My way to professional prototype at home.

[IanB]

his chemicals are not that specialDeveloper is Sodiumcarbonate or potassium carbonate.

In the YouTube video his plastic bottle says sodium hydroxide on the label.

I imagine sodium carbonate would work too, but would be a little less aggressive and therefore more controllable. Also much less corrosive to human skin.

[free_electron]

In the YouTube video his plastic bottle says sodium hydroxide on the label.

Sodium hydroxide is the 'stripper' for the photoresist ( the moment you see the blue stuff peel off in chunkss and float away is sodium hydroxide.
After exposing to UV light he develops in sodium carbonate or potassium carbonate.

anyway. you can buy that stuff on ebay. foil + developer for cheap.

By the way, Bungard ( well known pcb maker ) sells all that stuff too.

[mazza85]

I've only made pcb at home using store available board that came with the photosensible mask already(I don't know their exact name), so I have some question.
Why he don't work in low-light with the photomask?  He "activate" it when drops some liquid (which?) between the board and the printed acetate before UV exposition? 

Sent from my phone with Tapatalk

[mamalala]

Thru hole metallisation. That's just sticking a bunch of wires in holes.. I thought somebody finally had cracked the home rout to copper plating without all the chemicals.... A bit disappointed

Well, completely without chemicals would be a problem. But there is an easy way to do the hole-wall activation without any special chemicals.

I do my own through-plated PCB's occasionally if i want them quickly. First is drilling. I placed a small b/w camera module below my drill-press, looking at the drill. Made a small µC circuit that simply overlays a crosshair onto the image, adjustable in x and y direction using two potentiometers. The result is fed into a small b/w CRT monitor, the ones used for camping and stuff, running off of 12 volts.

As for the hole wall activation. There is carbon conductive ink available. I use the "SD 2843 HAL" from "Lackwerke Peters". It's a thixotropic paste that becomes a bit more "fluid" (like honey) when stirred. Dilute that with a little bit of ethanol, so that the consistency is like warm honey.

After drilling i sand the surface of the PCB with fine sanding paper to remove the ridges around the holes after drilling. Then i pour a bit of the thinned carbon ink on the PCB and press it into and through the holes with a squeege. Repeat the same from the other side with the excess ink. Then suck out the ink from the holes with a vaccuum cleaner. Place the PCB on a flat surface. Put some piece of cotton canvas or similar on a "sander block". It's just a block of cork. Wet it with ethanol and wipe over the board to remove the ink that sticks to the surface. Don't apply too much pressure or you will wash away some ink from the hole/copper boundary, loosing contact.

Cure the PCB/ink in an oven, at around 150°C for 10 minutes or so. After cooling down, another round of sanding the surfaces is done to make sure that any ink on the surfaces is remove, if needed.

Now place the board in a copper electrolyte bath for electroplating, add the copper anodes and turn on the current. My electrolyte mixture is:

1700 ml H2O
490 ml H2SO4 38% (battery acid)
280 grams CuSO4
300 mg NaCl
2 ml Tween 20

The latter is an agent that greatly reduces the surface tension of the liquid. Make sure to use pure copper for the anodes, otherwise you will get a bad plating. If you can't find pure copper, at least wrap the anodes into a paper-towel and then place it into pouches made of cotton fabric.

Grab an old oscillating desktop fan sans the blades. Attach some stiff wire to it, and attach the other end of the wire to whatever holds your PCB in the bath. That is used to gently move the PCB in the plating bath so that the electrolyte can flow through the holes.

Start the whole thing with a 18µ coper-cladded PCB. For a euro-sized PCB (160mm x 100mm) i use about 1.5 to 2 amperes, for about an hour or more.

When the plating is done, wash the PCB. If needed (due to impurities in the copper anodes, for example) slightly sand off the surfaces again. Laminate with tenting resist. Expose, develop, etch.

Attached are two images of how the final result looks like. One is chemically tinned, the other is "blank".

Just the ink gives a few ohms per hole between the two sides. In case you only want to make a test prototype, and you don't need high currents or precise analogue stuff, you may not need to do the electroplating. But i have no idea how long/robust that will be in the long term.

Greetings,

Chris

[Short Circuit]

Nice work!

Your PTH process sounds very similar to that expensive LPKF system
http://www.printtec.nl/contents/nl/d378.html
http://www.printtec.nl/contents/nl/LPKF-ProConduct.pdf

[jucole]

SUPERB!!! thanks for sharing that! ;-)

[mamalala]

Nice work!

Your PTH process sounds very similar to that expensive LPKF system
http://www.printtec.nl/contents/nl/d378.html
http://www.printtec.nl/contents/nl/LPKF-ProConduct.pdf

Well, _every_ PTH process sounds similar to each other. It always starts with drilling the holes, goes on to make the hole-walls conductive, and then electroplating. The LPKF stuff is so expensive because they use a silver based paste. And IIRC, it's not meant to get electroplated in the end at all.

However, the process i presented differs a little bit from how it's done in the industry. That is because i'm electroplating the whole board. The "real" process only electroplates the tracks and holes, not the whole board. First the board is drilled, then laminated with photoresist (or dipped in photosensitive ink, Peters has that stuff as well). The layout is then exposed inverted. That is, all the parts that are later etched away are covered. All that is meant to be left is "open". That gets electroplated, including the holes. After that it gets electroplated with a thin layer of tin. It is that tin that is the real etch-resist. The photoresist is stripped off, and the the board is etched.

I have no idea if, instead of electroplating the tin, a chemical tin can be used, never tried that. If that would work it would be great.

Anyways, my idea behind the process i came up with was to use as few chemicals as possible. It's, uh, 4 or 5 years since i came up with it, and it works. Of course it's not the "real thing", but it does the job well. There are a few perks with it. For example, it's hard to control how thick the deposited layer of carbon ink on the hole walls is going to be. So you have to add in that factor, and make the holes for THT parts big enough so that they still fit in after plating. That of course increases the minimum size for vias and pads.

But then, it's not meant (and very likely not usable) for replacing the real industrial process anyways. It's good enough for home use, and surely much faster and less work than using tiny bits of wire to make via connections. After all, the majority of the time required is just waiting for stuff to complete, the plating taking the most time. But even that can be shortened if you only need a quick prototype. No need to have a thick copper layer deposited for that purpose, just enough to get 0 ohms (and filling it up with solder if higher current demands exists in the circuit).

Greetings,

Chris

[mamalala]

Attached are three more images. "solderside.jpg" and "soldered_through.jpg" are from my proof-of-concept trials. Made a bunch of holes on a PCB, plated them as described, hand-milled a grid on one side after that to check continuity. Then i soldered in some wires to check if it nicels solders through to the other side. It does so nicely, as can be seen.

The third image is just showing a PCB that just finished the electroplating process. That's how it looks like when it's done properly.

Maybe i'll start a new thread, explaining the whole thing step-by-step, with images. Just need to find time for that.

Greetings,

Chris

[jucole]

After watching this video I went out and bought a cheap laminator;  I added a simple push button to start/stop the motor so I can manually slow it through the rollers - works nice!!

[Pat Pending]

Wow!!

Two questions

Are your photo-positives laser printed? if so, how as your plots look very good.
Is that water or mineral oil used to hold the positives in close contact with the copper surface?

Great technique.

[poorchava]

It probably depends in the printer. Some laser printers can give you good coverage of large black regions, but vast majority do not. This causes polygons to be spotted with tiny holes.

Quite actually affordable way to get perfectly black photomaska is to ask some company doing business in advertisment (ya know - billboards, stamps, gadgets, business cards, etc). They usually charge reasonably low (like $2-3 a pop?) and quality is much better than anything you could possibly do at home.

[flolic]

About printers and photomask printing:

https://www.eevblog.com/forum/general-chat/diy-pcb-inkjet-printers-and-paper/msg77444/#msg77444

[jucole]

It probably depends in the printer. Some laser printers can give you good coverage of large black regions, but vast majority do not. This causes polygons to be spotted with tiny holes.
Quite actually affordable way to get perfectly black photomaska is to ask some company doing business in advertisment (ya know - billboards, stamps, gadgets, business cards, etc). They usually charge reasonably low (like $2-3 a pop?) and quality is much better than anything you could possibly do at home.

In the printing business, generally speaking you need more resolution to print b+w "line art" than greyscale or colour in order to remove the appearance of the blocky "polygons".  Although the consumer laser printers are pretty amazing, a print house would typical use an "Imagesetter" which will print at ~3000 dpi.  The old imagesetters would use something called "Rip" (Raster Image Processor) , The Rip would take a vector format then rasterize it into raw pixel data very quickly then send it to the Imagesetter.,  But now I think most modern Imagesetters have the Rip's built in. 

[flodins]

New pcb in that technology

[lewis]

Well done sir! Keep up the good work!   

I don't see why lots of people in this thread are/were decrying Flodins for wanting to produce PCBs at home. Being able to produce your own PCBs has a lot of advantages. The main one being that if you get the design wrong first time around you don't need to wait/pay for a re-spin. Another big one is the speed at which you can have something up and running. And the unit cost is almost nothing, if you've got the gear.

We make our own prototype and small production PCBs in-house, and can do it extremely fast. After Christmas I designed and manufactured a customised product in two days because my customer needs it by 4th Jan. From initial concept, design, MCU programming, PCB manufacture, assembly, testing, through to packing it in the box was 2 days. We couldn't have done that as quickly or cheaply without in-house PCB production, and I applaud anyone who wants to have a go at it.

[fenclu]

Personally, I can recommend linuxgeek81's channel. He does some excellent work as well:

https://www.youtube.com/user/linuxgeek81

One of his boards:

[jerry507]

There's nothing wrong with wanting to do your own PCBs, and I don't think anyone was saying it was exactly. There were the safety concerns (very valid and meant sincerely) and the cost aspects.

For my small business and employer, we will never do our own PCBs. Most of ours decently complicated, and frankly TIME is the chief advantage, not money. At a standard billing rate of >=100$/hr, spending even an hour or two quickly approaches the cost of having them made. We use Advanced Circuits frequently and pay about 450$ for our prototypes. That's less than 4 hours of my time to the customer. By the time you add in the associated maintenance of everything, you're not saving that much.

So the trade off is time making pcbs vs time doing other stuff. Neither answer is the RIGHT one in general; it's very specific to your usage. For hobby, absolute cost is typically the overriding factor, not the money value of time.

[IanB]

For my small business and employer, we will never do our own PCBs. Most of ours decently complicated, and frankly TIME is the chief advantage, not money.

OK, but:

Time = getting PCBs done faster
Money = getting PCBs done cheaper

At a standard billing rate of >=100$/hr, spending even an hour or two quickly approaches the cost of having them made. We use Advanced Circuits frequently and pay about 450$ for our prototypes. That's less than 4 hours of my time to the customer. By the time you add in the associated maintenance of everything, you're not saving that much.

Your argument here considers cost, which does not support your initial premise that time is the main factor. If time was the factor, i.e. getting the PCBs done as fast as possible, then you would do whatever it takes no matter the cost.

So for your situation, apparently, you are prepared to wait to have the PCBs done outside even if it takes longer because overall that way is more economical. For someone else, like "lewis" above, a rapid turnaround may be absolutely critical and this may justify the increased cost of making boards in-house.

[MrsR]

I liked your video, but I will stick with may cad and Pen Plotter resist mainly because my boards are only SMD and measure 19mmx24mm and 50 parts. If I could do the masking system as easily as you seem to do I would swap for my test boards.
Rachael 5 Stars for your Videos though

[HeliEye]

Cool Video... Was not so keen on the laminate method, I make all my in house prototype artwork on Inkjet film.
The time from printing the artwork on film to having an etched board in my hand takes 25 minutes (assuming I had pre-heated the etchant bubble tank)
Most of my pcb's are all SMD, so rarely do I have to drill anything.. a bonus.

But... I really like one bit of the laminating.... The Film Solder Resist..
Just this one part of the process is making think about this film and laminator's 

Nice work

Steve

[poorchava]

Having CNC machine at home also helps tremendously. And yuu don't need any superb rigidity. MDF/plywood+drawer slides+cheap steppers+threaded rods+dremel will do miracles



Result:

[Psi]

Having made quite a few home PCBs myself i can say that drilling the holes exactly in the correct place is one of the hardest parts.

Without a CNC it's very hard to get them perfect.

[poorchava]

I actually think that getting alignment right is the hardest part. When you transfer first and then drill holes, you can get hole perfectly centered with the pad. When you have holes drilled first, you have to keep alignment of paper with trace patter with the hole pattern at all times, which is not that easy when using hacked laminator (it has a tendency to pull pcb and paper from your hand). On the other hand having holes drilled first really makes fabricating double-sided pcbs a breeze, because you have common reference points visible on both sides.

I'm thinking about making a video about how i'm fabricating my pcb's

[carloscuev]

Well, completely without chemicals would be a problem. But there is an easy way to do the hole-wall activation without any special chemicals.
...
...

Hello, I've been trying to do electroplating following this video: http://youtu.be/KTNuTv_IQp4 which uses the formula from thinktink (http://www.thinktink.com/stack/volumes/voliii/consumbl/cplatmix.htm) except for the Copper Gleam CLX he's using Polyethylene Glycol (PEG) 3350. I've used PEG 3350 and PEG 6000 with same results (see attached picture) would you have any clue about what I'm doing wrong or what I'm missing?

I've noticed that you are using less sulfuric acid but more copper sulfate than the thinktink recipe, also you are using Polysorbate 20, do you think that Polysorbate is better than PEG? I think I know where I can get polysorbate in my country, but I have to go to the other side of the city.

Hope you could give me some clue

[mamalala]

Hi carloscuev,

the polysorbate 20 (Tween 20) is simply to reduce the surface tension of the liquid, so that it can enter into holes easier.

I think there are two issues you have. From the picture you showed you have too much current, and your copper anodes are not pure copper. The high current causes high current density at the edges, so that more copper will deposit there. Regarding the grains on the surface, i get similar grains with my anodes too, which are an copper-alloy. The foreign metal cause to deposit as rather large grains onto the surface, which in turn causes a higher current densitiy at that point, causing more copper to deposit there.

So, first you should try to reduce the current. For a double-sided Euro-sized board (100 x 160 mm) i use roundabout 2 Amperes. If your sample is smaller, then you have to reduce the current accordingly. That is, for example, 1 Ampere for 80 x 100 mm, 500 Miliampere for 50 x 80, etc. In any case, start with a lower current first, then increase it a bit every 5 or 10 minutes, repeat until you reach the desired current. I spread it over 3 to 4 steps usually. Also, keep the stuff you electroplate moving through the liquid, and at a proper distance to the anodes. I have no less than 2 cm distance at all times.

Regarding the copper anodes, if the above does not help and you still get a grainy surface, the anodes are not pure copper. In that case remove the anodes, wash them under running water and then dry them. Then wrap them in one or two layers of soft paper towels, the stuff you use in the kitchen (alternatively you can use toilet paper as well). After that put them in a "pocket" made of cotton fabric. The idea here is that only the copper gets into the fluid (as copper ions), since the paper acts as a filter, while the cotton pocket holds back the wet and fragile paper (after all, you do not want any contamination in the fluid). but the other metals are held back since they do not go "into solution". That should definitely stop the grains from appearing then. Make sure that there are no holes or openings in the wrapping/cotton-pocket.

Last but not least, make sure that the part you want to electroplate is absolutely clean. It can help a lot if you micro-etch the PCB first. Simply place it into the etchant after you cleaned it, for maybe half a minute maximum. Remove from the etchant, rinse clean under water, then place into the electroplating.

Greetings,

Chris