You are here: HomeThe SciEngDIY PCB DevelopmentDIY PCB Fabrication, Construction and Assembly

DIY PCB Fabrication, Construction and Assembly


0000494In this article  the fabrication, construction and assembly of a single sided, Plated Through Hole (PTH) component, DIY Printed Circuit Board (PCB) is demonstrated. The discussion presented shows how a PCB layout is fused onto a  copper clad PCB using a laminator and toner transfer paper. It then goes onto to show how Sodium Persulphate can be used to etch the design.




The PCB design used in this article has been designed in Altium Designer, although any PCB designer software should suffice. In fact Eagle CAD is a particularly popular choice within the hobbyist electronics community and is worthy of an honourable mention! The layout of the PCB, seen in Figure 1 below, has been designed in such a way that the components are on the top or component side and all of the signals traces are on the bottom. Where necessary wires will be used to connect the power rails together as the top layer, in conjunction with vias, are not being used.

As we didn't have any single-sided copper sheets, this design will be transferred on to a double sided copper sheet, which approximately measures 100 mm x 78 mm.


PCB layout design

Figure 1: PCB Layout. The bottom layer signal traces are in blue, while the components and interconnects are coloured in green/yellow. The constructiuon layers are in the two shades of purple on the mechanical layers 1 and 2.

 In order for Altium Designer to printout the desired outputs it is necessary to set the required default prints, by navigating to the menu item File->Default Prints. This pop-ups the Default Prints dialogue window. When the "PCB Prints" option is chosen as the default prints, the output produced will show all of the objects on the enabled PCB layers.

This type of printout is useful for testing component placement and it helps ensure that all of the PCB footprints actually fit the components. Note, if the PCB printout covers the full page of an A4 sheet, as ours embarrassingly did, then the print scale value, in the page setup dialogue, should be set to 1.0. An example of using this printout to test component placement can be seen, in the Figure 2, below.


Figure 2: The component and connectors have been test fitted on a printout of the PCB layout. Despite this test, we still got it wrong! However, that it is the beauty of DIY prototypes, you can afford to get it wrong as many times as you like, as the method avoids expensive re-spin costs.

To printout the actual design, as it will appear on the Heat Toner Transfer Paper or Press-n-Peel Sheet, it is necessary to navigate back to the default print dialog. This time, in this dialog, the Final Artwork option is selected, as can be seen, in Figure 3, below. For this particular design the multi layer, bottom layer and mechanical layer 1 (which provides us with a border) have been selected as the layer print options. Note that we have selected to have via and component holes by selecting the holes option. This helps in locating the center of the hole when drilling, although sometimes, if the ring around the hole is not thick enough,  it can be completely etched away. As this is not the printout for the top layer we have not mirrored the design.


 Figure 3: In Altium Designer the user can select the layers that will appear in the final PCB printout.

In the Page Setup dialog, found in the File menu,  the colour of the printout should be set to mono, as this should help towards obtaining the highest resolution printout possible. In this case the printer's resolution was also set to its maximum, which we think on our HP Laserjet P1005 is an interpolated 1200 dpi. The Figure 4, seen below, shows the result of our printout onto plain paper. Note that we have thickened our component holes and vias where possible, which provides for a more resistive area to copper etching. If at this stage happiness ensures with the printout onto plain paper, then it should be ready to print onto Heat Toner Transfer paper.



Figure 4: Where possible the component and connector holes have been thickened to provide a larger etch resistant area.

As mentioned previously, initially the design has been printed on to a plain A4 sheet of paper to help in registering the design and testing component placement. It also allows us to estimate the amount of toner transfer paper to use. It is not unusual to cut out a small area of the heat toner transfer paper, similar in area to the size of the design, to cut down on wastage of the sometimes expensive toner transfer sheets. If the printout looks satisfactory then the design can be printed on the toner transfer paper. In this case this was done by sellotaping the toner transfer on top of  the design on the A4 sheet and feeding the sheet back into the printer. For the HP Laserjet, P1005, this is done by placing the shiny side of the toner transfer sheet facing upwards.


 Figure 5: The design printed on the heat toner transfer paper with (left) and without (right) a ground plane. From our experince dedicated heat toner transfer sheets and the more expensive press-n-peel sheets work equally as well. Others, on the Internet, have used the pages from  glossy magazines with excellent results too.

Before using this design we decided to add a ground plane, this helps save on etchant and reduce the total etching time. Once the design has been printed onto the heat toner transfer paper the next stage is to transfer it onto the  copper clad PCB. Firstly though, the PCB can be conditioned by scrubbing it with a PCB cleaning scrub block to help degrease and polish it. The board could  then be washed with soapy water and cleaned with acetone.

After cleaning the PCB, the design is  sellotaped and then fused onto it. This is done by repeatedly feeding it through a laminator. The one used for this purpose was a Texet A4 Laminator. We found this laminator easy to convert, as many Internet sites suggest, to increase its operating temperature. The PCB was feed through the laminator 10 times, as the number 10 seemed like a good number and by this time the PCB was incredibly hot to the touch. It was then placed in cool water, until the toner transfer paper was easy to peel-off.


Figure 6: The design has been fused on to the PCB. The blow up image shows that the ground plane could have been better. However, after printing out the design we realised that the printer was no longer set to its highest resolution, as it reverts to its default resolution after each print.

The result of fusing the design onto the copper clad PCB can be seen in Figure 6 above. Fusing the PCB layout using a laminator has allowed  us to obtain a very clean result. Even when viewed under a microscope. as can be seen in the top left-hand corner of the figure, there does not appear to be any broken tracks, very nice! The next stage, of the DIY PCB process, involved etching the PCB using sodium persulphate, the so-called clear stuff.


Figure 7: The PCB Etching Temperature Profile.

Sodium Persulphate, or Na2S2O8, is sometimes preferred as an etchant to Ferric Chloride, FeCl3 principally because it does not stain everything that comes into contact with it. Therefore making it a cleaner solution to work with. However, there are advantages of using Ferric Chloride, including being more effective at lower temperatures. The graph, in Figure 7, above shows a plot of the temperature profile of the etching process over time.

The PCB has been inserted into the etchant at a temperature of 46.4o Celsius.  The final temperature was measured to be 62.5oC when the PCB was removed, fully etched, 12 minutes later. Although the recommended operational  temperature of Sodium Persulphate is between 40oC and 50oC we did not notice any adverse effects of operating the etchant at this, higher than recommended, temperature. This slight anomaly was due to our new etch tank design, to be reviewed in a future article, not yet being calibrated. The result of etching the design can be seen, in Figure 8, below.



Figure 8: The etched PCB. The overall result was very pleasing, with a nice and clean etch. The width of the signals traces are 1mm for ground and power rails and 0.5mm for everything else.

The result of etching the design was extremely good, with no broken signal tracks.  However, we quickly realised that it may have been better to use a pair of rubber tweezers to retrieve the design from the etchant, rather than scissors, as the latter scratched the PCB. This can be seen, in the top-left hand corner, in Figure 9 below. Next, after washing the board in water, the etch resist is typically removed using acetone, although in this case nail vanish remover sufficed conveniently well.



Figure 9: The etched PCB cleaned with acetone ( or in this case nail varnish remover). The damage to the tracks shown in the photographs was caused when removing the PCB from the etch tank using scissors! Naughty!

Assembling the board was fairly straightforward with each component hole being drilled twice. Firstly, a pilot hole was drilled using an extremely small drill bit, typically between 0.1mm and 0.3mm. The second step involved using a larger drill bit of between 0.5mm and 0.7mm depending on the size of the component hole required.



Figure 10: The bottom of the assembled PCB. The red wires help connect the 3.3V power rails to the components when the top layer has not been used as a signal layer.


Figure 11: The PCB's top or component layer. It consists of a 320 x 240 LCD display, SPI Flash memory a USB 2.0 to TTL serial converter and a 2 x 20 way IDC connector, which is compatible with the DE0 Nano.



This was a relatively simple design, which was etched and assembled in only a few hours. The result, as can be seen in Figures 10 and 11 above, is extremely pleasing to the eye. It demonstrates the power of the DIY PCB when a rapid prototype of a design is required at minimal cost. To take this design to the so-called "next level" the next step is to try one of the many methods described on the Internet used to add a DIY silk screen layer.

Giving the ease of achieving success on this project our next design, using the same process, will  bravely involve designing a DIY PCB containing a 208-pin FPGA and synchronous DRAM. Stay tuned.

Go to comments start