30th October 2009
It may seem like an odd subject for me, but for the children's project I have got to find a way 0f producing PCB's very cheaply and I wanted to investigate the possibility of applying the acid resist, solder resist, artwork and solder paste using the screen printing process. I talked to several professional printers and they all came to the conclusion, that if they could print barcodes then tracks and pads should not be a problem. So I have made up a classic screen printing board, stretched my first screen and the photo sensitive emulsion arrived yesterday, so I only have to make a screen emulsion trough coater and a squeegee and I can give it a try. What I am aiming at is printing the following .....
The board is standard 1/16th inch thick fibre glass board with copper on one side of the board. The idea is we use screen printing to print the tracks and pads onto the board. When the ink is dry we them immerse the board in concentrated ferric acid and any copper that is not protected by the ink is etched away. We then wash and dry the board and remove the ink printing with a burnishing block, drill all the holes and we have a a PCB finished and ready for assembly. Normally we would use a PCB material that has been coated with a light sensitive emulsion and we produce the tracks and pads using a photographic method which is a lot slower and more expensive that screen printing .... providing it works !.
Until I screen print the first board I will not know if I will have to make any changes in the artwork, for example making the tracks narrower and the pad holes larger. There is also the question of the number of passes with the squeegee, squeegee angle and pressure. I will be using 43T polymer mesh for the first experiments because that was available via EBAY. 43T is equivalent to the American mess size of 110 and other mesh hole sizes go up to at least 300 so there is plenty of scope for higher resolution, if required.
Effectively the squeegee pushes the ink paste through the holes in the mesh, effectively extruding it to produce a series of ink dots on the material being printed on. In practice the ink tends to spread out under the squeegee pressure and the dots all join together to give a solid layer of ink (I hope). It is because of this ink spread that I may have to make tracks and pads a little narrower to avoid shorts on the finished board. It will be a case of suck it and see.
By changing the screen, it is also possible to print other colours and text. Test and artwork on a finished board makes them look a lot more 'professional' and at the same time the text can be used as essential labels for components like socket identification etc. In the photograph above, the text you see will actually be printed on the other side of the board in white. The original idea was that the student could assemble their own board. To make this easier and eliminate unwanted solder shorts it is normal to cover everything on the board with a layer of solder resist ink and this can be printed direct onto the board. For surface mount boards, solder paste can be accurately applied to all pads in one pass of the squeegee, so screen printing could be a very quick and cheap way of producing PCB's. We shall see !
31 October 2009
Screens and printing frames.
Since I did not know what I was doing, it made sense to copy the classic printing frame that contains the screen. The screen is the hinged part that moves up and down and has the screen material stapled to it's wooden frame. I made the inside dimensions about A4, which is a small size for screen printing, but then again I am only printing small boards
Before the screen mesh can be used it has to be stretched and stapled to the frame and any spare cut off. The mesh shown in the photograph is 43T and made of polymer although other screen mesh materials can also be used such as organdie.
Coating the screen with light sensitive emulsion.
I will be using the light sensitive emulsion method of making my stencil. The emulsion comes in two parts, the emulsion and a sensitiser. These two components must be mixed together in a dark place, or one fitted with a photographers 'safe' lamp. As soon as the two components are mixed the emulsion becomes light sensitive and must not be exposed to light as it will harden the emulsion. The mixed emulsion can be stored in a dark refrigerator for up to three months. I will probably only mix up half of my pack.
In the old day's a film of emulsion was spread onto the mesh using brushes or a squeegee, but today it is more common to use a 'trough emulsion applicator' which gives a good even coat without a lot of mess or waste. It also removes excess emulsion from the screen mesh as it coats, giving a continuous and even coverage. The screen needs to be coated on both sides to make sure the mesh holes are filled.
Here is one I made earlier today from some scrap PVC tubing. I may need to increase the angle on the left hand side, to make it a little easier to use, experience will tell. The trough is 5 inches long, which will produce a coat 2" wider than the board to be printed. The internal width of the screen frame is about 8 inches and the width of the PCB 3 inches.
Once coated, the screen must be kept in a dark place to dry out.
We will need a 'positive' film of whatever we want to print. In my case, the first one will look something like this .....
in the middle of an A4 size transparent sheet. In fact, as the board is so small I am going to put three of these images on the same positive, so that I will be printing three boards at a time. The positive image is printed onto transparent ink jet printer, 'overhead projector' transparent sheets. They must be sheets intended for ink jet printers and not LASER printers as the ink will never dry properly on the latter. We also need a good dense black image so what I do is pass the film through the printer several times. If the printer is not accurate enough to do this then we can print out say four sheets and make a sandwich of them. How well the sheets line up with each other is called 'registration' and we will be looking at that later.
Exposure test strip.
The screen printing process has several variables, so before we can actually start printing we first have to establish the optimum exposure time for a given lamp, operating a given distance above the emulsion and also the attenuation factor of the sheet of glass, used to hold the positive in close contact with the emulsion. The way we do this is to make a standard exposure test strip. A test area of emulsion is 90% covered by a light proof material and an exposure made for one minute. The shielding material is then moved back so that it is now only covering 80% of the test area and the process continued until that last 10% of the area has been exposed. The first 10% test area has now been exposed for 10 minutes, the next for nine minutes and the last for one minute. The test strip is then washed out and visually inspected to find the optimum exposure time for that particular set of hardware. If ten minutes exposure is not enough then a second test strip can be made with longer exposure time steps. Obviously a 1000 watt lamp will produce quicker results than a 300 watt lamp. Make a note of the optimum exposure time and you are now ready to expose your first screen.
The 'UV' exposure lamp.
Any purist will say that a special 'UV' lamp is required to harden the exposed emulsion, but the truth is that any quartz halogen lamp produces UV light at 405 nm and below. The only difference is that at 405 nm, it is 90% down on the spectrum peak. Which simply means that either we have to use a larger lamp, or a longer exposure time. Probably the cheapest and most useful lamp to use for this purpose is a standard 300 watt. security lamp, with the UV glass filter removed. If you decide to leave the filter on, you will need a longer exposure time and it also makes for a heavier assembly. To get an even light coverage on the base board, the lamp will probably have to be between 18" - 24" above the base board. It would also simplify life if a simple timer was fitted.
Exposure of the light sensitive coating on the screen.
The next step is to expose the light sensitive screen coating to UV light, with the parts we want to lose, covered with the black parts of the positive. To do that we fit the positive inside the frame, onto of the emulsion coating. Any part of the emulsion that is covered by the black parts of the positive, will not be exposed to the UV light and will stay soft. Any part of the image that is transparent will allow the light to shine onto the emulsion and it will harden. After exposure is complete we then 'wash out the screen' with a jet of water and that will leave holes in the screen that will allow the printers ink to pass through at those points. So in the above case there will be holes where ever there was black on the positive. We then let the screen dry and it is ready for use. It cannot be seen on the photographs but there is one long hinge rod that goes through both hinges and to remove the screen for cleaning etc, we simply pull out this rod. When we reassemble the frame it will always go back into the exact position every time (registration again).
Taping up the frame. We only want ink to pass through the screen holes .... and no where else !. So after the screen has been exposed we apply packing tape (the brown stuff} around the edges of the frame and over any part of the mesh that is not covered by emulsion. This is done on both sides of the frame.
After exposure, those areas which were covered by the black image on the positive remain soft, while those areas exposed to the UV light harden. Washout is achieved by using a water spray or jet to wash the soft emulsion out of the mesh holes. This is how we produce the 'stencil'. Exposure time is really a compromise between under exposure which would not completely harden the exposed emulsion and over exposure which can cause problems. Nothing is completely light proof and if we overexpose it is possible that stray light can find it's way under the positive and harden hidden parts of the emulsion thus preventing it being washed out of the screen without using a high pressure jet of water. The end result is that normally the emulsion on a washed out screen is never as hard as it could be.
Post washout exposure.
It has been found that if after washout and drying, the screen is subjected to another exposure period, the emulsion fully hardens and extends the working life of the screen by up to 15%. A further improvement can be obtained using the application of a curing chemical.
The squeegee has a rubber blade about 9mm thick and substantial handle in order to enable pressure to be applied.. They come in a range of widths and generally on the smaller frame sizes the squeegee needs to be 1 - 2" narrower than the inside of the frame. Two cross sections are available, "D" (rounded) and flat. I would suggest a flat blade for a beginner as it is easier to re-sharpen the edge (on sandpaper).
My 5" squeegee.
Normally people would be printing on paper or fabric. In my case I will be printing on a board. Either way we first need to charge the mesh with 'ink'. Actually the ink is more like a paste. There are many good videos on youtube that show this process and it would be a good idea to view them. In practice we 'spoon' some ink past across the screen at one end and then using the squeegee, with the right angle and pressure draw the squeegee towards you and that will fill all of the 'holes' in the screen with ink and transfer the image onto whatever material you are printing on. One then lifts the screen and pushes the ink back to the top of the screen with the squeegee and you can now place the next item to be printed onto the baseboard and repeat the process at a rate of one every few seconds. The ink is still wet at this point, so the items must be pegged up to dry in a safe place. A few pegs on a clothes line idea works quite well. If you have been printing a fabric then you will need to 'flash dry' the ink to make it permanent and waterproof. This would be the case if one was printing tee-shirts for example.
At the end of a printing run, the screen needs to be removed from the frame and washed to remove all traces of ink from the mesh holes. If one takes care of the mesh frames then one should be able to print off hundreds or thousands of items before the emulsion begins to wear. When it does, just use the same positive and prepare another mesh.
Old emulsion can be removed from a mesh with lacquer thinners and the frame used over and over again.
It can be seen on the photographs that the hinges are mounted on a block of wood secured to the base board by three bolts and wing nuts. This is the standard design approach and it enables objects of any thickness or height to be printed .... ie boxes, front panels etc.. One just uses longer bolts and include spacer blocks to bring the mesh frame to the new height. Most printers leave a small gap between the material to be printed and the mesh. Two or three mm is sufficient and this introduces 'snap'. During the printing pass, the squeegee presses the mesh down onto the material and as the blade moves on, the mesh snaps back away from the material.
Over Printing Layers.
The side of the PCB board that has the tracks and pads on it is known as the 'solder side'. The reverse side contains the 'through hole' components and is called the 'component side' and often has component positions and component numbers printed on it, usually in white. The solder side of the board may also have printing on it.
Solder cannot flow across dried ink and so to prevent unwanted solder shorts, the solder side of the board is often coated with ink, with the exception of the solder pads. Usually this is a green, or more recently a black ink coating. So starting at the top we may have to print the following layers ....
a. Component side artwork.
c. Tracks and pads.
d. Solder resist.
e. Solder side artwork.
In the above case it would require four screens, with one of those layers on each screen. That means we have to be able to remove a screen change it for the next one and it has to be positioned exactly. This is what is known as 'registration'. In the case of my printing frame the removable hinge pin achieves this. Also on the base board we have an L shaped guide so that every PCB blank aligns itself exactly. So the procedure is lift frame, place the PCB blank in the "L", lower the frame and then pull the squeegee towards you at the right angle (about 5 - 10 degree's). This forces ink through the holes in the emulsion and prints the image onto the board. The screen is then raised, the printed board removed, the spare ink paste pushed back to the top of the screen, and then the whole thing repeated for the next board. About every ten printing passes spoon a little more ink on the screen.
When one layer has been printed and the ink dry, the frame can be changed and the next layer over printed ... and so on.
Looking for trouble.
In an ideal world everything should work first time, but with any process that has a lot of variables, things can go wrong. I would like you to consider two statements, one from a professional who screen prints fabrics and the other a professional screen printer who prints artwork on boxes and signs.
The Fabric printer said " Lines narrower than 0.7 mm do not print well".
The sign printer said, "If I can screen print bar codes, on solid objects, then printing tracks and pads should be no problem"
My main cause for concern is the possibility of ink spread causing two closely spaced tracks or pads to 'short out'. If we examine the PCB layout above we can see that the two 30 thou vertical tracks in the center of the PCB, that pass between the two pads, have the narrowest spacing and so that is where we can expect problems to occur first. If we consider the two statements, one person is printing on fabric, which does not have a smooth flat surface and also one can expect ink to be absorbed, with a 'wicking' action into the cloth. On the other hand the sign printer is printing on a smooth, hard, flat surface that does not absorb the ink and therefore there can be no wicking action, spreading the ink. Either way we now know where to look if things do not go right.
The next question is could I have designed the board differently to have avoided the narrow gap between the two tracks. The answer is probably yes, but the design was originally for a double sided board which I had to change to a single side board design at the last minute. Now either there is sufficient space between the two tracks ... or there is not. If there isn't then there are options, for example we could make the top two pads narrower giving room for a wider gap.
Let us have a quick look at a real bar code label ......
Now compare the track widths and spacing of the PCB, with the screen printed barcode. I think that now you will understand why the sign printers remarks gave me such hopes, that PCB tracks and pads could be screen printed. Also since the thin bar code lines are only one pixel wide on the screen, you may also understand why I asked my wife to buy me a x300 USB microscope for Christmas !.
4th November 2009.
Made up a screen 'coater' using PVC tube and solvent.
The emulsion is poured into the channel and then the coater presented to the bottom of the screen to be coated. The coater is then tilted forward until the emulsion touched the mesh. Applying pressure to keep the coater against the mesh, the coater is then drawn upwards in one clean movement. At the top of the screen the coater is tilted back allowing any surplus emulsion to flow back into the coater. There are many school of thought but , the screen must at be coated on both sides, at least once to ensure that the holes in the mesh are completely filled with emulsion.
Copyright John Kent 2009