Adventures in Photoetching - or - Making Old School PCBs for Fun

At the first of the year I published an article on two new versions of the EHX Big Muff Pi effects pedal I'd completed. One was built on a home-etched board a friend had sent me using a pattern from the Effects Layouts blog. At the time I mentioned that I had done photo-etching in the past starting when I was in graduate school (getting a Ph.D. in chemistry - so applying chemistry to something guitar pedals, electronics, and music is "really cool" in my book!). Since I had some experience in the area (most recently around four years ago when I etched not one, but two sets of double-sided detail parts for 1/100th scale T-72 tanks), I decided it was time to dust off the old equipment and get cracking. This time I decided I'd try and take advantage of a few options unique to PCB etching (since I'm only doing single-sided boards!). 

Board etching in Ferric Chloride

I wanted to start with a board I wasn't able to find a good clone of - so sort of going back to my BMP roots, I chose the Cornish P2. It's another BMP variant favored by David Gilmour, among others, so it's a circuit I'm at least passingly familiar with. There doesn't seem to be a good clone on the market at this point (at least not from my usual haunts), and original pedals seem to be trending north of $1000 on Reverb. Effects Layouts has a project file however that is, as he says, "probably" the Cornish P2. The qualifier implies to me that there hasn't been a definitive trace done, which may be why you can't find a clone from any of the larger PCB makers.

In the past I'd used a negative tone resist for my photo etching. With a negative tone resist, the resist cross-links and polymerizes where light hits. Therefore when you develop the resist, the areas not exposed to UV light (the dark parts of your mask) are removed. Therefore when you go into an etching solutions, the dark parts of the mask is etched away. Most PCB patterns are printed out with the traces (in other words, what you want to remain), in black as many people use direct toner transfers for their etching - in other words effectively a positive tone resist. As the PDF versions of the PCBs were already set up for positive tone, I decided I'd give some of the pre-sensitized positive tone boards a try. 

Initial exposure set-up - I'd later raise the board to bring it closer to the light

I picked up several from Jameco, and they're pretty nice boards overall. The ones I bought were 4" x 6" and they come with a layer of positive photoresist already adhered to the copper side of the PCB. My first attempt at a board quickly overdeveloped - I think I let too much light get to it and despite using 1:10 parts developer to water, I think the developer was much too concentrated. The second go around, however, appears to have worked very well. I was able to get four P2 boards and two of the Cornish bypass boards out of the single 4x6 PCB sheet.

Four P2 boards and two Cornish Bypass boards

Unfortunately, my next mask set - a mix of Green Russian BMP, BMP project squares, and BMP tone stack mods, was an unmitigated disaster (see below). I can think of two potential root causes. Either the boards overdeveloped again, or the light intensity wasn't strong enough at the edges to completely expose the resist in the first place - meaning it had to develop longer, causing the issue. Either way I'm not sure any of the boards are actually usable as there is at least ground trace loss on all of them, and some are missing huge sections of the circuit itself.

Moderate to heavy trace loss on all edge boards

So given my issues with the positive boards, I decided to go "old school" and go back to my negative photoresist like I'd used four years ago. In fact it actually was the negative photoresist I'd used four years ago. I'd kept what parts of the kit I hadn't used safely stored until needed again. I decided to also go with a smaller PCB to ensure that lack of light intensity wasn't an issue during the exposure process. The process is a bit more involved. I started with a 70mm x 100mm copper PCB. Unfortunately they'd gotten scratched up a bit in transit, so I had to polish them with 600 grit wet sandpaper. This left a few slightly visible surface scratches, but they were so shallow I didn't figure they'd be an issue. I've ordered some of the 3000+ grit polishing sanders, but they haven't arrived yet.

After the copper is polished, the next step is to actually adhere the dry resist to the copper itself. The dry resist comes with protective film front and back. One film is removed and the resist is placed on the board while it is wet. Most of the moisture is then removed, and the board and resist is put in a paper carrier and passed through a laminator twice. I'd checked before committing to the resist that the PCB and carrier would actually pass through the laminator as I didn't want it to get stuck part of the way through!

Board with developed photoresist

Once the resist had adhered to the board, it was time to set it up for exposure. I'm using the Micro-mark set, and while they indicated that they did not recommend using laser printers to create masks (and provide special transparency film for ink jets to that end), I used the same transparency film and laser printer for both the positive and negative tone resist. As you can see from the photos the laser-printed masks actually worked even better on the Micro-Mark negative tone than they did on the pre-sensitized positive tone boards. 

To expose the resist, I used essentially the same set up as before - an incandescent bulb roughly 4" away from the top of the pattern. As before it took about 10-20 minutes, but this time I got a really nice purple color for the exposed photoresist which covered all of the traces as they'd appear on the final board. After developing I was really pleased with the outcome, the resist looked like it was solid, so it was time to etch.

Board post etch - the resist clearly held up!

I put the board into the ferric chloride and let it etch away. I'd added dumification to the board (extra lines around the boards) to limit how much material had to actually etch. As always, the etch took a little while (roughly 30 minutes), but the end result looked pretty good pre-resist strip. After the resist strip (below) I was literally blown away with how good the final boards looked. I can't wait to get them drilled and populated!

Final result! Wowzers!

So speaking of cutting, drilling, populating, and heck, even etching these boards. Please, please, PLEASE always follow recommended safety instructions when playing with chemicals and cutting/drilling PCB boards. Developer solutions typically contain sodium hydroxide, and ferric chloride will wreak havoc on metal surfaces, your clothes, your skin, pretty much anything it comes in contact with. When cutting and drilling boards, use appropriate PPE as the fiberglass (or other substrate) dust will do your respiratory system no good at all. Don't skip the instructions - please!

P-2 Board drilled - solder side

So with that being said, above and below you can see one of the P-2 boards drilled and ready for adding all of the components. I have a Dremel multi tool that I've got a small diameter chuck for where I can simply put a small pin vise bit into the chuck and run with it. It is also variable speed, so I can speed it up or slow it down as needed. Some people use a drill press with their Dremel, but I tend to freehand it - which does lead to a few small scratches, but once everything is soldered down it hasn't seemed to have mattered so far.

P-2 board drilled - component side

I'm still building up the board above as I thought I had a few of the components, but ran short! (DOH!) I also have several other builds planned that I've already printed out masks for, and I hope to get the boards etched before the weekend. Keep watching this space as I'm planning on showing off a few of the new board projects soon!