Posted by, Penguin on February 22, 2010 at 6:05 pm
The following will be a sequence of steps detailing the PCB creation process that I subscribe to. If you have any questions or comments feel free to email me directly or post a question in the comments. I will also include images which correspond to each step in this process, I will indicate that an image corresponds to a certain step by adding a reference id Example: (1) where the number in the parentheses corresponds to an image index in the gallery below.
I’m assuming in this process that you are familiar with creating dextrin glue for toner transfer paper, If not, no worries simply read this guide on creating dextrin coated toner transfer paper By Triston J. Taylor, the original process is attributed to a make.com user by the alias Cashsale.
http://www.radagast.org/~dplatt/hamradio/How_to_make_laser_toner_transfer_paper.pdf
Without the dextrin coating its nigh impossible to create such high resolution boards with any fidelity so I highly recommend investing in some dextrin or creating it in the process described in the above link.
as a side note my dextrin comes from my pyrotechnics / chemistry stash because dextrin is also used as a water soluble adhesive for creating fireworks stars / shells as well as compact smoke bombs and other items.
Places to buy dextrin:
United Nuclear
Ebay
Cheap Chems
Step 1: Initialize
Begin by printing out a to-scale representation of the front and back of your boards on separate pieces of paper (1), make sure that along with each side’s copper mask you also print any Vias and or pads which need to be present. In eagle layers that would be [Top Copper, Vias, and Pads] for the “front” and [Bottom Copper, Vias, and Pads] for the “back”.
If you look at the final board layout image you’ll see that I have also strategically placed alignment marks at the 4 edges of the board, these are marks that I can line up later to ensure that the front and back layers are perfectly aligned, which is important because if they are skewed then your via’s and pads wont line up and the board will become useless.
Step 2: Prepare Substrate
Next, cut out an amount of silicone sealed laser photo paper (2) which completely covers your design(3) (refer to this pdf if you don’t know what that means), you will need one piece for each layer you are printing, make sure to leave about 1/8 to 1/4 inch extra border which will make it easier for you to insert the copper clad later and tape the edges to align them.
Step 3 Apply Dextrin:
Next we are going to coat one side of these two pieces of photo paper with dextrin adhesive(4), again if you don’t have any please refer to the how to create dextrin PDF in the previous posts. Begin by using an eye dropper to place 3-4 drops of dextrin glue on the paper (use more or less depending on the surface area you are working with, in my case 3 or 4 was the most I needed) at first the dextrin glue will tend to bead up(5)(6), that’s OK, just keep spreading using a roller or a small paint brush until it begins to get tacky at which point it will begin to adhere evenly. Obviously repeat this step again for the second layer if you are producing one.
I have had some people ask me what a coated piece of paper is supposed to look like, so for reference i have included a photo(7) where the left piece is coated with dextrin (very light yellow coating) and the right piece of paper is not coated. You’ll also notice the coated piece is slightly curled up which is OK, if you want it to flatten out then just heat it for about 30 seconds under a warm light bulb to dry it otherwise rest assured your laser printer will dry it out plenty quick.
Step 4: Print
Next, using a small amount of masking tape go ahead and tape your dextrin coated paper over the original design template that you printed out(8).
And print!(9) Remember to mirror one of your sides otherwise you’ll find its curiously hard to line up your layers in the next step…
Step 5: Align Layers
Now remove the masking tape from the dextrin paper and arrange the two layers such that they are facing one another and line them up, this step is crucial, especially if you have a complex board, once the two layers are lined up then tape them together using more masking tape(10)(11), leaving an opening in at least one side to slide the copper in after we prepare it.
Step 6: Prepare Copper Clad
Next cut out a piece of copper clad to match your board size, in my case that’s 3.5 by 2 inches, use an Xacto to score the board and then snap it or even cut it out with a good pair of scissors, either way just make sure that after you have your piece cut out that you clean it. Copper clad oxidizes(12) while it sits in storage so you need to scrub that layer off(13) with some steel wool or a pot cleaner, after you have taken the layer of oxidation off it doesn’t hurt to wipe the entire board down with some acetone or alcohol(14) to remove any dust or oils that might be present.
Notice the difference in the copper before(12) and after(13) cleaning in the images.
Step 7: Prepare for Transfer
Next you’re going to slide your freshly cleaned and dry copper clad in between your two dextrin coated copper mask layers making sure to line the board up with the edges of your mask, its hard to show this step with an image(15) but it should be fairly self explanatory what you’re doing here.
Step 8: Transfer
Next you need to apply heat and pressure to the entire sandwich in order to transfer the toner to the copper clad from the paper (hence the name toner transfer method) I use a small lamination machine(16)(17) because it applies even and consistent pressure and heat which tends to produce excellent results, I have also used a clothes iron with mixed results, if you use a lamination machine 8-10 passes through should be plenty, if you use an iron the rules vary from iron to iron so you’ll have to play with it and see what works. Generally speaking 5 minutes of solid ironing should do the trick, and make sure you put something between the transfer paper and the iron like a layer of wax paper otherwise you will burn your transfer paper.
Step 9: Dissolution
After you transfer the toner to the copper clad you need to remove the paper that the toner is also fixed to, this is where the dextrin comes into play, by soaking the entire sandwich in warm water(19) you expose the dextrin to water making it dissolve and separate the toner from the paper effortlessly.You will see wrinkles start to form in the paper(20) as the water penetrates into the dextrin layer and dissolves it.
Step 10: Separation
After letting the board and transfer paper soak for about 5 minutes its time to peal the paper off the copper clad, Be gentle and peal back slowly, if a little paper stays behind its OK but ideally the separation should be effortless due to the dextrin(21)(22)(23). If there is paper remaining simply remove it by rubbing it off, you just want to make sure its out of all the spaces between the toner because if it isn’t it will act as an etch-resist and it will cause bridging in your board.
Step 11: Corrections
After you’re satisfied that there isn’t any paper left where it shouldn’t be look over your entire board and make sure there aren’t any breaks in your traces(25)(26), if there are use a fine point sharpie and draw them back in(27)(28), the sharpie ink will act as an etch resist and prevent the breaks from actually etching out.In the image you can see some red tinted areas where I have used a sharpie to correct some gaps in toner.
Step 12: Etching
Finally we are ready to etch the board, I typically use ferric chloride(29) which takes approximately 15 minutes to complete an etch at room temperature, after the board is done etching the copper will have been removed where there was no toner or other etch-resist(30)(31), but the toner will remain, use the same steel wool or pot cleaner that you used to originally clean your copper clad to remove the toner from the now fully etched board.
Step 13: Admire Your Work
Posted by, Penguin on February 22, 2010 at 4:51 pm
I have finalized a Board layout and schematic, I ended up going with a double sided board since I was unable to route the entire board on a single plane.
The advantage of a double sided PCB for those who aren’t aware is that there is copper clad on both sides of the board as opposed to a single sided board which has copper only on one side.
This allows us to run a signal from one side of the board to the other using a “via”, a small conductor which gives us one extra dimension of flexibility in our design, one extra layer in 3d space to route our signals.
Consider this puzzle as a simpler example of the same problem (good luck).
The file below contains the most recent Eagle schematic and board layout as well as several helpful part libraries and the design rules I used.
AetherTapEagle
Posted by, Penguin on February 22, 2010 at 4:16 pm
The second step in any project is to identify the main components that are needed to make your goals play out. In this case I have chosen a Microchip PIC24F series Microcontroller as the core of my project supplemented with a Microchip ENC28J60 Ethernet controller to handle the MAC and Physical layers of the Ethernet interface .
The PIC24F series has many advantages over its 10, 12, 16, and 18 series cousins, First and foremost the 16 bit architecture allows the chip to have a far expanded instruction set which makes it much more viable to program the chip in a higher level language such as C or even BASIC. I myself will be using Microchip’s student version of its C30 Compiler which comes fully loaded with individual header files for each chip in the 24 series and even libraries to help out with the on chip hardware peripherals like USART, I2C, SPI, Etc. The 24 series also has one advantage that most people tend to overlook, its hardware peripheral pins can be remapped to almost any pin on the device, meaning much less time working on routing a PCB. Its also interesting to notice the distinction between the 24F and 24H series controllers, while the 24H controllers can typically run at much higher speeds achieving in the range of 64mips, it typically lacks the expanded peripherals of the 24F series which on average has about double the hardware peripherals of the 24H series but performs at only around 16Mips
The exact model I’ll be using in this case is the PIC24Fj64GA002 with 64K of flash memory, 2 hardware SPI interfaces, and a max speed of 32Mhz (16Mips).
I wont be writing a guide for using the pic24 however I was able to find an excellent tutorial for any interested.
The ENC28J60 Ethernet controller handles everything below the network layer of the standard OSI Model, it handles the media access control layer and the physical reception and transmission of packets. The ENC28J60 interfaces to a main controller via SPI and receives and transmits information via an internal 8k buffer operating in a FIFO fashion. the ENC28J60 supports full and half duplex modes as well as unicast, multicast, and broadcast packets. It is also fully compatible with all 10/100/1000 Base-T networks making it a very plug and play device. The only qualm I have about it is that it it doesn’t support auto-negotiation which means its not safe to use full-duplex mode unless you can manually change your switch port to full-duplex.
During the course of this project I will be entering a lot of new territory, I have never before used a 16 bit uC, I have never before coded a Microcontroller in C, (until now I have only used the 8bit 12 and 16 series microchips and coded them in Assembly or PIC Basic), I have also never used the SPI hardware peripheral which many PIC’s provide and I have absolutely zero experience with the ENC28J60 module. So this project will without a doubt put me in a position to learn some new things.
Below I have included a schematic and board layout for a pic24 28Pin SOIC breakout board with an ICSP header for programming, A useful tool for familiarizing yourself with the PIC24 Series
PIC24BrkOutwICSP
Posted by, Penguin on February 22, 2010 at 3:48 pm
When you begin a project its important to identify what you hope to accomplish or learn in the process, This project will have plenty of both for me. First and foremost I would like to produce an inexpensive Ethernet enabled controller which I can interact with via the Internet. The device will be a product of its programming however, meaning that depending on what I need it to do it should also be able to…
1) Host web content and display real time data via the web
2) Send emails or alerts when necessary
3) Monitor and interact with external devices, sensors, and appliances
4) Allow control of the above mentioned devices via email or web based interface.
5) Log Network activity to the SD Card.
Its important to note that I don’t plan to incorporate relays or other direct appliance control hardware directly into my design, I’ll probably just end up making a bank of I/O pins available for external connections from sensors or relays.
As a rule of thumb its always a good idea to do some research and find out if anyone has done some of your work for you, in this case it happens I was able to find an excellent project to draw from, several years ago Ian from HackaDay posted an excellent How-To on building an embedded web-server which has a lot of the same functionality as my project. Its notable to mention that Ian is also the main developer of the very popular Bus Pirate module. There are also several other good sources of information about the ENC28J60 such as this article on Electronicfr.com or either one of these articles from tuxgraphics.org
So in the footsteps of giants I will lay out my project in the next several posts from Component Identification through Design, Prototyping, Building, Programming, and Testing.
As a final note, because I’m borrowing from other open source work and projects it seems only fair that I reciprocate and deliver my project documents and code as open source, and so it shall be.
Posted by, PHLAK on February 7, 2010 at 12:46 pm
This meeting was a particularly large meeting. Most “active” members were there at some point throughout the night and several of us overflowed outside. AltF4 gave his advanced presentation on cryptography, Zapperlink brought an Intel I7 powered laptop and an Atom powered IP phone, Rax gave away a couple of shirts (I got the Android one, woohoo!) and other miscellaneous conversations went on as well.
Posted by, Penguin on January 27, 2010 at 10:30 am
So I guess I have access to publish posts on this front page, that’s pretty awesome. I’m sure I wont ever abuse this.
Anyway just wanted to say I have setup a web-log to document my embedded Ethernet project, I’m already half way done with the project but I just started the web log so I’m sort of playing catch up with my posts.
I’m going to do my best to make one post per day until I’m caught up and then ill post as needed to document the project.
http://fgsfds.com
I may post other things on it as well but mostly the site is just for documentation purposes.
Also its worth mentioning for those who have been waiting on me to publish instructions on how to do the toner transfer etching that I will finally be documenting that as part of this project.
-Charles “Penguin” Bock
Posted by, PHLAK on December 15, 2009 at 5:40 pm
Rax gave a presentation on Tomato custom firmware and a talk was also given on DD-WRT custom firmware. Both for the Linksys WRT54G series of home routers. PHLAK, baby Breach and Huertanix went over code for the Haystack Project. Also, Dual made an appearance.
Posted by, PHLAK on November 10, 2009 at 11:41 am
The focus of this meeting was definitely AltF4′s talk on cryptography and basic WEP encryption.
Presentation Slides
WEP Slides – http://www.public.asu.edu/~dpetro/WEP.odp
Other slides – http://www.public.asu.edu/~dpetro/Crypto.odp
Posted by, PHLAK on September 20, 2009 at 4:48 pm

PHX2600 Flyer Preview
Get your official PHX2600 meeting flyers and help spread the word today!
Help out the group by placing some flyers wherever people will see them. If you frequent a coffee house, place some there. If your work is rather technical, set some out in the break room. If you attend a technical school, post some on the cork boards they have all over the place. Just make sure you get approval before posting these anywhere.
Download (PDF):
PHX2600 Flyer Full Color
PHX2600 Flyer Full B&W
PHX2600 Flyer 2-Up B&W
Posted by, PHLAK on September 5, 2009 at 11:26 am
September turned out to be quite a large meeting. Ugly was working on attaching a new power connector to a motherboard and helped Rax diagnose a problem with his monitors. The main focus of this meeting was on Penguins presentation on creating your own PCB boards via acid etching. Also, some ladies curbed their truck after attempting to get several members to sign a petition for legalization of marijuana (yeah, they were probably high). We got some lulz.
Click “Read more” for PCB acid etching videos.
Continue reading “September 2009 Meeting History” »