Making Printed Circuit Boards.

This article first appeared in  CQ-TV magazine, issue 200.

In this article, I hope to demonstrate that home made  boards are actually quite easy and cheap to make.

What I’m going to describe is the method and equipment I  use myself. The photographs and screen snapshot show the prototype of a simple  test card generator from design through to testing stages. The principles you  see here can be applied to almost any design, whether one of your own or one  copied from a magazine or book article. I want to stress that my way of doing  things isn’t necessarily the most efficient or cost effective way, it just  happens to be convenient for me to work in this fashion. The equipment I use to  etch the copper is made entirely from salvaged parts and could be improved  considerably. I’m deliberately not providing any design plans for my home made  bits, I leave that to your own ingenuity but by explaining how I do things you  should be able to make something at least as good and probably better!
I’ll guide you through the stages in sequence:

1. The planning stage.

If you are working from an existing design, such as one  published in a book or magazine, skip the next stages entirely and go straight  to stage three.
Firstly, make absolutely certain that a PCB is what you need.  Often, especially for simple designs, it is easier to use Veroboard or a similar  product. Also consider how many copies of your board are likely to be needed.  The method I will describe allows you to make an unlimited number of identical  copies and if stored safely, the master artwork will be as good as new even  after decades of storage. If all you want is a one-off board, bear in mind that  the time it takes to design and make it could far exceed the time to assemble it  by other methods.
If you have convinced yourself that a PCB is the best  solution, start gathering information on the components your design uses. A PCB  is, for most designs, nothing more than a neat and repeatable gadget for  mounting components, so it is a prerequisite to know how big the components are  and where their legs lie with respect to their body outlines. I find it useful  to lay the parts out on the bench top and move them around to find the best way  of orientating them. The rules here are simple, components must not touch each  other or cross over each other on the components side of the board and the same  applies for the tracks on the copper side of the board. It is permissible to  cross tracks on opposite sides of the board if you use double-sided board but  there are complications in doing this. Double sided boards are made as two  single sided ones back to back and it takes great care to get the two sides to  align properly. It’s also very difficult to put copper inside the drill holes so  you end up having to solder both sides of the board to connect a track from one  side to the other. Take my advice, until you have mastered single sided boards;  leave double sided ones well alone!

2. The track layout.

This is the step where a little artistic imagination comes  into play. Other than being sensible about the placement of components that have  to line up with something else or could interfere with another parts operation,  you are free to do just about anything you wish. All you have to do is join the  appropriate component legs together with lines. It sounds a lot simpler than it  really is but with practice it isn’t too difficult. With all but the simplest  designs I use a computer with a PCB design program. The one I use is Autotrax  from a company called Protel . Its DOS based and by any stretch of the  imagination, ancient and rather crude. It does the job though and as a freebie  on Protel’s web site several years ago I can’t complain. There are many  similar and more up to date packages around these days and their prices are very  reasonable. I’ve seen someone design a PCB using the Windows Paint program but  I would seriously recommend using one designed for the job. What all the design  packages give you is a library of component "footprints", the outline  of the component when it’s mounted on the board. They allow you to manoeuvre  the outlines until you manage to fit everything into place and then add the  lines between the component ends. The real advantage of using a computer is that  the components and track lines can be erased or moved after being placed, its  akin to moving text around in a word processor but you move parts of a diagram  around instead. Some packages will even "auto route" the board for  you. All you do is provide a list of components and a list of which pins on them  connect together, known as a node list, and the computer takes over and puts the  track lines in place. I must confess to having little faith in auto routers, I’ve  seen some of them pick two points diagonally across the board, join them  together and then refuse to connect anything one side of the line to the other  because the new lines would cross the diagonal one. I always route boards by  hand, I find it gives better results although it can take quite a long time to  finish. Manually routing also gives you control over which connections can  sensibly lie next to each other, for example, you would have the sense not to  run a live mains track parallel to a microphone input signal but an auto router  wouldn’t realise this was a bad thing to do.

These are the basic do’s and  don’ts when laying out the board:

Start by drawing an outline of the board, this gives you a  visible boundary to stay within.

Remember to leave holes for mounting the board and leave a  gap around the holes for screw heads and mounting pillars.

If any parts must occupy a fixed position, for example a  switch that has a shaft that must pass through a hole in a panel, place those  parts first. They can’t be moved so shuffle the other parts around them.

Leave a space around any parts that get hot so air can  circulate and cool them. If heat sinks are being used, remember that no parts  can cross the space they occupy.

Leave sufficient gaps around parts and tracks that carry high  voltage so the chance of arcing is minimised.

Ensure track lines are "beefy" enough to carry the  current expected through them. I use no less than 1mm wide per Amp as a rule of  thumb but it also depends on the thickness of the copper on the final board.

Try to fill large gaps between tracks with copper area and if  possible connect it to a ground. This helps by providing extra screening, extra  heat conducting area and as a bonus it means you dissolve less copper in the  production stage so your etching chemicals last longer.

Finally, remember that PCBs are designed from the component  side. Imagine you are looking at the components and the PCB substrate is  transparent so you can see the tracks on the far side. It is normal, but  worrying until you see the reason for it that your design will be a mirror image  when you reach the next stage.

3. Transferring the track to copper.

From now on we start to see something resembling the finished  product. First, if you bought a huge expanse of board at a rally, cut it down to  a manageable size. It ideally should be a little larger than your finished board  will be. I always place a thin copper track around the outside edge of my  designs so I can use it as a guideline for cutting and filing the board to its  actual size.

If you use plain copper board, it should be spotlessly clean.  Buy one of those abrasive blocks and gently polish the copper with it until it  shines. Then use a solvent to make sure no fingerprints are left on it. I use  Ultrasolve ULS for this. Then under dim lighting, spray the board with photo  resist paint. The full instructions for this are on the paint can.
Remember that  from now on, the board is sensitive to bright light so don’t switch the lights  on to see if the paint has dried!

I use pre-prepared board. Most of the electronics suppliers  stock it. This is board already painted and then protected with a peel-off black  plastic coating, which also keeps light off the paint. It is much less hassle to  use the ready-made stuff and it keeps for several years if you keep it cool and  dry. You can cut it with the protective film still on so the off cut can be kept  for use in the future.

The process of transferring the design from computer screen  to copper is quite clever and very similar to making photographic contact  prints. What you need to do is print the track layout, in black, on to  transparent film. I use the sheets designed for overhead projectors which most  stationery stores can supply. Be careful though, make sure the brand you choose  is designed for use in your kind of printer. Usually, one side of the sheet is  glossy and one side is matt, it’s the matt side you print on or the ink will  form droplets on the surface. If, like me you use a laser printer, make sure the  film is suitable for laser printing or for photocopying. Both these devices use  heat to fix the toner to the film and the wrong sort of transparency will feed  into your printer but never come out again. Instead you will find a seriously  gummed up printer mechanism, you were warned!

If the printing was successful, you should see black areas  where you want copper on the board and everywhere else should be see-through.  Hold it up to a light and check for any unfilled areas, if there are any, use a  black fibre tipped pen to touch them in. Unfortunately, even the latest printers  tend to leave voids when they try to fill large areas.
When the ink has dried or the film has cooled down you can  get ready to do some photography. The aim is to expose the paint on the board to  a source of ultra violet (UV) light but only in the places we want the copper  removed from. The paint undergoes a chemical change when the light falls on it  and it becomes soluble in developing solution. Where the black pattern on the  film provided shade, preventing the light reaching the paint, it remains  insoluble in developing solution.

You need a source of UV light for the next stage. I use a  commercially made light box but it is very easy to make your own. You need a  box, even my commercially made one is only made from plywood.

The box needs to be big enough that your biggest PCB can sit  in its bottom. Mine is about 15cm front to back, about 30cm side to side and  about 10cm deep. This comfortably holds a standard Eurocard (100 x 160mm) board,  a size readily available. The top of the box must be clear glass and if possible  it should either have a hinged lid or you should have a removable lid that sits  flat on the glass. Fit some thin spongy foam to the underside of the lid. The  plan is to place the printed film on top of the glass and then place the light  sensitised board on top of it. The weight of the lid and the sealing effect of  the sponge should hold the board and film flat against the glass. Its quite  important if you hinge the lid that it doesn’t shift the board relative to the  film when you close it or the image will be in the wrong place or even off the  edge of the board after exposure. In the bottom of the box you need at least one  UV strip tube as a light source these are inexpensive and will last for hundreds  of boards. The tubes are available from Farnell, Maplin and CPC amongst others  but don’t get the ones for disco lighting or EPROM erasers, they are  unsuitable for this application. I use two 8W tubes, spaced about 5cm apart as  this gives a more even coverage of light. The tubes should be at least 2cm from  the glass and each will need a starter and choke. You can add a mains switch or  timer if you want but I find that pulling the mains plug out works just as well.

The film goes on the glass with the printed side upward, the  board goes on top of it with the painted side downward. This is why the image  appears backward on the film; it is re-reversed when placed upside down. Doing  it this way means the printed surface is in direct contact with the paint so  there is no space for the UV light to scatter and blur the image. Don’t forget  to remove the protective plastic coating if you use pre-prepared board!
With two tubes, my light box takes about four minutes to  fully expose the board, your mileage may vary depending on the number of tubes  and quality of glass.

4. Developing the negative.

This stage is simple and quick. You need a plastic tank or  tray and some developing solution. Like all the other tools and chemicals, this  is easy to get hold of. I buy Farnell’s ready prepared solution, which is  already optimally diluted, but I’m almost certain it is actually caustic soda  dissolved in water. Remembering that the board is still sensitive to bright  light at this stage, remove it from the light box and place it copper side up in  the developing tank. Add enough developer to just cover the surface of the board  and then swish it gently from side to side so the developer is constantly moving  over the paint. Be careful with the chemical, it is corrosive and prolonged  contact with skin can cause irritation. Be especially careful not to splash it  into eyes where it can cause severe damage very quickly. After about a minute or  so you should see the pattern of tracks starting to appear in the paint and  after about five minutes the tracks should be clearly visible as paint areas against the plain copper background.

It's safe to leave the board in the developer a little longer  than needed for complete removal of the exposed areas paint but don’t leave  it more than about ten minutes or the unexposed areas may also start to be  attacked. Remove the board and rinse it under plenty of cold water. The  developer can be saved if you wish and re-used at least one more time. If  re-using it the time for development may be a little longer than before as the  solution will be partly neutralised. If using the same tray for the next stage  you must wash it thoroughly, the etching chemical and developer combine to make  an unpleasant brown sludge.

5. Etching the board.

At this stage you should have a copy of your computer artwork  painted on a bare copper base. The paint will protect the copper from being  dissolved in the etching solution so the copper underneath it will remain and  the rest will disappear.

You can do this in normal lighting; the board is no longer  sensitive to light after the developing step. Put the board in a tank or tray of  etching solution with enough present to cover the board to a depth of about 1cm.  I use the same tank as before which was originally a container for chocolates.  The solution needs agitating so fresh chemical (Ferric Chloride solution) passes  over the copper. I speed the process up by placing the tank in a sink of hot  water. The increased temperature speeds up the etching. I also put the lid back  on the tank so my "stirrer" can work. This is actually a small 12V  motor salvaged from a scrapped video recorder. It’s screwed to the top of the  lid so its shaft faces downward. The shaft is glued to a plastic rod which has a  paddle fixed to its other end. The motor is fed via a resistor to slow it down,  it should rotate about four or five times a second. Any faster than this creates  the risk of it splashing if the lid is raised.

The intention is simply to keep the solution on the move. In  hindsight, I would have fitted two motors and paddles to give better  circulation. The etching takes about ten minutes, lift the board occasionally to  see if it is finished, this time it isn’t wise to keep it in any longer than  necessary because the etchant will start to eat under the edges of the paint and  reduce the track widths. When finished, thoroughly wash the board in cold water  and return the etchant to a plastic bottle so it can be reused. Depending on the  amount of copper removed, it should be possible to use it about ten times.  Ferric chloride doesn’t only dissolve copper; it likes most metals, including  stainless steel so take care to keep it from splashing over kitchen sinks.

6. Drilling lots of holes.

You now have a bare substrate with copper tracks in the right  places and paint still covering the tracks. Leave the paint there for now; it  protects the copper from oxidising.

You can either cut the board to size now or after drilling. I  drill first while the board is a little larger and easier to handle. Drilling  the holes is laborious and a bit messy. I use a small low voltage drill in a  mini-pillar mount. Put a block of wood beneath the board to support it then  manoeuvre the board beneath the drill bit.

The optimum size of hole for chip legs is 0.6mm and for  resistors 0.8mm but larger holes can be tolerated if there is enough copper land  surrounding them. It takes quite a long time to drill the holes and a quantity  of dust will be created. Its best to use a vacuum cleaner after every 20 or so  holes to suck the dust away as it can cause skin irritation. No matter how many  times you check all the holes, I guarantee there will be one you missed until  you try to fit a component through it!

7. Cleaning up.

Wash the board again after drilling. It is safe to leave the  paint in place until you are ready to start assembling the board. The paint is  usually "solder through", it acts like flux and burns away when you  apply a soldering iron to it. I find it leaves unsightly marks and does little  to improve solderability so I use Ultrasolve ULS to dissolve the paint just  prior to placing components.

8. This step is entirely optional.

You can if you wish, coat the copper with tin to prolong its life and make it easier to solder to. You need another tank and some tin plating  solution for this. Remove the paint and clean the board with ULS then  immediately drop it in the plating solution. Leave it for several hours until a  silvery film has settled on the copper. Remove the board from the solution and  wash it in water. Tin-plating solution is very expensive and although I have  plenty, I rarely use it. Making a PCB without plating takes about one hour from  computer print to etched board, adding the plating stage makes this much longer  and doesn’t really offer many benefits. The choice is yours of course. A spray  of plastic lacquer or conformal coating does just as well and is quicker than  plating.

Using existing designs.

Copying designs from magazines or books is easy if you use  the photographic method I’ve described. All you do is either photocopy the  artwork from the page onto film or use a scanner and printer to do it via  computer. You may need to adjust the scale of the copy if the article didn’t  show it at full size. Scaling is normally a feature of photocopiers and print  programs so this shouldn’t be a problem. Once you have the design on film,  simply go to the UV exposing stage described earlier.

I hope I’ve managed to persuade you that no black magic is  involved and nothing more complicated than pouring a few chemicals is demanded.  It may sound a little complicated at first but in practice the whole operation  is quick and gives professional results. The investment in equipment is very  modest and it can be used time and time again. What you don’t make yourself is  easy to find in component catalogues and the chemicals are readily available.

Warning!

The chemicals used to make circuit boards are quite safe to use if handled with respect and stored safely. However, as with most chemicals, misuse can create a hazard and serious injury. The developing solution in particular is caustic and will cause severe damage if it gets into your eyes. Please read the safety instructions on the chemical container labels and be familiar with the actions necessary should an accident occur.
You are advised to wear goggles (eye protection) and waterproof gloves while handling hazardous substances.

I've been making boards for over 30 years without incident but I cannot stress enough that accidents can happen and you should be prepared for them. I do not accept responsibility for other peoples actions!