Etching PCB's at home

[juggle]

A couple of people have asked me to make this post so I’m going to do it. But - right up front…I don’t know what I’m doing. I’ve got a system that seems to be about 98% reliable now…but it took me quite a bit of trial and error despite following a number of great on-line tutorials, videos, blogs and other resources.

I can now go from a print ready design to a finished board in under an hour. The cost per board is a bit tricky for me to nail down since I went down a couple of dead ends in my attempts…but one of the best resources I found calculated a price in 2017 and came up with $0.54 per 70mmx100mm board. Prices have gone up a bit since then though. If you really shot around…maybe you can still get close.

Now - right off the bat there are some limitations here. Be realistic at first. There are people using this method to do really fine pitched SMD parts…but I’m mostly making Yusynth’s boards which have nice thick traces and big clerances…or my own designs which I create with 0.5mm traces and 0.4mm clearances. Though now that I’m having more success I’m thinking about trying some less forgiving specs again.

Also - I’m only doing single sided boards right now. When I design them I do put a few traces on the front side - but those are just to represent jumpers. You CAN do double sided boards and I plan on trying one soon. But it’s a lot more to go wrong so start out easy.

And before I really get into it the other big question is…why? You can get boards professionally made dirt cheap now! I know, I’ve used several services and they’re great! But you don’t want to or can’t wait a week or two or pay high amounts for fast processing/shipping. Maybe you don’t want to deal with selling extra boards or dealing with the guilt of just tossing them in the trash since you usually need to order at least 4-5 boards with those services. Maybe you don’t want to risk good money on a batch of boards if you aren’t sure of the circuit yet (one bad board isn’t a big deal…but it’s no fun to order a run of them and then find a problem!)

Maybe you want to make a board you found art for on the web - but there aren’t any gerbers (like the MFOS and Yusynth boards) so you can’t send it off to a board house. Sometimes you can order them from someone but if they’re out of stock or closed or just not available…the only option is to roll your own. (or design your own board and generate your own gerbers.)

Bottom line, being able to make your own boards isn’t that hard or expensive and can be VERY handy! Don’t be intimidated by it - just give it a go!

A lot of people like the toner transfer method. And I tried this first. It can work. But I’m not a fan. It’s hard to get a really good clean transfer consistently. Even with special purpose bought paper I had a hard time getting consistent results. I have tried using toner transfer for “silkscreen” on one of my boards and had some partial success and will probably try that again. But overall toner transfer is just too iffy of a process even with the expensive special paper.

So I’m using a photolithography process. You can buy pre-sensitized boards but they’re kind of expensive. And I found it’s not that hard to sensitize your own.

Before I go into the gory details of how I do this I’m going to share a link that would have saved me a ton of time and effort if I had found it first: http://www.johncon.com/john/PCB/

That is hands down the best resource I’ve found for DIY PCB’s. He has a ton of great detailed information and details on how he came up with his recommendations. Mine are mostly an echo of his. Anytime I had problems with some step of the process if I tried his methods all my problems went away.

So let’s get into this. There are a few things you’re going to need so let’s get that out of the way first:

The most expensive is an inkjet printer. I tried my laser because I really love laser printers and hate inkjets…but…I just couldn’t get dense enough prints out of my laser. If you have a laser than can give really good dense prints on transparency film (or vellum) - use it! If not…save yourself a lot of heartache and just get the cheapest inkjet you can. I got a Canon Pixma MG2522 for $25 at the local mega mart:

Sitting on top of it is one of the next most expensive things you’ll need. Inkjet Transparency film for screen printing. This is key! I got it off Amazon - 50 sheets for $17 delivered to my door: https://www.amazon.com/dp/B07S5GKGSJ/ref=cm_sw_r_tw_dp_y5mRFbEBGTQDR?_x_encoding=UTF8&psc=1

You can also use vellum - but it’s harder to print on with an inkjet and requires more exposure when you make your board. If desired I’ll do a detailed post later on all the alternatives I tried - but this gave the best results overall.

The other “bit ticket” item is a laminator:

I got this one at the mega mart for $24. There’s an amazon basics one that’s a dollar or two more that’s really popular for this, but it’s smaller. This one isn’t great but it works very well for me. Note - this is technically optional. You could just use an iron. I did a few with an iron and it worked…but it was much easier to get consistent results with the laminator. And at <$25 it’s pretty cheap and kind of nice to have on hand - I laminated a nice cheat sheet on how the CV/Gate outputs from my sequencer map in different modes so it won’t get destroyed

You will also want scissors and tape - but I’m guessing you have those laying around. Note - do NOT use your wife/girlfriend’s fancy sewing scissors as I have pictured here. I’m widowed so I can do these things - but I know I’ll regret it next time I need to cut fabric…I can live with that.

While you’re at the mega mart there are a few other things you’ll want:

Washing soda is what you use to develop the UV sensitive film. Acetone is what I use to degrease my boards before applying the UV film…it can also be used to strip the film after etching the board - but more on that later.

And you can also buy your etchant at the mega mart:

Yep. Just ordinary white vinegar and Hydrogen Peroxide make a functional etchant. You do need to spike it with some salt to get it working well…and it’s a bit slow but it works.

Not pictured but a better choice is HCL - Hydrochloric Acid, sometimes sold as Muratic Acid here in the states. You can find it with pool supplies at most hardware stores or pool stores. It’s pretty cheap. I think I paid $7 for 2 gallons a few years ago when I bought mine. It works MUCH faster than Vinegar - but does fume a bit more so I use it outside.

Also at the mega mart you may want to pick up:

A few containers, a shot glass, and some cheap foam brushes. The bigger container you can just grab from your kitchen…nothing nasty will happen to it. The smaller ones…get something disposable. I use these little ones that my lunch meat comes in.

The shot glass is optional…but it’s the perfect solution to mixing up your etchant. The foam brushes won’t last super long. You’ll need at least two but you can usually get a bag of a bunch of them for a few bucks. It’s worth having spares the process is a bit hard on them.

There are two more specialty items you’ll need both of which you can get off Amazon. The first is the UV film itself. You want “UV Sensitive Dry Film” It’s sold by a lot of places under a variety of names…but it’s suspected that it all comes from just one or two sources. This is the stuff I bought: https://www.amazon.com/dp/B07MMVPY1Z/ref=cm_sw_r_tw_dp_ZfnRFb2GJJ45K?_x_encoding=UTF8&psc=1 That’s a 30cm x 5m roll…it should last a long time.

The other is blank PCBs! These are the ones I’ve mostly been using: https://www.amazon.com/gp/product/B01MCVLDDZ/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

They’re only $7 for 10 and are pretty good quality. But…they’re slightly undersized. They measure more like 65mm x 100mm. That’s ok if you’re doing your own boards…but a lot of designs out there are for 70mm x 100mm boards (because that’s the largest size Eagle could do for free for a long time) so if you’re going to try and do some of Yusynths boards his are 75mm x 100mm. For those I’m using these: https://www.amazon.com/gp/product/B07R58VXKT/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1 They aren’t quite as nice of boards…but they’ve worked just fine for me so far and at $17 for 15 boards are still pretty cheap.

Ok, unless I forgot something that should be all you need to make your own boards. So I’m going to break this into another post and show my process step by step start to finish to prep and make a board!

[juggle]

Alright! Let’s make a board!

First step is to prep your PCB. You need to clean it and sensitize it.

No photos of the cleaning, I couldn’t hold my phone and wash a PCB at the same time. But basically I use a scotchbrite pad and some clorox spray first. I’ve been told you want to use a cleaner with chlorine here but never saw a good explanation why. I’ve also started trying some bar keepers friend on a damp paper towel and that seems to work well. So I guess right there first step are a few things I forgot in the last post - scotchbrite pad, cleaner, paper towels.

The site I linked in my first post suggests using the blue non-scratch scotchbrite pads. But I found when I tried them I just can’t get the board looking shiny and clean. He also suggests using TSP as a cleaner though and I haven’t tried that because my mega mart is out of it and I keep forgetting to get some when I go to the hardware store. I’ve been using the green scotch brite pads. They do leave a few small scratches in the copper…but so far I’m ok with that. Some people suggest using 0000 steel wool…but I’ve heard some sources say it’s not a great option since bits of steel wool can get stuck to the board and cause problems.

Once the board has had a good scrub I then give it a really good scrub with paper towels and acetone. This would be a decent time to wear gloves (hey, something else I forgot to list since I usually have them laying around!) more to protect the board from the oils on your hands than to protect your hands from the acetone to be honest. Until the board is sensitized it would be a good idea to only handle it with gloves from this point on. I don’t follow my own good advice so I just try to handle it as little as possible and only by the edges. This is a big part of why I wind up with some small flaws on the edges of my boards. Learn from my bad example so I don’t feel as bad about being one.

Ok, you’ve got a shiny clean degreased board. Let’s get some UV film on it!

The film is UV sensitive. So you need to work in subdued light. LED lights and incandescent lights are generally fairly safe here. If you have a yellow safelight - use it. I actually do have one why is why I was able to get some photos of this part. But the film isn’t OVERLY sensitive…so as long as you avoid fluorescent lights, direct sun, and other UV sources you should be ok.

The film comes rolled up in this black plastic - I roll it back up in that to keep it in the dark when I’m done with it. The film is kind of a pain to cut because it won’t lay flat and keeps curling back on itself. But…it’s fairly cheap so just cut a piece a bit bigger than you need and it won’t be that bad. I tend to cut off a strip the width of the longer edge of my boards then cut it into 3 pieces so I only have to fight with the roll every 3 boards.

Now, fill your bigger container with some cold water an inch or two deep toss your clean PCB in it shiny side up and grab some tape.

The UV film has protective coatings on both sides. You want to peel off the protective film from the side on the inside of the curl. So put a bit of tape on both sides at a corner…then pull them apart and hope for the best. 9 times out of 10 it will work and the inner film will peel off. If it doesn’t…hope you left some extra material when you cut your film. Just pull the tape off and try another corner.

With the inner film peeled off the UV film won’t curl nearly as bad…but will still try to curl. But go ahead and put it in the water with the inside curl towards your PCB:

Try not to let any air bubbles get trapped under it, and get it centered on the PCB. You should have left enough material here that you don’t have to be exact and can just be close enough. Carefully lift it out and put it on some paper towels to suck up most of the excess water.

You can gently drag a paper towel across the top…but do it GENTLY. You want to leave a thin layer of water between the film and the PCB! This is one of the keys to getting a clean layer of film. That water will be there until the laminator squeezes it out and vaporizes it.

Now - grab a piece of ordinary typing paper and fold it in half. Put the board in the middle of it:

This is basically to support the board as it goes through your laminator and to keep the film from gluing itself to the laminator.

You’re going to run it through the laminator 4 times. twice with each face up and feeding the opposite edge every other time. I’m not sure which side is hotter in my laminator so I’m not sure which way is best here…but the goal is to reach 140f and with most laminators that takes 4 passes. I do the flip and rotate just to try and get even heat…it may not be necessary but it’s what works for me.

When you’re done the film will be stuck to your board…and the paper:

Which brings us to yet another item I left off the shopping list. A razor blade. Just get in there and gently cut the board free. I kind of slide the blade under one edge and then just cut between the board and the paper leaving a bit of film hanging over the edge of the board. If your boards have nice smooth clean edges maybe you can just cut along them…mine don’t and I’m too lazy to sand them smooth (though it would be a decent idea…though fiberglass dust is really nasty so maybe my laziness is helping keep me healthier here…at least until the last step…drilling.)

Alright! You should now have a nice sensitized board that’s ready to be masked:

Which seems like a good spot to end this post. Coming up next - exposing the board!

[juggle]

So you’ve got your UV sensitive board - either you took the easy route and just bought one or were cheap ^h^h^h^h^h resourceful like me and made your own.

Let’s dig in and find out just how many more things I left off of the shopping list! (there are going to be a few in this post…I’m not good at making shopping lists!)

Now you need some artwork. I’ll do a detailed post on this later. But basically you’re going to print the artwork onto a sheet of the screen printing film. It should come out nice and dense like this:

For best results you’ll probably want to crank the contrast and density controls in your printers control panel to the max. Every printer is a little different but poke around in the advanced settings and you should find a way. Also - make sure any kind of ink/toner saving mode is off and your printer is set to print at it’s highest quality setting.

If you’re making your own artwork - make sure to mirror and invert the art. Anything that’s black at this point will be etched away while anything white/clear will remain copper. You also want the ink on the side that will be against your board for ultimate sharpness.

BE CAREFUL. I’ve noticed that the ink never fully sets on these screen printing transparencies. It’s not like printing on a normal transparency where the ink will stay liquid…but the ink is easily damaged. So when you put your board against the transparency you’re going to want to be very careful and NOT slide it around.

So - things you’re going to need for this next step that I forgot to list in the first post - some kind of printing frame and a UV light source!

The printing frame can be as simple as a sheet of glass and some kind of backer. The website I linked has some detailed instructions that both show you how to build one and prove that they guy who wrote all that is an engineer. I think he overthinks the print frame a bit. A bit of glass, a piece of something stiff, and a few clips are all you really need.

I go overboard in my own way. I worked in darkrooms for years so have a bunch of photography equipment including a nice contact printing frame someone gave me years ago:

So - put your artwork in the frame ink side up. Then place your board against it very gently and carefully. Then clamp it in. and flip it over. Note - the board is still UV sensitive so I’m working under a safe light here - be sure you stick with subdued light. Usually I just leave a small lamp in the corner on instead of messing with the safelight.

Now it’s time for exposure. You could just use the sun. But that’s a bit hard to do consistently. Thankfully it’s pretty cheap to build a really nice UV exposure unit. I first tried using a UV lamp I used for curing 3d printed resin prints…that was a nightmare because the light was very uneven. Instead I built the setup suggested by the site I posted in the first post. $15 for a string of UV LED’s which I applied to a sheet of 1/4" MDF cut to the size of my frame. I added some 3D printed feet that hold it so the LED’s are exactly 1" above my PCB. This gives incredibly even illumination:

Oh - I also have my fancy darkroom timer - but the exposure we’re looking for is in the 40-50 second range…so a watch with a second hand or a timer on your phone works just as well. The timer is just easy and I had it on hand. You will probably want to do a few test exposures first to dial things in for your setup…I’ll do another detailed post on that later. For mine I use 43 seconds:

And when it’s done it will look about like this:

Note the image is VERY faint at this point. The darks aren’t very dark and the lights are kind of dark. Don’t worry…that’s ok. Keep this board in the dark and mix up some developer.

The developer is made from 1.5 cups of baby bottle warm water - I aim for about 90f or so. The hotter the water the more “active” the developer is and faster it will work…but you don’t want it to go too fast at this point. To the warm water you add 2 heaping 1/4 tsp’s of washing soda.

I know…that’s an odd measurement. Why not just use 1/2 tsp and be done with it? Well…you’re actually looking for 3.6g here…and 2 heaping 1/4 tsp’s will do that more accurately than one 1/2 tsp. You could also weigh it out if you’re like that…but you’re here so I’m guessing you’re not like that.

It should dissolve really quickly. You can wear gloves for this part if you have sensitive skin, you probably won’t actually get much on you and it’s basically just laundry detergent anyway so I don’t.

Before you dunk that PCB there’s one more step! The UV film still has a protective layer over the top of it that you need to remove. So get out another strip of tape and put it on a corner, then gently pull it off. Your mask should stay on the board and the clear protective layer should come right off. If like me you left some ragged edges you could wind up with a few tears or imperfections around the edge of the board. I try not to get my designs that close to the edge…you may want to prep your boards more carefully.

Now - dunk it in! Grab one of your foam brushes and gently wipe it as it develops:

The developer will get whiter and whiter but you should see the exposed parts of the board get darker and the unexposed parts turn kind of whiteish. The brushing helps remove that whitish layer exposing the copper underneath. If you’re lazy and patient you can just let it sit in there…I brush. It only takes a minute or so. This step is pretty forgiving unless you made it way too hot. Eventually you’ll notice no more white stuff forms and you’ll be able to feel the thickness of the mask if you run your fingers over it. That’s when it’s done.

So pull it out and wash it off under some fresh cool water. It will feel kind of slimy and slick from the developer - once it’s fully washed off that will go away.

You should now have a more distinct mask on your board! This is a good time to look for any flaws or errors. I have a few small ones near the top where the UV film was damaged. Thankfully they’re in a ground plane and won’t cause any issues so I ignored them.

Next step is to beef up the mask a bit more with a bit more UV light

This second exposure is probably optional…but I’d rather play it safe than risk the mask letting any etchant through. I just re-expose the board for the same 43 seconds…I’ve heard some people say it should be re-exposed twice as long as the initial exposure - but even this gives a VERY tough mask which is kind of a pain to remove at the end so I don’t go that far. I’ve yet to have a mask fail once it reaches this point:

You’ll notice it’s even darker and more purple now. Note - it’s a good idea to have some text on your designs so you can make sure you didn’t get something flipped around backwards

Ok - we’ve got a masked board! Time to etch!

But…it’s also midnight and I’ve got to work tomorrow. So…etching will have to wait until tomorrow.

Don’t worry - the mask will be just fine, this is a perfect place to take a break in the process.

[juggle]

Also - please note. It’s taking me MUCH longer to write up this series of posts than it did to make the actual board itself. The process really does go pretty quickly. Even stopping to take photos at every step along the way this board took me under an hour from start to finish.

In fact. From when I thought “Hmm, it would be nice to have another oscillator in my rack” to when I had a fully soldered board was less than 24 hours - including a full 8 hours sleep in the middle It took me much longer to reverse engineer the circuit from Sam’s stripboard layout and draw it in kicad and layout the board than it did to etch the board or assemble it.

With a bit of practice you can go from print to solder ready board in 30-45 minutes. If you want to do silkscreen and/or soldermask it will take a bit longer…but we’ll get into those options once we’ve actually etched this board.

[eric]

Thanks for these very detailed instructions/tutorials !

Once I have assembled all of my backlog, I may try this…

[juggle]

Ok, so we’ve got a masked board ready to etch.

Let’s get out our chemicals and make some etchant!

My first boards I used 1:1 vinegar and hydrogen peroxide with a pinch of salt. It works. But it’s slow. I then switched to 1:1 HCL and H2O2 which worked much better…and after doing some more research I settled on 1:2 HCL:H2O2 which is what I now use.

Remember - always add acid to water - or in this case H2O2.

To start - fill your bigger container with water and toss it in the microwave for about 2 minutes. This will be a water bath we’ll float our etching tank in to heat it and help it work quicker.

Then take the shot glass and measure 2 shots (about 3oz) of H2O2 into the bottom of your smaller disposable tray. Now it’s time to glove up and measure out 1 shot (1.5oz) of HCL and add it to the H2O2. Finally float this little tray in the big tray of warmed water:

And in goes your board!

After a few seconds you’ll probably (hopefully) notice some of the exposed copper starting to darken:

At this point I start gently wiping the board with another foam brush (not the same one from before!) You can just let the board sit…but it will take longer. Instead I keep wiping it to remove the dissolved copper and expose fresh. If you let it just sit it will quickly darken more and more - as copper dissolves into the solution it actually makes it strong and more efficient at etching:

You’ll also note the solution start to turn green as the copper dissolves into it.

After just a few minutes of this with gentle wiping all exposed copper will be dissolved:

If you had more exposed copper your solution will be more green. Usually it gets almost as dark as simple green - this board just didn’t have a whole lot of exposed copper to etch.

Once all of the exposed copper is etched pull your board out and rinse it clean with fresh water. Note - do not do this in a stainless steel sink - this etchant is strong stuff and will attack the sink.

Heck - I made the mistake of just storing my jug of HCL in my outside SS sink 2 years ago when I noticed the jug was starting to crack. Big mistake:

Looks like I get to replace the sink when I redo my countertops now

And yes…that’s a 2 year old gallon jug of HCL which is cracked and leaking. I need to move it into a better container ASAP.

Speaking of containers. What to do with the etchant when you’re done?

Well…you can store and re-use it if you plan on doing more boards. Just keep it in an airtight plastic bottle and you should be good. If it starts to get weaker you can either bubble some air through it with an aquarium pump or add another shot of H2O2 and perk it back up. If you’re particularly analytical you can measure the specific gravity and do some titrations and determine just how much more copper to dissolve in it and oxygen to add…but…we’re hobbyists and this is a pretty forgiving process so unless you enjoy that kind of thing you don’t have to do that.

When you are ready to get rid of it…you have a few options. The solution is cupric chloride…which is many areas is sold in solid form as a root destroyer for drains. So…just dumping it down the drain with some water isn’t as bad of an option as you may initially think. Still - it’s not the best habit to get into tossing chemicals down the drain.

One solution is to add plaster of paris and turn it into a solid which you can then toss in the trash. I haven’t tried this option because I live in the desert so evaporation is an easy solution for me. I just set the tray out side with a screen over it to keep critters out and in about two days I’m left with just some green crystals. These can either be tossed into a campfire (preferably one you’re NOT cooking over) to make some pretty colored flames…or tossed in the trash.

So - with the board out of the etch and washed off it should now look like this:

We’re almost there! Next we need to get that mask off to expose our copper.

For this you’re going to re-use your developer from when you made the mask. Toss the tray with it in it into your microwave and give it 2 minutes. That should get it good and really hot. Then add 1 heaping tsp of additional washing soda so you’ve got a really strong and hot solution. Then dunk the board in.

You can wipe it at this point to help speed this up…but it’s still going to be one of the slowest parts of the process. I usually just walk away for 10-15 minutes. Then come back and give it a brush…it may be all the mask came off at this point. If it didn’t then I usually nuke the solution for another minute after brushing off as much loose mask as possible then put it back in. I’ve had some boards strip clean in 10 minutes and others still have some mask clinging to them after 45.

You can also use acetone to remove the mask…but I’ve found that’s trickier and still takes a good soak and I don’t have any trays I can fill with acetone (acetone will destroy most plastics) and who really wants a tray of acetone sitting out stinking up the kitchen anyway. If the mask seems particularly tough then while re-heating the developer I will sometimes give the mask a gentle scrub with a blue scotch brite pad and a bit of acetone to help weaken it and help the developer finish it off.

Once the mask is stripped you’ve got a board!

Now is the time to bust out your magnifier and continuity meter and check out your work. Make sure all the traces that should connect have continuity and none of them short to ground or each other. If it’s not perfect you may be able to save it. If there are shorts you can get out a razor blade and cut through them. If there are breaks you can use a little bit of cut off component lead and solder it across to repair it.

My first few boards I had to do quite a bit of that. Since I got the printing process dialed in to make a good negative and got my exposure dialed in I haven’t had to deal with it and just get boards that work…assuming the circuit itself is designed correctly

If you’re doing a SMD board you’re now done. Go solder down your devices and enjoy. If you’re building through hole…then there’s one more step. Drilling.

Drilling sucks. It’s my least favorite part of the process. It’s fiddly and if you mess up it can be hard to repair the mistake. So let’s take our time and go slow here. First off…one more thing you need that I forgot to add to the list:

PCB drills. These are really easy to break…but thankfully they’re also cheap. I got 5 sets of 10 sizes for $18 off Amazon: https://www.amazon.com/gp/product/B07S636JHY/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1 You can also get them off aliexpress, banggood, ebay and other places for equally low prices.

For most components you’ll want the 0.8mm drill which is the one with the orange ring in that photo. Some things are spec’d for 0.7mm but unless they have really small pads I usually just drill them as 0.8 anyway. You’re going to want to start with the smallest size and work up - you can always make holes bigger but its incredibly difficult to make them smaller. On things like IC’s, headers, and IDC sockets it can help to use a larger 0.9 or 1mm bit so there’s extra wiggle room. (plus some of them have wider pins that call for a larger hole.)

I’ve used three different methods to drill

1. By hand. A pin vise makes this a lot easier but I couldn’t find mine so my first board I did the absolute hardest way possible. I just held the drill bit in my fingers and did it by hand. I used one hand to steady the board and hold the bit in place putting gentle pressure it the back of the bit, and with the other hand I grabbed the ring on the bit and twirled it back and forth between my fingers. This actually works. On the plus side it’s the easiest way to be super accurate with your hole positioning. On the downside…it’s slow…your fingers will get sore…and the fiberglass dust is itchy.

Note - Any of these methods will create fiberglass dust. Which is BAD stuff. You do NOT want to get it in your lungs. You could drill under water to keep down the dust - but that raises it’s own set of issues. Thankfully we probably all have facemasks available now. But I use an actual respirator because I have one and would rather be safe than sorry. Usually a decent respirator with replaceable filters is pretty cheap. I paid $17 for mine and love it. I use it while doing yard work to keep the dust out of my mouth and nose, I use it when woodworking or painting, I use it when working with my 3D resin printers…it’s comfortable and has interchangeable filters so I can use a particle filters in the yard and woodshoop and switch to charcoal vapor filters when working with chemicals. Unfortunately right now due to COVID they’re more expensive and harder to find

2. Drill press. I have a “small” benchtop drill press. I used this for my next few boards. It works…but it’s not ideal. Even on it’s fastest speed it’s not really fast enough for these little bits. It’s accurate…but at these sizes not quite as accurate as I’d like. It’s also loud and sits at a height that’s too low to use while standing but too high to use comfortably while sitting. It works…but I wanted something better.

3. DIY micro drill press. So I bought a cheap micro drill motor off Amazon (https://www.amazon.com/gp/product/B06XGQ24PS/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1) (yes, my mailman is getting really tired of my Amazon habit) and 3D printed a drill press around it:

It’s not bad. But it’s also not great. I’m working on improving it. The original design is here: https://www.thingiverse.com/thing:1444938

My upgrades are here: https://www.thingiverse.com/thing:4633213 I basically redrew almost the whole thing. I added bearings on the shaft, made it a few mm taller to fit the smooth shafts I found on (you guessed it) Amazon, added some extra bolts to keep the top on better, added threads to the handle so you just screw the rod in, snugged up some of the clearances, added a hook on the side to hold the chuck key…little things. I still need to redo the motor mount and bearing carrrier. The motor mount because it’s kind of a pain to install and just threads into the plastic of the bearing carrier (I want to add some nut traps) and the bearing carrier because the bearings fit too loose and cause some slop - which is my biggest complaint about it as that makes it hard to get accurate holes.

Anyway…with either drill press a lighted magnifier REALLY helps:

If you do your own design in kicad (or work from an existing kicad design) you can choose “small drill marks” when generating the negative. The nice thing about those is they act kind of like a center punch mark to help center the drill. Just go slow and as the tip of the bit meets the board it will pull the board into position…then hold the board well and pull the bit in.

When you’re done a bit of backlight will help you check for missing holes:

Well, in theory it will. I still missed two holes that I didn’t notice until I went to install the components in them. Can you spot them?

Now…you’ve got a usable board! Break out the soldering iron and:

If it’s a proven board design you should be in business.

If it’s your own design…and it doesn’t work…now comes the fun part

A few cuts with an xacto and a few bits of 30awg wire can work wonders for fixing circuit design errors:

Oh - and that messy area just under where I did the re-work? That one trace was just a little iffy so I chose to reinforce it with some component lead…but I got a bit of solder on the ground plane. Doesn’t hurt anything but does make it a bit messy looking.

Two big things to watch out for assembling these self-etched boards:

1. You don’t have a solder mask (you can though…I’ll do a post on that later) so be careful of creating solder bridges…it’s a lot easier than it is with a solder masked board.

2. You don’t have a silkscreen. So you have to be a lot more careful placing components. There’s a nice plugin for Kicad that will generate an interactive BOM that helps a lot:

When you select a row from the BOM on the left it will highlight the components on the board layout. You can then check them off to track your assembly process. Very nice and super helpful when you don’t have a silkscreen to help you along!

But…you could have a silkscreen. I’ll follow up later with a few options on that.

For now I need to get back to work. So updates later tonight on how to do silkscreen and solder mask. Details on how to build the exposure rig, how to get good negative prints, and how to dial in your exposure coming soon!

[juggle]

Not sure how much detail I should go into on the other side-topics I’ve mentioned since there doesn’t seem to be quite as much interest as I expected. If anyone wants me to go into details on the various printing options I tried before settling on what I’m using let me know and I’ll gladly share. Same with anything else I offered to expand on.

I did etch up a power supply board today though and took a few extra minutes to show you you can do both soldermask and silkscreen with the same basic process. I didn’t do this step by step but here’s the results:

Now, admittedly I did kind of rush through these. So the results are super great. But the process is basically the same. Just for the silkscreen you don’t want to invert the output so you get black dots on a mostly blank sheet:

And the silkscreen you should not make the mistake I did and mirror it. You want that non-mirrored but inverted so the stuff you want to have show up is white and the areas you don’t want to print are black:

Which means you have to generate gerbers 3 separate times with different options selected each time. And be sure to select the correct layers to generate with each set of options.

Honestly though…I think the silk and soldermask are more trouble than they’re worth. The solder mask can be nice since it helps protect the copper and makes it harder to bridge something when soldering…but it’s REALLY tricky to get it to line up well.

The silk is tricky to line up…and the lines are so fine it’s hard to not damage them when developing the UV film. It may benefit from a little extra exposure…or hitting the blank PCB with some fine sandpaper first to give it something better to attach to.

I did do both sides at once as practice for doing a double sided board. I pretty much just applied UV film to one side…then once the board cooled did it on the other. Though I think this may be part of why the soldermask got a bit damaged…so it may be better to laminate both sides at once.

I did expose them one side at a time. Then peeled the protective film off both sides and developed both at once. Then re-exposed both sides again to really set it.

One other issue I have with the soldermask…the holes it leaves for soldering can be pretty small. On this board I had to re-drill the holes for the power terminals larger so the ones I want to use will fit…and there’s so little pad left I had to scrape away some soldermask to have enough board to solder to. That’s probably mostly my fault for not using quite the right footprint. But…still something to watch out for!

[Farabide]

Keep posting. It’s all good. I’ve always painted on soldermask after I’ve built the project - saves a lot of faff.
I say always … Ive rarely had to bother. Spray lacquer protects copper just as well and I’ve once sprayed an entire project black on completion but that was for a hitchhikers guide computer game gift.

[juggle]

Yeah, I’ve considered just coating after finishing. I do have liquid soldermask as well…I just haven’t tried it yet since I haven’t worked out a good consistent way to apply it at home.

But…I find myself re-working my boards enough right now that I’m not ready to “lock them in” just yet. Once this second case is fully operational and they’ve proven themselves then I’ll be ready to declare them “finished” and seal them

[juggle]

Note to self. It really IS worth it to include some text on your copper layers so you can easily tell if you get something backwards.

What I meant to do:

What I actually did:

The -12v regulator let out some nice magic smoke

[itsnotworkingatall]

Instead of HCl you can use citric acid that can be bought in any grocery store It works quicker than muratic acid, and way more safer: you can work with it inside (just open a window a bit if you smell something).

30 grams of citric acid + 100 g H2O2 + 5-10g of salt as catalyst is enough for etching one side of 10x10 cm board.

[juggle]

Interesting. I hadn’t heard of using citric acid before. I actually have some on hand so I may give that a try on my next board!

[itsnotworkingatall]

If you are interested, here is an article (google translate widget in top right) where a guy compared 5 different methods of etching with pros & cons of each. Citric acid is the cheapest and the safest one. I’ve always used this method of etching and never wanted to try something else. Usually, it takes somewhere around 20-30 minutes to etch a board. Sometimes even faster - depends on an amount of copper that is needed to remove.

РадиоКот :: Безопасный общедоступный состав для травления меди в домашних...

[juggle]

20-30 minutes is quite a bit slower than I’m getting with HCL + H2O2. When the weather was warmer I was etching them in 5 minutes or less. Now it’s more like 7-8 minutes since we’re under 80f outside finally.

The article seems to make an interesting case for it and I’ll definitely give it a try. I already had the HCL on hand. I tried vinegar first and it worked…but even heated it took longer than I wanted so I gave the HCL a try and haven’t looked back. Of course…you can see from the photo of my sink the risk of improper storage. (though that damage was done well before I decided to try etching PCB’s again!

[Farabide]

I have an old desktop plotter somewhere and wondering if I could combine it with conductive ink. Just a whim. Soldering will probably be an issue. Thoughts?

[juggle]

I have a big old HP plotter as well. Though I should probably get rid of it since I haven’t used it since I got the CNC and it takes up quite a bit of space in my tiny shop. (I used to use it to print full size 1:1 plans for RC planes - but now I just cut them out on the CNC directly.)

But getting regular ink for it is a big enough pain…it’s an inkjet style plotter and uses old HP cartridges that are getting harder and harder to find. (And of course are always dried up when I go to use the darn thing.)

The plotter was actually a fun story. A friends boss had it sitting in a storage shed and didn’t want it anymore so offered it to me for free. When I got it home I found that mice had been living in it The sides were full of nests and packed full of dogfood the mice had stolen from his dog. I had to do a TON of cleaning…and of course the mice had chewed through various wires. Still kind of amazed I got it working it was so bad.

Conductive ink is something I’ve wanted to play with before but never quite got around to…may have to try it one of these days…

[Farabide]

Mine is only for A3 and I was thinking if nothing else I could use it to help etch a number of circuits or a really big one onto a board for a controller or large face plate. Not sure if it’s working and I can’t remember if it’s centronics, serial or a weird SCSI interface. I just remember stowing it away for some odd PIC project I never got round too. Stick a laser or drill where the pen is and slow it down. The mechanics on plotters are amazing. An old friend of mine made wallpaper for his dining room using a huge plotter.

[binarymob]

What’s a good source for single side designs. I’d like to give this a go over the holiday.

[juggle]

Most of the Yusynth boards are a good starting point. The ARP VCF is an exception where the PDF has bitmap instead of vector data - so it’s a bit harder to get a clean negative from it. But the rest of his boards I’ve tried have been really easy and fit on the cheap 75x100mm boards. (don’t quote me on the size…I’m horrible at remembering numbers and I’m about to go to bed for the night so my memory may not be it’s sharpest right now!)

MFOS also has some single sided designs with published artwork - but I haven’t tried any of those yet.

I’ve also shared my take on the M4TM 808Kick (a GREAT Kick!) https://github.com/jhitesma/Modular-Synth-Build/tree/main/808Kick

And my layout of LMNC’s simple 3340 VCO:
https://github.com/jhitesma/Modular-Synth-Build/tree/main/cem3340/cem3340 (The files are a bit of a mess…I haven’t got around to cleaning them up and doing a README yet…but I’ve made two of them - one with an edge mounted board and one with pots on the board and they both work impressively well for how simple they are.)

I’ve seem some others out there as well but those are the ones that I started with and that come immediately to mind.

[binarymob]

rad thanks.
I’ll peruse the Yusynth archive. I’ve been holding off on his designs till I’ve leveled up a bit more.
Do you know if anyone on this board has applied any of the AS3340 PWM fixes to their designs yet. I’ve been meaning to build another.