You are the visitor n.
Best view with 1024x768 |
||
You are the visitor n.
Best view with 1024x768 |
||
This document has the aim to place at disposal of whoever wishes the necessary techniques to obtain with a few means the realization of printed circuits using photoetching. The quality of the result, though if obviously not just equal to manufacturer's one, is indeed very good. The subject we are going to approach will start from the basic principles arriving to the practical realization of a printed circuit; as to the exposure method of the breadboard, we immediately say that it's not necessary at all to have a contact printer, usually rather expensive unless you decide to build it by yourself. What will be reported here should however be valid also in the case you own one; though exposure by means of UV or Wood light tubes has not been personally tested, some e-mails we received seem anyway to confirm it. The material contained in these pages is the result of tests and experiments executed in first person by the undersigned. The only contribution I have to recognize to other people's work, drawn from the net, is only related to the use of a solar light bulb for the exposure and to the very basic methodology for the determination of needed times. I have personally tried, infact, the use of other means of lighting, for instance direct solar light, but with such technique the result is strongly dependent by rays' inclination towards the breadboard, moment of exposure, atmospheric conditions (haze, fleeting clouds, etc.). As instead to the advanced methodology to determine the correct exposure time, the power of the light bulb, the type of support, the type of needed material to create the master, the number of copies of it, some advices and devices used in various phases of the entire process, the experimental parameters for exposure, development and etching, the advices on drilling and on the realization of a personal layout, I guarantee the absolute originality of the material, though obviously not excluding that on the net it could be possible to find something analogous. Though I have tried to be as more scrupulous as possible, and I have therefore included also advices to take care in some phases of the process, I don't assume any responsibility for possible direct and indirect damages if you decide to use the informations you have got from this article. The photoetching of a printed circuit needs first of all a copper-plated breadboard covered with a layer of photoresist, that is a photosensitive material able to change its chemical properties when exposed to light. In short, the photoresist layer has the function to protect the breadboard's copper during the chemical attack for the removal of the metal in excess; if it is possible to make so as to maintain such protecting layer only where you want tracks to exist, these will be conserved, so obtaining the wished layout. To obtain the selective removal of photoresist, the entire breadboard's surface has to be exposed to the light, after, of course, applying on it a mask, for definition opaque in some areas. About this subject, it's first of all necessary to make a distinction: on the market you can find breadboards presensitized with positive photoresist and breadboards on which, instead, negative photoresist is applied. In the first case only areas in which the copper has to be removed have to be exposed to the light, while in the second case it's the contrary. Usually positive photoresist breadboards are used, and those we are going to deal with; masks to use with them are opaque in correspondence of the tracks and transparent in correspondence of the areas in which the copper has to be removed. What follows, mutatis mutandis (changing what has to be changed), is generally valid also for negative photoresist breadboards. We immediately say that the greatest part of layouts that you can found on electronics magazines is apt for positive photoresist (I have personally always met this type); of course with a computer it is very easy to make the image negative and to obtain this way a suitable layout for the other type of photoresist. Beyond buying already presensitized breadboards (protected by an opaque film, of course), it is possible to sensitize normal breadboards by means of spray photoresist. It's surely true that, realizing a certain amount of circuits, the cost is lower, but I have personally never used such technique as it needs a rather accurate distribution of photoresist, a totally lacking in dust atmosphere, long times for drying (some hours or even one day). More, spray photoresist should be used within short times, usually about one year from production; this is also one of the reasons of why you can't find it in every electronics store. The realization of layout mask is different if the image of the layout is available on paper, as an example on a magazine page, or if it is instead possible printing it, maybe in the case you have designed it by yourself. If the layout is on paper it's necessary to make a photocopy of it on an acetate (transparent) sheet, taking care to obtain a good contrast; the photocopy should therefore to be the darkest you can in the parts that will have to be preserved from light, but also the most transparent you can in the parts to expose. A single acetate sheet is not enough: I advice three of them overlapped. It's important to check that the layouts are exactly with the same dimensions, that is they have to coincide perfectly: if this doesn't happen the quality of the final result is partially compromised, as some tracks could result too wide and overlap to adjacent ones, with easily imaginable effects on circuit behaviour, or, above all the thinest ones, could be not present at all because of insufficient mask's opacity in that area. If you have the possibility (or the necessity) to print the layout, the process is not very different from the just discussed one. In what follows we will refer to ink-jet printers, but basic principles should be valid also for plotters and laser printers. You have to use acetate sheets for printers; they are different from those for photocopiers (the first ones, for instance, are not perfectly transparent); the printable side can be easily identified as it's slightly rough in order to allow the ink to join to the support. TAKE CARE! Don't use acetate sheets for ink-jet printers in the photocopier, as you could seriously damage it (a person of my acquaintance has tried this interesting experience; it's not very nice seeing acetate melting inside the photocopier). If the printer allows it, it's necessary to select the type of support in printer settings, as printing on acetate should usually occur at a speed lower than on paper. Also with printed sheets you have to overlap more than one copy: I personally advice to print two copies of the image with double printing each; if, due to the support, such modality is not allowed, it is necessary to insert again the slide in the printer after at least a pair of minutes to let the ink dry. If all works as planned the new print is perfectly aligned to the previous one (you have obviously to pay attention not to leave clearance between lateral guides and the acetate sheet, of course if the printer allows it). In this way I personally haven't any problem to obtain two or more identical copies. If the second print on the same sheet produces an image not overlapping the first because it's not easy a perfect alignment of the sheet, it is then possible to try with four single-print copies (better five or six). With both types of support, as previously said, you have to overlap acetate sheets: if possible, while cutting out the parts to use, I advice to leave a transparent area at least some centimeters wide externally to layout's area, so to extend also outside of the breadboard to expose. One after each other, overlap acetate cuttings taking a very great care to tracks' coincidence. If cuttings have dimensions greater than the breadboard's ones, it is possible to use common clips, better a stapler, otherwise glue of the kind of " Super Attak" can be used, obviously externally to the layout. Take care: the precision in overlapping and joining masters has a fundamental importance for the result. If you notice that a master has been printed badly don't hesitate to print another one in substitution. I advice to protect overlapped sheets with a transparent acetate sheet (for photocopier) on every side, so that the printer's ink (or the photocopier's toner) couldn't come in contact with breadboard's photoresist, and above all that the master itself couldn't be scratched or soiled accidentally. If you have a contact printer, you probably have a certain familiarity with exposure, and in such case you could do without reading a part of what follows; if instead you don't own a contact printer, or if it's the first time you use the etching method we are talking about, it could be useful to read this and the next chapters. As previously said, the contact printer it's not at all indispensable to obtain the chemical transformation of photoresist, as this one is moderately sensitive also to visible light, in particular to blue light. It is therefore possible to use for exposure, with some devices we will treat later, a common light bulb with blue glass, commonly known as " solar ". In comparison with the use of the contact printer, the only disadvantage is the necessity of remarkably longer exposure times: compared to very few minutes for UV or Wood light exposure, with light bulb times between ten-twenty minutes and a maximum of two and half - three hours are needed; results, however, are indeed very good. First of all you have to get one or two glass sheets of the size you prefer: I advice against too small glass sheets, also when exposing very small breadboards (some cm of side), as the upper glass sheet has above all the function, with its weight, to hold the master firmly in contact with the presensitized breadboard. If the glass sheet is too small, or if it is absent, nearly surely the master will not join perfectly to the breadboard originating blurred tracks, and therefore results similar to those derived from an imperfect overlapping of the masters. As a purely indicative example, suitable sheets have approximately a thickness of 5 mm and dimensions of 20 cm x 30 cm. For exposure you have to know correct times: you will be able to get them with one or at most two-three tests in the way that will be described later on. It's advisable to execute exposure in a dark or poorly lightened room, for instance with an abat-jour some meters far from where exposure to blu light will take place place; it is however advisable to have the solar bulb as the only light source during exposure, but it is possible without any problems to switch on even a chandelier for some minutes during the deposition of the master and of the upper glass sheet, or during the exposure itself, without any collateral effects and without shortening sensibly the needed time. What follows is valid for single-side breadboards; in case you have a breadboard to be exposed on both sides, to create a reference it is possible to act in at least two ways. In the first one, you should drill some holes on the breadboard, with the two sides still protected, in the center of some pads or vias, taking care to obtain quite vertical holes; if not so, in fact, the pads on the two sides will not be aligned. The second way, mainly appliable when the master is self-made, is designing or just creating some signs on the two masters in correspondence of the breadboard's corners; it is worth to point out that some ones are not regular at all, and breadboards of the same declared size and of the same brand can be different for some points of milimeter. It is also needed to take a great care not to scratch the already exposed surface; I therefore advice to expose the first side without removing the protective film from the other one, and, when having to expose the second surface, to protect the first one with an acetate sheet for photocopiers or simply with the master itself (already protected), and maybe with a black thin pasteboard between the acetate sheet and the basement (possibly the lower glass sheet). Put the (eventual) lower glass sheet on a flat support, and if possible (but it's not absolutely necessary) cover it with a sheet of black thin pasteboard in order to avoid reflections on the more external part of the breadboard. Place the breadboard on the glass sheet (or the thin pasteboard) after you have removed the protecting film; it is now possible to place the master on the breadboard, taking care that no evident dust traces are present neither on the breadboard itself, nor on the two sides of the master, nor on the glass plate with which you will fix all. It could appear superfluous, but it is good to remember to pay attention to the side to overlap the master: if (as it usually happens) it represents the layout viewed from the copper side, then the printer ink (or the toner of the photocopier) must be in the opposite side as to photoresist; viceversa if layout is seen from the components side. It is now possible to place the solar light bulb with its center approximately above the breadboard's center; if you haven't done it previously, it is advisable to find a position such that no shadows due to the small metallic supports of the filament are present on breadboard's surface; to such an aim the light bulb can be turned until finding one satisfactory lighting. This operation is not truly necessary, but advisable, as if the exposure is globally hardly sufficient a shadow line could determine unintentional tracks in correspondence of it. The correct distance depends on the dimensions of the breadboard; if you don't want to have more than a parameters' table you could use in all cases a relatively great distance, suitable also for not too small breadboards (for instance 15 cm x 15 cm). If in fact these are not only a few centimeters wide (for instance 5 cm x 5 cm), and light bulb is placed very near the surface to expose, the zone of minimal distance will receive an amount of light much greater than some centimeters farrer, being therefore impossible to obtain a correct exposure for the whole breadboard (overexposure of the central part or underexposure of the peripheral parts). As an example, parameters currently used by the undersigned are here reported: Solar 100 W light bulb on a simple lamp-holder without reflector; distance of the filament from the center of the breadboard: 25cm; exposure time: 2h 30 '; three master in double printing overlapped and protected on both sides by a transparent acetate sheet. It is obviously possible to use a lamp with a reflector, obtaining one greater lighting on the area under the light bulb, and therefore shorter times at parity of distance from the breadboard; problems are only born in the case breadboard's dimensions are not too little, because of the smaller peripheral lighting compared to the central one. It's not possible, at the moment, to personally test Wood light, but, with tubes with a power of ten-twenty Watt, and with distances of about five-ten centimeters between light source and breadboard, exposure times should be, approximately, between three and fifteen minutes. TAKE CARE OF YOUR EYES! Please remind that correct exposure times depend also on the company supplying the photoresist; changing mark, if you don't get the usual results it could be useful to carry out the exposure and development tests again. Determination of exposure times In order to determine the correct exposure times it's necessary to execute some tests. You have to cut a breadboard of small dimensions, long and narrow, for instance 2 cm x 8 cm. Once you have applied the master on the breadboard, place the glass sheet on the whole and a sheet of black thin pasteboard on it, covering the breadboard; take care that the thin pasteboard joins well to the glass. Ideally divide the breadboard in a certain number of zones, for instance eight-ten; it would be better to have a master with tracks of different widths in each zone, for instance 10 mils, 20 mils, 50 mils, 100 mils (1 mil=1/1000 inch; 10 mils=0.254 mm). Discover the first zone, and begin exposure. After a certain time move the protective black thin pasteboard until enlightening also the second zone; do the same periodically. In the end develop the breadboard and observe the results you have got. As to the solar light bulb, with parameters by me used and previously indicated, you can as an example enlighten the first zone for a quarter of hour, then move the thin pasteboard, and you should repeat this operation every ten minutes. With Wood light, not personally tested, you could try with three minutes for the first zone, then move the thin pasteboard, as an example, every two minutes. In such a way, as the exposure time is known for each zone, it will be easy to determine the correct time in such lighting conditions. With solutions at 10 or 20 g/l, exposure has been correct if, after some tens of seconds, the design of tracks is clearly visible and, after some minutes, the photoresist has been removed correctly (as said previously, copper appears perfectly shining); if the photoresist trends to dissolve in a too short time, the exposure has been excessive; viceversa if the photoresist trends to remain and the design of the tracks appears with a little contrast with regard to the rest. If, last possibility, the design of the tracks is insufficiently visible, and the photoresist trends to be removed everywhere, then the master is insufficiently opaque: in such a case it is necessary to overlap a greater number of copies of the master itself. A still more meaningful test could be carried out using a test breadboard, as the one previously described, just under the center of the light bulb, and at least another one some cm far from it (for instance 8-10 cm). If the exposure of both is correct with the same times of development, then it will be possible without any problems etching breadboards until 16-20 cm of side; if this is not the case, you can use only smaller breadboards or you have to increase the distance between light source and copper. The development of the exposed breadboard has the function to remove photoresist where the removal of copper has then to be obtained. In the case of positive photoresist it will obviously be the part exposed to the light to be removed, viceversa in case of negative photoresist. The development solution is simply sodium hydroxide (NaOH). Take care: it should be known that sodium hydroxide can provoke serious burns, though if in effects the solution we need is not very concentrated and its effects on the skin are substantially very limited; I ignore which could be effects on eyes; I WARMLY ADVICE TO USE GLOVES AND ABOVE ALL PROTECTION GLASSES WHEN MANIPULATING SODIUM HYDROXIDE, BOTH AT SOLID AND LIQUID STATE. In every case please read warnings on the packaging. You could find sodium hydroxide in small papers in any electronics store, but also in ironmonger's shop or in supermarkets (at really lower prices). The cost of a 10 grams paper shouldn't however exceed a few dollars. To prepare the solution you can use the common water of the tap; in this case, the only problem is related to the repeatability of solution's properties, and therefore times of development, if the concentration and the type of dissolved salts are not constant in time. These last ones, in fact, can determine, in different times, a different PH (>7) of the development solution. It's so useful, but not really necessary, using distillate or demineralized water. I personally use a plastic container with 500 mililiter of demineralized water in which I dissolve 10 grams of sodium hydroxide; 5 grams of such a base can be good the same, the only difference is an increase of development time; the first times it could be useful to use the less concentrated solution in order not to leave unintentionally the breadboard in the bath for an excessive time. After exposure, dip the breadboard in the solution by means of plastic pliers, for instance; the important thing is anyway not to scratch its surface. If the photoresist has been exposed correctly, after some seconds from the immersion you should see a darkening of unexposed parts, and at the same time exposed photoresist starting to dissolve. In order to assure a uniform development you should shake the breadboard or the container; if the breadboard is a double-face one, you should avoid, for obvious reasons, to lean it on the bottom of the container. Periodically, let's say every 30 second-1 minute, it is advisable to extract the breadboard from the container, to wash it abundantly with water, taking care that eventual drops of solution do not blot dresses or other things, and also chromium-platings (though I don't know if it is possible to damage them), finally to examine it. The development is complete when the copper that has to be removed appears shining, without any opacity (the residual photoresist), while the design of the tracks is clearly visible; above all, to the touch it should be possible to feel a step due to the photoresist that protects the copper parts to be conserved. If the development is not complete the breadboard has to be dipped again. It is necessary taking care not to exaggerate with development; if in fact you persist too much, also some unexposed parts of the photoresist could be removed. Keep in mind that if the photoresist to remove has not been dissolved completely, but only a thin layer remains in some points, it will then be removed from the copper attack solution. If after ten minutes of immersion in the solution at 10 grams/liter, or after three-four in the doubly concentrated one, tracks still do not appear, or appear with a very little contrast with regard to the rest, it means that exposure has been insufficient, and though it is possible to try with a prolonged development, it would be better to use another breadboard, assigning the first the role of a normal one (pen, transferables etc.) after you have removed photoresist, operation that you can carry out with denatured alcohol or acetone in dependency of its composition. Once the development is complete you should examine the breadboard to be sure that there are not any considerable scratches on the photoresist; otherwise it could be necessary to retouch the tracks in such points with the pen for printed circuits in order to avoid breaks on tracks. The solution for copper attack can be as an example ferric perchloride (I have personally always used that). Keep in mind that ferric perchloride, beyond being probably not too much beneficial for skin, has the power to spot almost indelibly; I therefore warmly advice to use gloves and protecting glasses and to take anyway a great care. Dip the breadboard in the solution, also in this case as an example with plastic pliers, for a time from some minute to some tens of minutes; this depends mainly on how much the liquid has been used and on its temperature. On the market you can find heating coils that allow you to shorten the etching time; you can also find containers with a system of diffusers that allow to air coming from a small compressor to gurgle in the solution; also in this way it is obtained a remarkable reduction of development time. These two accessories are not necessary at all, as they don't condition the result's quality. As for the development, also in this phase you should take care to guarantee a quite uniform erosion on the entire surface, therefore periodically slightly shaking the solution and moving the breadboard; this is valid mainly for double side breadboards, that for obvious reasons have not to be leant on the container's bottom, but have to be kept suspended some way. You should periodically extract the breadboard in order to control the degree of the process' advance: as soon as all the copper in excess has been removed wash it very well and then dry it. All is done! If instead you note that, also after a long immersion, in some points the copper is not dissolved, this could mean that the photoresist has not been removed completely; if so, you could try, with a great delicacy, to execute this operation with some tools, as an example a screwdriver, taking care not to scratch also the photoresist that must protect the tracks. It should be possible, but I advice against it, trying to dip again the breadboard in the development solution. When you have completed these operations you can remove photoresist with alcohol or acetone, or with very fine grain abrasive paper (grain 500 or greater), though the photoresist is generally weldable and it could protect the tracks from a slight and unavoidable oxidation. At this point the breadboard is nearly ready: only left to execute the drilling operation. It would be better to own, to such an aim, an appropriate high speed drill (more than 5000 rpm) for breadboards or for modelling, or also a column drill. Drilling is easier if where holes have to be done the copper has been removed during etching; if it's not this way, maybe because in the layout holes are not reported, it is useful to use a cross screwdriver or other similar tool to graze the copper where you have to drill the breadboard; in such a way the tip will not slip on the copper. If the drill is a variable speed one, above all if it's neither a high speed nor a column drill, it is very convenient to put the tip in correspondence of holes to make and to start drilling at a very low speed, until the tip doesn't slip anymore. The diameter of the tips to use usually varies between 0,8 and 1.5-2 millimeter for largest pins; generally a 1 mm tip is very good for the greatest part of holes, still better a 0.8 mm one. Take a great care, above all with normal drills, not to act with an excessive pressure on the breadboard: thin tips are very fragile, they break very easily and can sometimes jump away (take care of your eyes). Just to finish, it's important to take care that during drilling a part of copper still remains around the hole, to allow then an easy soldering of pins. If you have the possibility to create the layout of a circuit projected by yourself, the following advices could be of some usefullness. Keep in mind that, referring to printed circuits, very often as a length measure unit "mils" are used (1 inch=2.54 cm = 1000 mils), and that the distance between adjacent pins of many types of components is exactly expressed in an entire number of tens of mils; as an example, the 2.54 millimeter step of very many integrated circuits' pins is exactly 100 mils. As to the the distance between adjacent tracks I advice not to make it less than 20 mils (0.5 millimeter approximately); in case that on the tracks high frequency signals should be present, interaction effects should probably be estimated. For feeding tracks or for those which should carry currents higher than some tens mA, you should use, very approximately, 40 mils (about 1 millimeter) every 100-200 mA. For signal tracks I advice not to go below 20 mils, though if with the previously described method I have obtained perfect test tracks 10 mils wide. In order to guarantee an easier welding I advice to realize the pits with a 70 mils diameter or width, both for discreet components and for chips with 100 mils step pins; in order to make drilling faster and more precise I advice to design also holes in the layout, with about a 30-40 mils diameter. What I have said is valid for the greatest part of components, but there are very common ones demanding pits with dimensions greater than those indicated, as an example the 78xx series in TO-220 case. Hoping I have been useful to you, only left to wish you a good work! Fabrizio Iacopetti
|
| Last update: 8 July 2002, 18.45 |
