Fullscreen HTML5Med Res (??MB)HTML4

One-off boards for the hobbyist, prototypes, etc Yes, You Can Make PhotoResist PC Boards At Home Making your own PC boards has almost become a lost art. Last month we showed how easy it was to transfer laser-printed or photo-copied images to a blank board using an ordinary iron. While that method works, it’s not real good for fine tracks. Here’s a way to get pro quality PC boards using laser prints or copies. . . by Ross Tester P rinted circuit boards have revolutionised electronics over the past forty years or so. It’s no exaggeration to say that they make some projects possible – it would be well nigh impossible to wire up many designs involving ICs, for example, using point-to-point wiring. Just imagine a modern computer without PC boards! And they also make life easy for hobbyists. Providing you know how to solder AND you start with a clean, bright PC board, assembling a project on a PC board is arguably the most foolproof and mistake-proof method of building (even for projects which could be done other ways). even given a thickness – it is expressed as a weight of copper per square metre. 1oz (yes they still use ounces) PC boards are common, as are 2oz. But you can get 3oz and even more where a thick copper is needed. Most PC boards, especially the type you will come in contact with, have the copper on one side only. But it’s not unusual to find PC boards with What’s a PC board? Their full name, Printed Circuit Boards, gives you a fairly good clue! Once upon a time they had an even more accurate name, Printed Wiring Boards – but this name didn’t “stick” whereas PC boards did. A PC board starts life as a piece of thin fibreglass or phenolic material (and occasionally others) which is a very good insulator – as far as we are concerned, about as perfect an insulator as we can easily get. Onto this is glued a thin (no, make that very thin) sheet of pure copper. The copper is so thin it usually isn’t 70  Silicon Chip Most of what you need to make your own PC boards at home: the large packs contain pre-sensitised blank PC board (150 x 300mm sheets). In the plastic sachets are measured amounts of sodium metasilicate developer, while the plastic jar contains 600g of etchant – in this case ammonium persulphate. Not shown are the exposure box or the etching tank. (Courtesy Computronics Corp). The proof of the pudding – here’s a selection of boards, as yet uncut and undrilled, which we made using the method we’re describing here. Both SRBP and fibreglassbased material was used. Some boards are better than others: we certainly got better as we experimented with exposure times (with this resist and light source 6 - 6.5 minutes was about the best). We used “EPS” files similar to what you would download from our website, flipped them, printed them out on bond paper on our laser printer then used the images to produce the boards. Total time elapsed: about an hour! copper on both sides. In fact, many computer PC boards have many more layers – four, five, and more, with each copper layer sandwiched between a layer of fibreglass. But we’re getting a bit ahead of ourselves here. By various means, which we’ll cover in a moment, areas of the copper are removed from the board leaving “tracks” and “pads” which connect to each other. These tracks and pads form the “wiring” which connect the various circuit components together. In each of the pads and often in various places on the tracks, tiny holes are drilled right through the fibreglass and what is left of the copper. The circuit components are soldered to the copper, connecting them into circuit. Almost always on a single-sided board (ie, copper on one side only) the components are mounted through the board from the non-copper (eg fibreglass) side and their legs soldered on the copper side. If you think this is blindingly obvious, you’re perhaps a bit cleverer than the customer who some years ago sent a Musicolor kit he’d built into the Dick Smith Electronics service department, saying it was faulty. The service manager at the time (g’day Garry) commented that he’d never seen such a well constructed kit, especially the way all of the components were so neatly and carefully glued to the copper side of the board with Araldite… How are PC boards made? There are many ways to make PC boards, depending on the use, who’s making it and the number being made. However, all methods involve three main steps. Step 1: preparation Unless you’re very lucky, the blank PC board you buy (or have in your junk box) won’t be the right size. Not only must you cut the board prior to use (usually a centimetre or so larger in each direction than your finished board), most importantly you must ensure that the copper side (at least) is scrupulously clean and dry. Even though a board might look clean, it probably isn’t. It could have fingerprints on it; it could have lint and dust on it but worst of all the surface could be slightly oxidised (copper in the presence of oxygen, ie from the air, quite quickly oxidises), making other steps in the board making process difficult or impossible. Step 2: the image transfer This is usually the most difficult for the hobbyist: getting the image of the tracks and pads onto the copper. Usually, this involves some form of “resist” – a material which resists the action of the etchant, leaving the copper underneath intact. (We’ll look at etchants at shortly). Remember a moment ago we said the blank board might have fingerprints on it: the oil in fingerprints is a pretty good resist! Getting the image on is where the processes differ greatly. We’ll look at just a few: • Painting the pattern onto the blank board using a tar-based or similar waterproof “paint”. It’s messy, it’s not easy to get a good result and it’s almost impossible to produce fine tracks and inter-track spacing. • Tracing the pattern onto the board using, say, carbon paper, then going over this with a “Dalo” resist pen or similar (pens which contain resist). While easier, and capable of better results than the paintbrush method, similar problems remain. Dalo pens, by the way, are often used to repair faulty resist in other methods. • Transferring the pattern onto the board using a method such as the iron-on process described last month. This uses the carbon black and plastic binders of a laser-printed image or photocopier as the resist. The biggest difficulty here is getting consistent results. MARCH 2001  71 The difference between positive and negative: at left is a POSITIVE image of part of a PC board pattern – black tracks on white/clear background. At right is the NEGATIVE image of the same board: white/clear tracks, black background. Incidentally, you can buy material specifically intended for this process. We’ve tried them from time to time but have had little more success with them than using ordinary bond paper. • “Silk screening” the image on. This first requires the image to be photographically transferred to a silk screen and then the resist is applied by forcing it through the silk screen in contact with the copper, using a squeegee. This is the method most often used by PC board manufacturers because it lends itself to mass production. It’s not really one for the hobbyist or even commercial prototypes. • Direct photographic transfer of the image onto a photo-sensitive resist which has previously been applied to the blank board. This resist can be applied from a can or bottle, or you can buy blank PC boards which have the resist pre-applied. The latter usually give the best results. In either case, the resist must be processed in a suitable developer and dried thoroughly before etching. It is the last-mentioned method which we will be describing here. Normally, this method is used for one-offs or prototypes in industry but has been rather difficult for the home constructor due to the materials and equipment involved. Commercial users normally employ a relatively expensive photographic film positive or negative which has very high contrast, resulting in excellent results. However, it is possible to do a poor man’s version using an ordinary laser or inkjet printer. You should get perfectly acceptable results – maybe not quite as good as with film but acceptable nevertheless. Positive or negative resist Photosensitive resist can be positive-acting or negative-acting. Posi72  Silicon Chip tive-acting resists require a PC board pattern which has black where the copper tracks are and white or clear between them (ie, a “normal” looking pattern as you would see published in the magazine). Both types of resist have the image transferred by exposing them to UV light through the image while it is held tight against the resist. With positive-acting resist, the black areas stop UV light affecting the resist but the white or clear areas allow the UV light to “soften” the resist, allowing it to be “developed” away. Negative acting is the reverse: the copper tracks are white or clear and the areas between them are black. UV light hardens the exposed resist while the unexposed areas can be developed away. As a general rule, most commercial operations use negative acting resist; most hobby or one-off prototyping is done with positive acting resist. If in doubt, read the label. Step three: etching Once the required image is on the blank PC board it must be prepared for etching. Etching involves the use of chemicals which dissolve copper –they eat away at any area of the blank board not protected by resist. There are two common types of etchants used for PC boards. The first is Ferric Chloride, (FeCl2), a brown liquid (or more correctly a brown powder which dissolves in water) which has the habit of staining anything it touches! Its big advantage is that it works very well at room temperature. And for commercial users, it is a relatively easy process to extract the etched copper back out of Ferric Chloride – copper is a valuable mineral which they can sell to metal recyclers and make a few bob on the side! The other common etchant is Ammonium Persulphate ((NH4)2S208). When dissolved in water it makes a clear liquid, which is much cleaner to use than Ferric Chloride. However, it has two major disadvantages. The first is that it must be heated significantly (at least 60°C) to make it usable; the second is that because it is colourless, splashes tend not to be noticed until such time as they’re busy eating away at the kitchen sink, adjacent pots and pans, etc! Despite these two hassles, Ammonium Persulphate is by far our etchant “of choice”. Some sources suggest Hydrochloric Acid as an etchant. We have just one word for that: don’t! Step four: finishing What’s this? We said there were only three main steps. OK, we lied! One way of producing a double-sided board in perfect registration. The two sheets of film or paper are first aligned on a light box then stuck to PC board offcuts. The blank PC board to be exposed is then slid between them. Finishing off is just as important as the other steps. First, you have to drill all the holes out. Usually, we use a 0.8mm drill bit for most component holes and a 1.0mm for the larger (ie PC stakes, some semiconductors, etc) holes. You may find that some components such as on-board pots require larger holes – 2.0mm for example. And mounting holes tend to be 3.0mm. By the way, you’ll find drilling a lot easier if you use a drill stand. Even better is a small drill on a stand intended for the purpose (eg, a “Dremel”) but that might be going a bit overboard for hobbyist use! Then again, there are some hobbyists who maintain you aren’t serious if you don’t have a Dremel drill in your arsenal! But that’s not all there is to finishing off. You also need to cut the board to the right size. Commercially, this would be guillotined but you’re probably going to have to cut it slightly oversize with a hacksaw and then file it back to the correct size. And finally, there’s the little matter of getting the resist off those copper tracks. Sometimes you don’t need to – some resists are specifically designed to be able to solder through and are supposed to stay in place to protect the copper surface. Other resists must be removed with a suitable solvent (otherwise you won’t be able to solder to the board) and then once again the copper surface needs to be protected with a suitable protective coating (one which will allow soldering through). You can, by the way, make up your own “flux” coating which protects the copper surface as well as making soldering real easy: simply dissolve a few rosin crystals in a small quantity of metho and paint a thin coating onto the board. Where do you get rosin these days? We don’t know either! Double sided boards We mentioned double sided boards a while ago. These are not all that common but are still well within the scope of the hobbyist if care is taken to keep the alignment perfect (it’s called “registration” in the trade). This can be done by making a sandwich of the patterns, glued down one side to hold them in register. The double-sided sensitised board is then stuck in position to one only of the sheets as the “meat” in the sandwhich. Exposure is done as for a one-sided board but we would place some black plastic or other light-proof material against the resist on the opposite side while exposing. Some exposure systems, such as the Kinsten one shown, expose both sides at once. Another challenge for the home builder making double-sided boards is how to get the sides linked together. In commercial boards, this is done with holes that are plated through, making contact with both sides. The easiest way for the home constructor to do it is to use component lead offcuts and solder them to both sides. Likewise, where components go through holes with copper on both sides, they should be soldered on both sides. The technical name for these connections, by the way, is “vias”. (Current flows from one side to the other via the via...) Making PC boards at home (or small scale prototyping) Simple: follow the steps above! Seriously, though, folks(!) there really isn’t a great deal more to it than that. Let’s just expand on the steps above where they need expanding. “Milling” or “Routing” PC Boards We have been asked if it is possible to use an X-Y plotter or table, with the appropriate head, to mill or rout PC boards. The answer is yes, but… For a start, you need more than an X-Y table – you need the Z axis as well to be able to lift the bit clear of the board when traversing wanted sections. You also need the Z axis to raise and lower the drill(s) and cutting bits. Good X-Y-Z tables should have enough accuracy to mill a PC board. The difficulty lies in having the software capable of driving your particular table to do the job. None of these problems are insurmountable, of course, and many quality PC boards are made using this process – with nary a grain of etchant (nor any other chemicals!) in sight. A big advantage in producing PC boards this way is that very complex board shapes can be realised as well as cut-outs within the boards themselves. And a milled board will never have any undercutting or bridges (assuming the software is OK!) Some of today’s PC board design software has the capability of driving a miller or plotter instead of a printer; if it can it will generally also be able to automate the drilling (always a tedious part!). Most tables, though, will require some translation to be able to be used properly. There is yet another use for a table or plotter: using resist ink and plotting the PC pattern direct to the blank board. This is then etched in the normal way. We once did all our PC boards at SILICON CHIP this way; we gave it away for two main reasons – the difficulty in keeping plotting pens clean with this type of ink; and also because of the time it took to produce a board. Sometimes it’s cheaper for a business to get them done commercially, drilled and all: time is money! If all this is double dutch to you, we suggest you read a recent article in SILICON CHIP which reviewed a commercial PC board milling machine: “Quick Circuit 5000 PC Board Prototyping System” November 2000. An ar ticle on plotting patterns to blank boards appeared in the November 1994 issue. The Quick Circuit 5000 PC Board Prototyping System mills boards instead of etching them. It’s capable of cutting a variety of shapes as well as milling the unwanted copper away. It’s not real cheap, though! MARCH 2001  73 (1) Cleaning the blank board (2) The resist (3) Your PC pattern As we said before, your blank PC board needs to be cut about 1cm or so larger than the finished board in each direction. You should also file off the edges to make sure there are no bits of copper poking up. Of course, if you are using pre-sensitised board it comes already clean as well as coated. So you can skip straight to step 3! There are a couple of conflicting aims in cleaning. One is that you need to have the copper clean – very clean – but you don’t want to scratch deep gouges in the copper surface. That would appear to rule out steel wool (in fact, the text books say so!) but to be honest, we’ve used steel wool on badly oxidised boards and achieved perfect results. Normally, though, we’d use something like powdered “Ajax” and a new, non-metallic scouring pad. You shouldn’t use the old scouring pad from under the sink because invariably it will have bits of grease and grime trapped in it, which could be transferred to the copper surface. When you are sure the copper surface is very clean, give it a good rinse under fast-flowing water and then stand the board vertically in the sun to dry. Don’t wipe it clean because this could leave lint or fibres on the surface. Have a good look at the board (even use a magnifying glass) to make sure there is nothing on it, then protect it from dust. Having just gone through all that, there is a way which you can avoid all of the above steps and hassles (and some of the next!) and that is to use a pre-sensitised blank PC board. These are available from a number of suppliers – those shown are “Kinsten” brand boards (from Computronics, 08 9470 1177). Another popular brand is “Riston”. If you use the pre-sensitised boards, all this is done for you. You simply have to cut the board to the required size under subdued (ie, normal household) light. Sunlight is a no-no. Once you open the light-tight packaging, avoid unnecessary exposure for the remainder of the boards in the pack and also for the board you are handling. A couple of minutes, a couple of metres away from a fluoro light won’t worry it too much; much longer or closer you will risk “fogging” the resist and therefore making it useless. If you must apply your own resist to blank boards, first make sure the resist you are using is positive acting (otherwise you’ll end up with the reverse of what you want). Photo-resists are commonly available in either liquid or spray-on form. In both cases, the idea is to get a nice, even coating on the copper surface, not too thick and not too thin. Apply with a “swirling” motion to move the resist around and into missed areas. While resist is fairly liquid, it starts going thicker fairly quickly so you need to work reasonably fast. Most spray or liquid resists do not dry hard enough naturally and must be baked in a just-warm oven/frypan. Follow the instructions carefully when baking – and remember that as it dries the resist becomes more and more light sensitive (that light in the oven?). Once your resist-coated board has dried properly, it’s much like the pre-sensitised ones (including handling and light sensitivity). As you probably know, as well as being published in the magazine, most PC board patterns for SILICON CHIP projects are available from the website (www.siliconchip.com.au). Download these and you can make your own PC boards. However, there is a choice when it comes to printing out the pattern. You can usually achieve a more-than acceptable result by printing the pattern on plain bond paper (ie, photocopier paper). People who use plain paper report “10 thou” tracks (small enough to fit between IC pins) are no problem. But you will probably achieve a better result by using clear film, as used for an overhead projector. First, plain paper: you need two things: one is a very good quality print with absolutely black blacks. Most modern day laser or inkjet printers will achieve this for you. The other, and most important, is you need a reverse direction, or “mirror image” print – that is, any writing is back to front when you look at it. The reason for this is simple: you want the black image in intimate contact with the resist so that the light which exposes it doesn’t have to then pass through the paper. Otherwise light scatter occurs in the paper which results in a much inferior result. Most printer drivers have the facility for printing a reverse direction, “mirror” image. (Note that you don’t want a “negative” image – that reverses whites and blacks). Ensure also that the size is right – PC board sizes are given in the project parts lists for this reason. Hey, we’ve seen boards made 200% or 50% of original size. They look good but gee the components are hard to fit! At far left is the laborious task of cleaning blank PC board. It’s important to remove all gunk and oxidation prior to coating with resist. Both cleaning and coating are already done when you use presensitised board such as this “Kinsten” brand board from Computronics. It’s available in a wide range of board and copper thicknesses. We’d take that “less than 10 minutes” claim with a chunk of salt, but! 74  Silicon Chip (4) Exposing the board Finally, if you have a choice of paper, print on the lightest weight which gives good, consistent blacks. Now to the alternative, film: most laser printers and inkjet printers can print to film, as you would use for an overhead projector. Unfortunately, though, the density (or “blackness”) of most isn’t quite good enough for PC board making. (Hold one up to the light and you’ll see what we mean). This can be easily overcome by printing two copies, then very accurately aligning them and sticking them together. You will see the difference when you hold this up to the light! As with paper, print the film reverse direction so that the bottom layer of film will be in intimate contact with the resist. And before use, check the size one more time. Any flaws in the printed image (paper or film) can be retouched with a fine felt-tip pen. This includes breaks in tracks, pinholes, etc. We ’ v e b e e n t a l k i n g a b o u t down-loading and printing PC board patterns – but if you can get a good quality photocopy from the patterns published in the magazine, these too can be used. The major problem you’re going to have is that few photocopiers have the ability to print reverse or “mirror image”. If you must have the ultimate quality, download the PC board “EPS” file from the website and take it to your local DTP service bureau, who should be able to output the file on high-contrast film for you for a few dollars. Tell them you want a film positive, right reading, emulsion side down. This puts the PC board image right next to the resist when you expose it – that is, no layer of film in between. Here’s where you might have to use some ingenuity. The aim is to have that black image of the PC pattern in intimate contact with the resist. Commercial organisations doing a lot of prototypes should invest in an exposure box, such as the Kinsten KVB-30D shown here. Once again, this comes from Computronics. It really is the “Rolls Royce” and has everything you need for great boards: a vacuum pump to ensure the pattern is held tight against the board, a digital timer and even upper and lower UV lights so you can do two sides at once (on double sided boards). All this comes at a price, of course: you won’t see any change from $700 when you This automatic UV exposure unit from Computronics would have to be the ultimate: vacuum include GST! pump, digital timer, capable of double sided So what does a hobbyist do? boards in one exposure . . . but the price tag You have two problems to puts it a tad out of the reach of the hobbyist. overcome. The first is to ensure that intimate contact we talked signed to emit UV and while most is about before; the second is the light converted to visible light by the phossource. phors, enough “escapes” to be usable. The first problem can be solved It is possible to buy special UV as simply as placing the board and fluorescent tubes which glow pale pattern between two sheets of glass, blue (similar tubes are in the Kinsten held together by large bulldog clips. exposure unit). These are available Alternatively, you can buy small exin 20 & 40W sizes to fit standard 2ft posure frames at art and silk screen & 4ft fluoro fittings. However, these suppliers (or you could make one). aren’t recommended for domestic use Just remember, the thicker the glass, because the UV they emit could be the more opaque it is to UV light. harmful to the eyes and skin at close The second problem also has an easy range. And they’re not real cheap! solution – in fact, two easy solutions. Just remember before exposing If it’s a fine day, you could use that big pre-sensitised board to remove the bright yellow thing up in the sky – it backing paper! emits tons of UV light along with visible light (which won’t matter). Exposure times Or you could use ordinary houseNeedless to say, exposure time hold fluorescent tubes. They are devaries enormously according to your light source and your PC pattern type. As an example, even for the Kinsten A high contrast unit recommended exposure varies laser print on bond paper (ie, from 60 to 90 seconds using high qualvery black blacks ity film (ie, very black blacks and clear and nothing in the whites) to five minutes or more using whites) is OK when a laser print on plain paper. you can’t get (or afThere is only one way to determine ford) a photo-graphthe exposure for your setup: experiic film positive. You ment with small pieces of PC board. can get very good And the only sure way to determine results from laser success or failure is to follow the next and inkjet prints. step and develop the image. MARCH 2001  75 (5) Development Developers vary according to the type of resist and also their source. For the Kinsten presensitised PC board, the developer is sodium metasilicate. We’ve also used resists that develop in a weak solution of caustic soda (sodium hydroxide). Some developers are simply labelled “developer” with no hint as to what is in them (which is probably illegal these days). Prepare the developer as per the instructions packed with it. If it is a powder or crystal type, you need to ensure that it is totally dissolved before use. Whatever the type, always use plastic gloves when preparing and using developer. Most are caustic or alkaline and can do wondrous things to your skin. Also use plastic developing trays and implements for the same reason. A pair of plastic tongs is handy. Another useful tool we’ve found is a plastic fork, á lá the local Chinese take-away. To develop the board, place it pattern-side-up in the developer and gently rock the tray to give a slight sloshing motion. You should start to see the pattern emerging after just a few seconds (depending on resist and developer) and then the developer start to lift off the exposed areas of the board within about 30 to 60 seconds. Soon, all of the exposed areas should be free of resist. Most resists will develop fully between about 30 seconds and two minutes. Less than this, the developer is probably too strong and is likely to start attacking the wanted tracks. (7) Etching Longer than this and the developer may never do its job in clearing off the unexposed resist. Developing is normally done at room temperature. Higher temperatures will result in shorter times but again, may make the developer too active. Each pack of developer will handle a number of boards. Most instructions say to mix up a fresh batch of developer for each batch of boards being done as it will only last a day or so. We’ve found that some developers, especially those based on caustic soda, will last for weeks or months. And if they lose their punch, we just throw in another couple of flakes of caustic soda! OK, so it’s not technically correct. But it works for us – and saves us having to buy developer all the time! If the solution is really badly discoloured, that’s when we make up a fresh brew. (6) Drying or “post-baking” Some resists are fairly soft and require “post-baking” (ie, baking after development) to ensure they are hard enough to withstand the rigours of etching – particularly when using hot etchant (ammonuim persulphate). This step applies more to the sprayon or pour-on photo resists. We generally place the board in a just-warm oven or frypan (and we mean just!) for an hour or so after development. Post-baking is not necessary for the Kinsten resist – as soon as it’s developed and rinsed, it’s ready for etching. Developing is done in a shallow tray. Keep rocking the tray to ensure the board is continually being agitated. This board is almost developed. The dark patches on this PC board are where we tried to repair a positive before exposure: the ink attacked the resist! 76  Silicon Chip We’ve already mentioned the common etchants. Simply mix them up according to the directions in a plastic (not metal!) container. We always mix ammonium persulphate with hot water (close to boiling point) to make sure it’s hot enough to use when etching. But be careful – both with the hot water and then with etchant splashes. Ammonium persulphate is supposed to be mixed at about 200g per litre of water – we usually use about a cupfull to the litre. Near enough is close enough! Back in the good old days, a common mix for ferric chloride was “a pound a pint” – probably way too much but it was easy to remember! Again, keep in mind that etchants will attack most metals. As far as etching methods and equipment are concerned, there are also a couple of different routes you can follow here. If you’re only going to do the occasional board, a largeish, flat, heavy-duty plastic tray will suffice. The type used by photographers is ideal. The board sits in this tray pattern-side-up and you rock the tray back and forth to get a wave action moving the etchant over the board. It usually only takes ten minutes or so to etch a reasonable size board this way. Even better is to have two trays, one of which “floats” in the other containing hot water, keeping the etchant warm in the second tray. If you’re going to do a number of boards, it will pay you to invest in an etching tank. They’re fast, convenient and produce less mess. You can use the same tray to etch a PC board. This one is well advanced with blank board appearing at the top. Use a “sloshing” motion to keep the hot etchant moving over the board. It’s a lot slower than using the etching unit shown above right . . . but it’s also a lot cheaper! This etching tank contains a heater and air blower, both of which speed up etching times significantly. Boards hang vertically from the clips visible at the top. Most etching tanks, at least for smallscale use, are similar to the type shown, the Kinsten ET-10 from Computronics. It is a clear or near-clear plastic thin vertical “box on legs”. The idea is for the board to hang vertically in the tank so that, as the etchant eats away at the copper, it can fall away from the surface, allowing the etchant to keep doing its work on the copper underneath. There are a couple of things which will speed up etching. We’ve already mentioned heat: it’s one thing to mix the etchant with hot water at the start but it’s another to keep it warm. One option for this type of etching tank is a submersible heater, preset to keep the etchant at about 60°C or so. It looks for all the world like a tropical fish aquarium heater – probably because it is, just set a bit higher than normal (tropical fish in 60°C water become tropical floaters!). The other item to speed up etching is an air pump, designed to bubble air through the etchant along the PC board surface. This dislodges copper particles much more quickly than hanging or even agitation. Dare we say this pump looks for all the world like a fish tank air pump? To use the etching tank fill the tank with warm (not excessively hot) etchant. You need to avoid thermal shock on the heater glass. Turn on the heater until the pilot light goes out – the etchant is then at the required temperature. Hang the PC board in the etchant using the clips supplied (vary the height of the clips if necessary) and turn on the air agitation. The air pump should always be placed higher than the tank to avoid syphoning etchant into the air pump. Etching should take somewhere between about 3 and 10 minutes, depending on (a) the size of the board, (b) the amount of copper being removed, (c) the strength of the etchant and (d) the temperature of the etchant. It is complete when all the unwanted copper is removed – but be careful not to over-etch because some of the wanted copper may be either undercut (where the etchant starts attacking the tracks from the sides after removing the unwanted copper) or in some cases, completely destroyed. If your PC boards consistently have etched scratch marks, it probably means you were too vicious with the cleaning process and the etchant has found very thin copper to eat away. If it has numerous pinholes, your exposure time is too long OR your PC board pattern doesn’t have enough blacks. Make sure you empty the etching tank and rinse it out – otherwise you’ll find crystals forming in the bottom. (If you use ammonium persulphate etchant, it will crystallise out to copper sulphate). (8) Finishing Off If the instructions for your resist say that it can be soldered through, leave it in place. It will prevent oxidation of the copper. However, many resists must be removed – the usual solvents for these are alcohol (methylated spirit) or acetone. If you do remove the resist you should coat the board with a solderable lacquer or flux (see above). The only remaining tasks are to drill the board and cut it to size – again, we covered these areas above. Contact: Computronics Corp Pty Ltd 8 Sarich Way, Bentley WA 6102 Tel: (08) 9470 1177; Fax (08) 9470 2844 Website: www.computronics.com.au PC Board soldering tips While on the subject of making PC boards, perhaps a word or two about soldering PC boards would be in order. 99.9% of problems with kits are in the soldering of components, especially to PC boards. The biggest mistake constructors make is using a soldering iron which is too small for the job. A 10 or 15W iron used to be a popular choice by many, believing that it would minimise the risk of heat damage to semiconductors and other sensitive components. Believe it or not, it’s not necessarily so – in fact, it can be the exact opposite! Because the light iron cannot supply sufficient heat and because the copper of the tracks is such a good conductor of heat (taking heat away from the joint) invariably you have to leave the iron on the joint much longer – maximising the risk of damage! For the hobbyist, a much better choice is a 20-25W iron, either mains-powered or (preferably) low voltage, with a fine tip kept in bright, shiny condition. The best choice is a temperature-controlled iron or soldering station, where the iron is usually rated much higher (perhaps 60-70W) but only supplies the heat “dialled up”. Be careful not to use a temperature-controlled iron at too high a temperature. A bad choice is any heavy-duty iron because these are made to supply a lot of heat and can do a lot of damage to fine copper PC board tracks. Except in case of emergencies, you shouldn’t use a gas-powered iron to solder to PC boards – they too can develop far too much heat and they can be hard to control. Finally, always use solder intended for electronics work. Don’t buy solder from the local hardware store – you’ll probably end up with plumbing solder which contains a corrosive flux (it can eat through the thin copper tracks sometimes within weeks). Multicore electronics solder, preferably of a thin rather than thick gauge, is the way to go. That’s what you would normally be supplied in electronics kits. If in doubt as to which solder to buy, ask at your usual electronics dealer. SC MARCH 2001  77