOV DC or 40VAC) YADPOT POE Mount (ELAY2) Contact rating: 24y DG/4OVAC Altronics $ 4147, Altronies $ 4150, Altronics $ 4152 SADPDT PCB Mount RELA) Contact rating: 30v DC/4VAC BA DPDT PCB Mount ELAY!) Contact rating: SOV DE/MOVAC Jaycar SY-4052 Jaycar SY-4053 Altronies $ 41800 Altronics $ 4270 Altronios $ 41950 Altronios $ 4272 aoa ET ‘Board Relays (Maximum Switched Voltage Limited By lay Contacts) 30 (RELAYS)* Aronies $4211 SPOT | Use 12V relay. R1=1800 (for $4211), 1200 Contact rating: 14V DC/240VAC_ Jaycar SY-4040 SPST (for SY-4040) 5W, LK6 out ‘30A SPST Horn Relay* Altronics S 43354 Altronics $ 4332 Gontact ating: 4V BC “Jayoar SY-406B~ | Jayear: Use 12V relay. RY 8202 SW, LK6 out 30A SPDT Horn Relay” fe as Contact rating: 14V BC : ___sayear 4070 Use 12 relay, 1-820 5, LKB ot 60A SPOT Horn Relay” Atrios § 4339 vy Rte 7 Contact ating 14V OC “aycar S¥4074 Use 12V relay, R820 5, LKB out ‘Notes: LKG installed (jumper in) unless stated ** Install 1N4744 16V zener ZD2. * Bolt on and quick connector type. Requires 2 x 6.35mm PCB-mount male spade connectors with 5.08mm pin spacing (Altronics H 2094) plus 4 x 6.35mm insulated female spade quick connectors with 4-8mm wire diameter entry (these are not suitable for the 6DA relay), A variety of relays can be used with this unit, such as DPOT (double-pole double- throw), SPOT (sinale-pote ouble-throw) and SPST (sngle-pole single-throw). Double-pole (OP) simoly means that there are two separate sets of contacts that can be used indepen dently to switch power (or even signals) Single-throw (ST) and double-throw (OT) contacts each havea common (COM) contact and both ST and DT types have acontat that isopen when te relay sof, the normally open or NO contact. This NO contact closes against the COM terminal when the relay is 1on ie, the coil is powered) In relays with DT contacts there is also a normally ctosed (NC) contact. Tis connects tothe COM terminal when the relay is off and ‘opens wien the relay ison, Both SPOT and DPOT relays give the op- tion of powering something when the relay 's either switched on or is switched of. For example, you can set up the TVS so that power is switched on when the relay is off by connecting te oad to its supply vi the NNG and COM contacts. The main reason to do this isto minimise the current drawn by the circuit. The TVS typically draws les than ‘mA hen the relays offbut when terelay is on, the current drawn by isco wl typically be around S0maA or upto 100mA, depending on the tlay used Table 3 shows the various relays that can be used withthe Threshold Voltage Switch The choice depends on the supply voltage and the current to be switched by the reay's contacts PCB-mounting relays are accommodated on the PCB and their contacts brought out | from a 24V supply and can obtain a suitable to screw terminal block CON2. By contrast, relays with quick connect terminals are ‘mounted off the board. You can either use leads fitted with quick connectors or you can solder the leads directly to the terminals. Since relays with 12V coils are more common than 24V relays, the TVS has been designed so that it can use a 12V relay even when operating from 24V. It's just a matter of removing LK4 and installing a dropping resistor (R1) on the PCB, in series with the relay’s coil Having said that, if you are operating relay with a 24V coil and the correct pin-out this will generally halve power and current Consumption when the relay is energised In that case, leave Rt out and install jumper KA instead the relay, then adjust VR3 until the threshold voltage is reached. LED1 will light when the input is above the threshold, while LED? will light when the input is below the threshold, With hysteresis trimpot VR2 set at maximum, the threshold for a rising input voltage will be similar to that of falling input voltage. This hysteresis can be increased by reducing the value of VR2 (ie, turn VR2 anti-clockwise for more hysteresis) 32 SILICON CHIP Changing the hysteresis will also affect the threshold voltage previously set using VR3, so you will now need to readjust VR3 to correct this. Once that’s done, check that the hysteresis using VR2 is suitable and repeat the above steps if necessary Jumper LK3 determines whether the relay turns on or off for rising or falling threshold voltages. Install LK3 in the HIGH position if you want the relay to turn on when the input volt- ‘age exceeds the threshold. Conversely. install LK3 in the LOW position if you want the relay to turn on when the in- put voltage goes below the threshold. Finally, to reduce the current drawn by the Threshold Voltage Switch with the relay off, jumpers LK5a & LKSb can be removed (to disable LED1 & LED2) conce the set-up procedure has been completed. Alternatively, you may leave them in to monitor the unit’s operation, siliconchip.com.auravac D2 mq TWA ‘curour Qo TRANSFORMER WITH CENTRETAFPED SECONDARY Fig.4: typical battery charger circ winding transformer with's four-diode bridge movac | TRANSFORMER WITH UNTAPPED ‘SECONDARY jing either a centre-tapped transformer with two rectifier diod tifier (B), Charging With the lihkesholdaVoltagexswitch) ANY READERS have asked for a simple solution to prevent over- ng of lead-acid batteries. Most simple battery chai any end-of-charge detection and will continue charging at their full current even though the battery may have reached 14.4V. If allowed to continue for too long, such over-charging leads to severe gassing, excessive fluid loss as the battery overheats and even buckling of the plates. Ultimately, the battery will fail much sooner than it should. Over-charging can also lead to a build-up of hydrogen closed space, which is an explosion hazard, especially in the pre: sparks (often caused if the battery is disconnected during charging). An elegant solution to this problem istouse our Threshold Voltage Switch asa battery charge cut-off device and you can then adda trickle charge facil: ity as well. So why do most battery chargers not limit or stop charging when the battery reaches 14.4V (in the case of 4 12V lead-acid battery)? The is that most chargers simply comprise a transformer and rectifier supplying ers do not have nce of raw full-wave rectified voltage to the battery Fig.4 shows two typical battery charger circuits. These use either a centre-tapped transformer with two rectifier diodes (A) ora single winding siliconchip.com.au transformer with a four-diode bridge rectifier (B) The charger will usually also include temperature cut-out that switches the charger off when the transformer runs, too hot. But there is no facility to sense the battery voltage or stop charging above a certain voltage. You may have a commercial bat- tory charger that uses a circuit like one of these or you may have built the Bits'n'Pieces Battery Charger from April & May 2013 SILICON Cit Either way, the charge process can be monitored to ensure that the battery isn’t overcharged, Overcharging can easily occur since these chargers use a nominal 12V (or higher) transformer. The output after rectification is pulsating DC with a peak voltage of around 17V. If the charger is left on charge for too long. the 17V peak can overcharge the bat- tery easily, reaching well beyond 15V if left unattended. This solution is the Threshold Volt: e Switch. It can monitor the battery and switch off the charging current as soon as the voltage reaches 14.4V. Ad ditionally, the hysteresis can be made sufficiently large so that charging does not recommence until battery voltage falls to its 12.6V (typical) resting volt- after charging ceases. Fig.5 shows the required arrange- ment. The output from the charger VOUAGE SWmCH * OFTONAL TICKLE CHARGE RESISTOR (IW RECOMMENDED} Fig.5: here's how to add the Threshold Voltage Switch to a battery chargé jomatically ceases when the batt rickle charging (see text). so that charging ai Resistor Ry is opt nal f y is fully charged. Juty 2014638ToaFOR SV = Ste Fig.6: follow this diagram to assemble the PCB and wire it to an external relay and battery ch circuit. Medium- duty hook-up wire can be used for all connections to the PCB but be sure to use heavy duty cable for all connections ‘between the charger and the battery and to the relay contacts (30 & 87). is switched using a 60A 12V relay (Altronics $ 4339 or Jaycar SY-4074), This heavy-duty relay is mounted externally, since it is too big to fit on the PCB, It works like this: when the Com- mon (COM) and normally open (NO) contacts are closed, the output from the charger is connected directly to the battery and the battery cha {As soon as tho battery roaches 14.4V, the relay switch contacts open, thereby disconnecting the bat- tery to prevent overcharging, ‘The supply for the Threshold Volt- age Switch is derived from the charger (rather than the battery), so that the battery doesn’t begin to discharge when charging ceases. We do, how: monitor the battery voltage this process results in a current drain of less than 32A. That's much less than the battery self-discharge current. Note that the wiring to the TVS for voltage sensing is run separately from the battery terminals. This ensures that voltage drop across the charging leads does not affect the measurement. Adding trickle charging Switching to trickle charging at the end of full charge isa good idea, since it ensures that the battery is always fully charged (without the risk of over- charging). The trickle charge must be low enough toallow the battery voltage to drop to below or be held at 13.8V. ‘Typically, the trickle current should 34 SILICON CHIP be 0.025% of the battery’s Ah capac orabout 10mA for a 40Ah battery. ‘an beachieved by adding a 2200 ss the relay contacts jor value is calculated as- suminga charging voltage of 15.8V (ie, 2V more than the 13.8V battery volt- age). A 2200 resistor will dissipate less than 0.25W but we recommend using 2 1W resistor as it is more rugged and. has thicker leads to make the connec- tion to the relay terminals. Fig6 shows the PCB layout and external connections necessary to connect the TVS to the battery and the charger. The relay is mounted externally, with its coil terminated to the contacts on the PCB using spade quick connectors. Note that Fig.6 shows the arrange- ment for charging a 12V battery. Zener diode ZD1 is now a 10V 1W type (1N4740) instead of the original 15V zener and provides a regulated 10V. supply for comparator IC1a. This regulated supply is necessary because the hysteresis must be made quite wide and because supply variations would affect the voltage at which the TVS switches off charging, For a 24V charger and battery, use another 10V 1W zener diode for ZD2 and leave LK2 open. In addition, the 1000 resistor (R1) needs tobe changed to 2200 0.25W. You will also need a relay with contacts rated for 28V DC. Medium-duty hook-up wire can be used for all connections to the TVS but note that heavy duty cable must be used for all connections between the charger and the battery and for the connections to the relay contacts (30 & 87). We used 25A cable on our prototype but you could use 10A cable if the charger is a low-current type rated at less than 5A. Asshown in the photos, we installed the PCB and relay in aUB1 plastic util- ity case measuring 158 x 95 x 53mm, ‘The PCB mounts on M3 x 9mm tapped stand-offs and is held in place using M3 x 6mm screws. The relay is bolted to the base of the case using an Ma x ‘12mm screw and an M4 nut. Finally, the connections to the rela contacts are all made via quick cor nectors and the external leads are fed through a 10-14mm cable gland at one end of the case. Setting up the TVS The TVS must now be set up for battery charging following this step- by-step procedure: Step 1: feed a voltage (eg, 9V) to the signal input on CON, then accurately measure this voltage using a DMM (no need to apply power). Step 2: connect the DMM between TP1 and TP GND, make sure LK1 is installed and adjust VR1 for a reading that’s one-tenth the measured voltage in Step 1. Thissets VR1 to divide by 10. Step3: measure the resistance between ‘TP2 and the LOW position of LK3 (with the LK3 jumper out). Adjust siliconchip.com.auVR2 for a reading of 90K@ to set the hysteresis appropriately. remove the input voltage, short the signal inputs on CON1 using a wire link and apply power tothe circuit us- ing the battery charger ora 12V supply monitor the voltage between TP2 and TP GND and adjust trimpot VR3 for 1.44V. This sets the TVS to disconnect the battery from the charger when it reaches 14.4V. The hysteresis setting ensures that the TVS will not switch the relay back on again to recon: nect the charger until the input voltage falls below 12.6V remove the shorting link on the signal input and +" input to the positive supply rail ead. Now, with LK1 out, check the voltage between TP2 and ‘TP GND; it should be close to 1.26V. TP2 should lun to 1.44V if the signal input is again shorted to ground (ie, to OV), install LK3 in the LOW po. LK5a & LKsb can either be «dor left in circuit to show the status, In practice, leaving LED & LED2 ope ‘onnect the signal rom ting is a good idea because the 33 Getting D The PCB and relay for the battery charger cut-out be installed in ‘The external leads exit through a cable gland at one end of the case and the leads for the battery terminated in large alligator clamps. The two leads with the bared wire ends go to the battery charger. relay indicator LED3 now glows even wher his is due to the supply coming from the pulsating DC of the charger plus various capacitive n effects which cause the LED to light. e Bycontrast, with anormal constant DG supply, the relay LED is extinguished when the relay turns off. st fy BR OC CSS LOU US LUNA Ce ama) URE Re COO CR Cam ROM ie stay Bleed eee erect siliconchip.com.au Ly 20435