DIY PCB creation

[kkissinger]

Down the rabbit hole

Hello. I started this thread due to the "scope creep" of my original project to integrate Arduino-based circuitry into my modular.

Well, there are a number of interesting chips available however they are all tiny SMD devices and I decided I wanted to try something different from the toner transfer method that I've used up 'till now.

I stumbled across a couple of people who are creating PCBs with Resin 3d printers.

What does a 3d printer have to do with PCB creation?

In case you are unfamiliar with resin 3d printers they have a horizontal LCD base onto which a tray of liquid resin is attached (the bottom of the tray is transparent). The resin becomes solid when exposed to the UV light from the LCD screen. However, 3d printing is a whole other topic. The thing to bear in mind is that the base of the printer is a flat LCD screen that has a UV backlight. And, we can "trick" the printer into displaying the PCB traces on this UV LCD screen. Thus, one can simply place a copper clad pcb with its dry etch resist film directly onto the LCD screen (in place of the resin tray). This process reminds me a bit of contact printing from the darkroom photography days.

What is a slicer?

A slicer is software (often provided with a 3d printer) that converts a 3d file into the slices that the printer will print to create the 3d object. There are numerous 3d filetypes but two ubiquitous ones are .step and .stl.

KiCad (pcb design software) will export in .step file format -- and note that the bottom "layer" of the file are the pcb traces. The slicer software can, in turn, import the Step file. If I were to 3d print the entire step file, I'd basically get a plastic block that resembles the pc traces, board, and components. We are only interested in the PCB traces -- which happen to be the first layer.

Thus, one can export a PCB from a package (I use KiCad) into the printer's "slicer" software (in my case, the AnyCubic Photon Workshop that comes with the printer). (My printer is an Anycubic Photon Mono M5s.)

I was a bit frustrated with the lack of troubleshooting information on the subject with many people noting that there is a good deal of trial and error.

What I discovered from Trial and Error

My first few attempts had some problems and I am going to share with you the problems and the solutions.

Problem 1) -- the etch resist would separate from the copper while washing the board in the Sodium Carbonate solution.


1) if the copper is not REALLY, REALLY clean, the etch-resist won't adhere very well.
2) It is important to get a good lamination between the film and the PCB. I found that I had to do multiple runs with my laminator to assure a good bond.
3) the Sodium Carbonate solution should be relatively weak (I settled on 2 grams of Sodium Carbonate for .5 liter of water
4) the Sodium Carbonate solution must be cool (room temperature or less) so as not to soften/melt the etch resist traces



Problem 2) -- the traces were wide and adjoining traces would touch each other. Also, I couldn't clear the etch-resist from the drill holes in the pads. The traces appeared overly-thick and the edges were ragged.


1) this was caused by over-exposing the board on the printer. After some experimentation I have found that 20 seconds works for my particular printer.

Needed chemicals
The only chemicals needed are Sodium Carbonate, Etchant, and Acetone.
I still use Ferric Chloride etchant though there are others that are supposed to be safer and more environmentally friendly.

I purchased the Sodium Carbonate from Amazon though it is probably available in grocery stores in the Laundry section (Sodium Carbonate is often used for laundry). Having said that, use rubber gloves with the stuff and, whatever you do, don't ingest it or get it in your eyes!).

Acetone is in the nail-polish section of your local drugstore -- just be sure to get pure Acetone rather than nail polish remover with other stuff in it.

One also needs to purchase dry etch-resist film -- this is also available on Amazon.

The process for a single-sided board with traces on the top

1) Design the PCB with PC design software (I use KiCad)
2) From the PCB editor on KiCad, export the PCB as a STEP file

1. Set Coordinate Option to “Drill/place file origin
2. User defined Origin can be the defaults (Units=mm, X=0, Y=0)
3. Other Options: Uncheck: “Ignore 'Do not populate' components”, Uncheck “Ignore 'Unspecified' components', and Uncheck “Substitute similarly named models'”
4. Check “Export tracks, pads, and vias
5. Uncheck “Export zones”
6. Board outline chaining tolerance = “Standard (0.01 mm)”

3) Import the Step File into Anycubic Photon Workshop and slice it (it requires no scaling, or anything else). Set the option for printing the first layer to 20 seconds via the “Slice parameter” page
4) Remove a protective layer from the photo resist and press it flat onto the copper (cut the film a little larger than the PCB. Note, use scotch tape to grab the protective layer – press the scotch tape firmly to the film so that it will pick up the layer.
5) Place the PCB inside a folded paper that covers the entire board. Preheat the laminator to the 5mil setting. Run the PCB through the laminator four times changing the board's orientation each time.
6) Place the PCB on the LCD surface and center it with the printer's “expose” function. Be careful not to expose the PCB with any stray uv light,
7 ) Expose the first slice of the pcb for 20 seconds (need to stop the printer once the first layer is printed).
8 ) Remove the PCB from the LCD base. Note that the traces (etc.) will appear to be faint.
9) Dissolve 2 grams of Calcium Carbonate into .5 liters of cool tap water
10) Peel the remaining layer of protective film from the PCB
11) Insert the PCB into Sodium Carbonate solution and agitate gently (wear gloves for this)
12) Gently rub the unexposed areas away – the exposed areas remain.
13) Rinse the board with water and pat dry.
14) Etch the board in the usual manner.
15) After washing/drying the PCB, apply a thin coat of acetone to the board. In a minute or two, the etch-resist will separate from the copper and can be easily wiped off.

The process for a single-sided board with traces on the bottom

1) Design the PCB with PC design software (I use KiCad)
2) From the PCB editor on KiCad, export the PCB as a STEP file

1. Set Coordinate Option to “Drill/place file origin
2. User defined Origin can be the defaults (Units=mm, X=0, Y=0)
3. Other Options: Uncheck: “Ignore 'Do not populate' components”, Uncheck “Ignore 'Unspecified' components', and Uncheck “Substitute similarly named models'”
4. Check “Export tracks, pads, and vias
5. Uncheck “Export zones”
6. Board outline chaining tolerance = “Standard (0.01 mm)”

3) Import the Step File into Anycubic Photon Workshop and slice it (it requires no scaling, or anything else). Set the option for printing the first layer to 20 seconds via the “Slice parameter” page
4) Cut a piece of photo-resist film slightly larger than the PCB
5) Place the photo-resist film directly onto the LCD surface and place a PCB on top of it so that the film rests flat against the LCD screen. Be careful not to expose the PCB with any stray uv light,
6) Expose the photo-resist film for 20 seconds (need to stop the printer once the first layer is printed).
7) Remove the photo-resist film from the printer. (Note the traces will appear to be faint).
8 ) Remove the protective layer from the lower side of the photo resist and carefully center and press the photo-resist film to the PCB centering the traces as needed. Note, use scotch tape to grab the protective layer – press the scotch tape firmly to the film so that it will pick up the layer.
9) Place the PCB inside a folded paper that covers the entire board. Preheat the laminator to the 5mil setting. Run the PCB through the laminator four times changing the board's orientation each time.
10) Dissolve 2 grams of Calcium Carbonate into .5 liters of cool tap water
11) Peel the remaining layer of protective film from the PCB
12) Insert the PCB into Sodium Carbonate solution and agitate gently (wear gloves for this)
13) Gently rub the unexposed areas away – the exposed areas remain.
14) Rinse the board with water and pat dry.
15) Etch the board in the usual manner.
15) After washing/drying the PCB, apply a thin coat of acetone to the board. In a minute or two, the etch-resist will separate from the copper and can be easily wiped off.

I should probably do a video on this and I will put it on my "do list".

Why 3d printing?

I have kind of wanted a 3d printer and there are a number of things I want to be able to make with it in support of my synth-DIY. Indeed, I want to make mounting brackets, frames for OLED displays, and other niceties. So, I anticipate getting good use out of the printer. btw -- I am learning to use "Blender" which is a free, open-source 3D modeling and animation program. Well, that's something for another time and another thread!

[PHOBoS]

[blue hell]

What pho said

And some more .. mainly nostalgia .. and a bit of curiosity of course.

I like that 3d printer idea - indeed for what more it can do than make circuit boards. Oh and also that resin printers seem to be able to do more precise constructions than the wire based printers (which always seemed to have a bit rough results - from what I saw).

I used to make PCBs with boards pre-sprayed with a photo sensitive layer. Layout I'd make with tape and rub-off symbols on transparent paper - at a later stage 2:1 and would then go to a printhouse to have made a 1:1 en-small-ment - Jurgen - the printer - would do it for a really nice price (also for the xmas cards I then did ) (which may have been because I had done a screen printing course given by him - which was great fun btw).

Anyway - later on when doing more processor based things I needed better PCBs than i could make myself - so had them made by some company(with through hole and smd and all).


Pwew - one mosquito killed


Erm .. where was I .. ah yes . .. what is the reason for you to use a light sensitive film instead of pre-sprayed PCBs? (except that the internet seems to be full of that (for me new) technique ...)

I have never used the glossy paper tricks you mentioned and that I've seen to be used here - seems messy ...

[kkissinger]

Thanks for your replies.

For many years, I used the toner-transfer method and got excellent results but in recent years, I have not.

Upon some introspection, it occurred to me that a few years ago I changed printers. I think that my former printer was an HP (I don't remember the model) and that my current printer is a Samsung. I suspect the Samsung does not apply the toner at as high a temperature as did the HP and, as such, the black areas of the printout (i.e., the traces, pads, etc) were not adhering to the toner transfer paper very well.

I have done a bit of research and appears the Brother, Canon, and HP printers are best for toner transfer. I may go ahead and replace my laser printer with a model that is better-suited to toner-transfer printing.

One printer that looks promising is the Brother HL-L2320D -- it is a relatively inexpensive printer but has some decent reviews for its use with toner transfer paper.

Do you have any recommendations/opinions on this?

[kkissinger]

Hello again!

I couple of posts back (you can scroll up to see it) I posted my procedure for creating PCBs with the MSLA printer screen. I discovered that while the resulting traces looked good, they were not "mirrored" and, as such, could be used on the top side of the PCB. However, for a single-sided through-hole PCB, we want to the traces to be on the bottom.

The solution is to print the etch-resist film before laminating it to the PCB.

As such, I am printing the film, removing the protective layer from the bottom side of the film, then laminating it to the copper.

I updated the previous post and included both methods (one for the top and one for the bottom.

I will keep you updated!

My test of this method looks good and I will keep you updated.