My way to professional prototype at home.

[mamalala]

Forgot to mention:

Check pharmacies for the Tween 20, or shops that produce beauty related stuff. Also, you can find it on Amazon as "tween 20".

I have no idea what the hydrochloric acid is used for in the receipe you linked to. I don't use that but use regular salt. The salt helps to get a shinier surface.

What is that "Copper Gleam" supposed to be? Sounds like some super-special-snake-oil stuff that is only there to fill the pockets of the company who makes it. Never heard of it, never used it.

It's always best to stick to simple receipes that do not use any exotic stuff. All you need it water, sulphuric acid and copper sulfate. The salt and Tween 20 is already optional, so to speak, although the Tween helps with the surface tension (and thus the tiny holes) and the salt helps with a nice surface. But neither of them are really needed, it would simply mean that you have to put a bit more effort in the process to, for example, get rid of air pockets in the holes.

Greetings,

Chris

[carloscuev]

Forgot to mention:

Check pharmacies for the Tween 20, or shops that produce beauty related stuff. Also, you can find it on Amazon as "tween 20".
...
...

Thanks so much Chris, I've made a test with a bag made of 1 micron polypropylene felt (conveniently I have a close source of this material) and results were way beter, I'm having small grains but now it's a lot less.

Your current density is 6.25mA/cm², given by 2000mA/(10cm*16cm*2) The small PCB in the picture is only 1.25cm x 2.9cm, has 6 0.5mm holes, and 6 1mm holes and I designed it for testing purposes (also it's a small boost converter to test zener diodes) and I have plated it at 50mA, so it's 50mA/(1.25cm*2.9cm*2) = 6.89 mA/cm² how much time do you leave the board plating? have you came up with a formula to calculate the plating thickness?

The surfactant I'm using is PEG 6000, I believe it's a surfactant somewhat used in copper plating, I found this interesting abstract about using PEG in plating process http://jes.ecsdl.org/content/152/11/C769 but can't access the entire article. I have read some people are using PEG 3350, commonly found as a laxative MiraLax or in my country ContuMax. I want to test Tween 20 but first I want to make more tests with the PEG 6000 I bought from eBay.

The Copper Gleam is a "special" surfactant for copper plating PCBs (http://www.dow.com/products/product/copper-gleam-125-s-2/), would be nice to test, but not interested as they probably wouldn't sell small quantities and ship anywhere for everyone to make their home plating baths.

About the NaCl vs HCl, I have only read that it's necessary to have a few Cl- ions present in the plating bath, the quantities need to be very small, in the thinktink website recipe they state only 12 PPM of HCl, don't know which function they serve. Also it would be nice to know which one of the 2 Cl- providers (NaCl or HCl) is best, and now many PPM are optimal.

I'm making 60ml tests each time, I'm going to do another clean solution with more copper sulfate and less sulfuric acid, the relative amounts of your recipe or course. And test a new mini board, I'll be posting results tomorrow.

Greetings,
Carlos

[mamalala]

Regarding the current, i'm using a somewhat lower current than what would be possible. I never bothered to do some formula for the time, i have it in the bath for 2-3 hours. I came up with that time once using a micrometer screw to check that about 40 µm copper was added (i start with 18 µm copper cladding and wanted a bit over 35 µm as end result). From then on i just sticked to that time. Sure, not the way to do in a professional environment, but more than good enough for home/hobby use.

As for the grains, the cotton fabric i used is really tightly woven. Since the fibers also slightly expand when wet, there a virtually no real "openings" left. My first try using only paper towels was better than without, but i also still got grains on the surface. Only after adding the cotton fabric it was OK.

Good luck,

Chris

[carloscuev]

I went to visit a local electroplating shop here in my country and I learned that good plating results are achieved using a Phosphorus-Copper alloy as the anode. He gave me a small used piece of anode to test and that worked as magic, see the attached files. In fact if you search google for "phosphor copper pcb" you'll see that is what is used by manufacturers to do vias.

In the first picture I show that the hole wall activation is made with 2 layers of simple waterproof indian ink, in this case Higgins Black Magic ink with NO graphite added as some may recommend, I found that adding graphite clogs very easily small holes, I'm looking to buy very fine graphite and do further testing, but the actual technique of applying one layer, curing, another layer, curing and then sanding the copper surface to remove excess ink is working wonders.

BTW. The first 3 small holes are 0.5mm, then there are 3 holes of 1/32" size, then 3 holes of 1/26" size.  The big one at the edge is 1/8"

[mamalala]

That looks very good indeed!

True, in professional setups they use phosphorous copper anodes. However, with a little care-taking, other copper types work fine as well, at least for a hobby/home setup. At least that is what i experienced.

Regarding the ink, that looks good as well. What i use here is a carbon conductive ink, SD 2843 HAL from "Lackwerke Peters" (a German manufacturer). They hand out samples  if you ask nicely, but not sure with international customers. But i think any other screen-printing ink of similar type/composition will do just fine as well. That ink is normally used to screen-print conductive traces/patterns on a PCB. I'm thinning it down using alcohol to give it a warm-honey-like consistency. Then i smear it into the holes, wipe off the excess and cure it in the oven. After that i sand the surface with a rather fine-grained sandpaper. After that comes the plating. So far i was able to get holes down to 0.3mm in diameter, with repeatably successful results, that is, no failed VIA so far.

If you run into troubles with the contact between copper surface and VIA, try reducing the current you use at the beginning. At the edge you have a rather high current density. If you have a very thin copper/conductive-ink layer there while using high current, you may (literally) burn it away at that edge/point.

Anyways, good job that you did there, congratulations!

Greetings,

Chris

[chickenHeadKnob]

This thread is drifting from Flodins original post, but drifting in a good way
Maybe there are other people interested in copper plating who are not following here because they do not know?

Are you (carloscuev) sucking ink through the holes  with a vacuum? it appears mamalala isn't. Also would be interested in hearing if either of you has experimented with limited -or asymmetric AC instead of pure DC to achieve a smoothing action.
Very interested in trying this, thanks guys.

[SeanB]

With copper plating using high ripple DC is best, simplest is to use a half wave rectifier ( single diode) with a resistor across it to provide a reverse current of about 5% of the forward current. Gives a better plating action but not as smooth. Only time you want pure ripple free DC is if you are plating chrome to a mirror finish. If there is ripple there then it comes out matt. With copper this is not a problem. Best to start off at low current and gradually go higher, sharp edges are current concentrators until they accumulate enough metal to round off some what. If you want to practice without drilling holes take a leaf and paint it with the ink then electroplate it till it is thick then peel it off. You can then get a negative of the leaf surface.

[amspire]

Thanks for this thread. I had just about dismissed the idea of making my own plated through hole PCBs but you have got me interested again.

I just went to my local art shop and got a bottle of Daler Rowney FW Acrylic Black ink and did a quick test. Without any graphite, I can get a about 500 ohm resistance in a 1mm hole using just the ink. Probably lower with a second coat.

Does that sound like enough to be useable or do I need to get a resistance down in the low ohms?

[BravoV]

Noob question, once the holes plating is done, how do you cover or seal those holes at the pcb traces etching process ?

[amspire]

Noob question, once the holes plating is done, how do you cover or seal those holes at the pcb traces etching process ?

I was wondering that as well.

My guess is that after copper plating, you have to add a negative mask to both sides of the board and do tin electroplating. Then you remove the mask and etch the remaining exposed copper.

Any methods that do not require the tin plating?

[SeanB]

500R is fine, start at a very low current and slowly ramp it u with time, pull the board out the first few times and look to see when you get a full coat of copper then you can ramp it up.

[mamalala]

Are you (carloscuev) sucking ink through the holes  with a vacuum? it appears mamalala isn't.

Well, i use  a vacuum to suck it back out after the ink i'm using is squeezed through the holes. The idea is that only a rather thin layer of that conductive ink stays on the hole-walls, making contact with the copper-cladding on each side of the PCB.

Greetings,

Chris

[mamalala]

Noob question, once the holes plating is done, how do you cover or seal those holes at the pcb traces etching process ?

I was wondering that as well.

My guess is that after copper plating, you have to add a negative mask to both sides of the board and do tin electroplating. Then you remove the mask and etch the remaining exposed copper.

Any methods that do not require the tin plating?

Yes. Use dry-film resist. Laminate it onto each side of the PCB, then expose and develop. That stuff is called "tenting resist" for a reason: it will tent over the holes Then etch the board and strip off the resist once done. After that you can either use chemical tin and tin the whole thinng, or use a solder-stop resist like Dynamask...

Greetings,

Chris

[carloscuev]

Oh sorry for not replying, I thought the forum would auto-notify me about any replys. Now I've configured my account to do so

Anyways, good job that you did there, congratulations!

Thanks, you also gave me some tips!

Are you (carloscuev) sucking ink through the holes  with a vacuum? it appears mamalala isn't. Also would be interested in hearing if either of you has experimented with limited -or asymmetric AC instead of pure DC to achieve a smoothing action.

In my case, the Indian Ink is too thin so if I suck the ink with a vacumm I'm left with a very thin layer with high resistance, I use a booger sucker haha (see attachment) just to unclog the holes because if curing with the holes cloged of ink, it will dry clogged. So that way, with a fat layer of ink in the hole, when curing it deposits a nice conductive layer. As mamalala said, his conductive ink mixture is about warm-honey viscosity so in his case, vacumming the holes is the way to go.

I just went to my local art shop and got a bottle of Daler Rowney FW Acrylic Black ink and did a quick test. Without any graphite, I can get a about 500 ohm resistance in a 1mm hole using just the ink. Probably lower with a second coat.
About the Asymetric AC, I haven't done that. To get smooth copper deposition you need a surfactant in the solution, I use Polyethylene Glycol 6000 (buy it on ebay) and mamalala uses Tween 20

Does that sound like enough to be useable or do I need to get a resistance down in the low ohms?

500ohm in a 1mm hole of 1.6mm PCB is what I'm getting so it's enough. Proportionally we're talking about 1kOhm for a 0.5mm hole and so on. And as SeanB said:

500R is fine, start at a very low current and slowly ramp it u with time, pull the board out the first few times and look to see when you get a full coat of copper then you can ramp it up.

Yes, I start at 25% of my calculated current, and ramp it up 25% every 10 minutes until I get to 100% By the time I've reached that point I can see in the microsope a lot of particles of copper already deposited over the ink, looks rather beautiful

Noob question, once the holes plating is done, how do you cover or seal those holes at the pcb traces etching process ?

I cover them with paint, the same process of the conductive ink but now with paint to mask them from the acid. I got this idea from a youtube video, keep reading this reply for the link please.

Yes. Use dry-film resist. Laminate it onto each side of the PCB, then expose and develop. That stuff is called "tenting resist" for a reason: it will tent over the holes Then etch the board and strip off

I'm using dry-film resist, but big holes don't get tented, however my smallest ones (0.5mm) do get tented , but anything up from there won't, what dry-film resist are you using?

Some very valuable notes (I think):

About the process in general:
As amspire I had dismissed the idea too, until I saw this video: http://youtu.be/bD9imNIkTKM and browsing that user channel found this one too: http://youtu.be/KTNuTv_IQp4 If you're interested in plating vias at home it's a good idea to watch them too. Here they cite this thinktink website (http://www.thinktink.com) that has lots of valuable information.

About curing the ink:
When using just indian ink, as in my case Higgins Black Magic with no added graphite there's a problem. If you cure the board too fast you'll get a thin non conductive layer on top of the ink, I think it's because this ink has some acrylic compund or something and if you cure it too fast you'll have good conductivity from layer to layer, but the copper won't deposit. I observed this because doing a test I cured it with a lighter, raising the temperature very fast, I could see the ink bubbling on the big holes. And then when plating no copper was adhered to the surface. So I changed to curing it with a heatgun slowly ramping up the temperature (1 min) and then letting it cool down. You'll see that the conductivity gets very affected with the curing temperature you're using. I'm looking to do further tests with and temperature controlled oven to find out the best curing temperature and ramp-up time for this ink. Also as I mentioned, I've got (from ebay too) 5 micron graphite powder, I'll run some more tests. Here's the 5 micron graphite from eBay: http://www.ebay.com/itm/121066283612

About the anodes
Make an effort to get Phosphorus-Copper alloy anodes, search for electroplating shops in your country, search in alibaba and ask for a sample (almost always you pay shipping) or do anything you can to get them, the results are way better than using other types of anodes. The guy from the youtube video says he found Phosphorus-Copper pipe in Home Mart being sold as heat exchange pipe or something like that, watch the video. The small anode I got from the local electroplating shop looks very ugly because it's used and highly eroded, but produces amazing results, I'm looking forward to buy more, I've phoned them and I got the addres of their provider. I attch a picture of the ugly but magic anode. Also in order to keep your plating solution usable for long time you'll need Polypropylene felt bags to cover the anodes as the anodes will generate a black slug and you don't want that in you solution. The local electroplating shop has been using the same solution they first made since 1994! Again, in the youtube video you can see how they look like.

About the plating solution
In the thinktink website you can get the recipe (http://www.thinktink.com/stack/volumes/voliii/consumbl/cplatmix.htm) for the solution. And as the user in the YT video says, instead of that rare Copper Gleam GLX as surfactant, you can use Polyethylene Glycol 3350 available at drug stores, even in my country (Mexico) I was able to find it, then I bought from ebay Polyethylene Glycol 6000 but both of them work well, haven't made a deep comparison of results but they were at plain sight the same. The problem I had is that I got 98% sulfuric acid instead of 35% and the problem is that this ratios are not volumetric, they are mass rates, so for a 35% sulfuric acid, 35% of its weight is sulfiric acid, and the rest is water, but pure sulfuric acid has a density of about 1.84g/cm³ so the calculations turn out to be a bit harder than expected when not dealing with the concentrations of that recipe. However I studied some days a bit of chemistry and made a PHP script to easily find the right amounts, you can find it here: http://escalalibre.com/edwt/ just click in "6. Via Electroplating Solution Calculator" and fill the requested fields according to the chemicals that you can get and hit "Calculate" to get the volumes and weights of each component for a specified amount of solution. Very Important Note: as you can see, the amount of HCl is very small, but AFAIK is very critical not to mix more than the specified amount of Cl^- ions, as a tip, if you got 35% HCl, you can mix in a separate container 34 parts of distilled water by weight to 1 part of 35% HCl by weight, with this you have 1% HCl so if you recalculate for 1% HCl using the web tool, the quantity of 1% HCl would be higher and more manageable than using 35% HCl.

About the plating time and layer thickness
In the thinktink website you can read about this: http://www.thinktink.com/stack/volumes/volvi/copplate.htm but I'm not sure if the same would apply if not using their Copper Gleam GLX surfactant, something tells me it doesn't matter, so I also made a simple calculator for the plating time and current. Again, you can find it by going here: http://escalalibre.com/edwt/ and clicking in "7. Via Electroplating Current and Time Calculator" and as you can read there, the current density is recommended to a maximum of 20 amps per square foot (ASF)

Hope all of you can find this large post useful.
Cheers!

[amspire]

A lot of great information there. Thanks. I will give the process a go.

So it looks like for double sided boards, I will have to give up the toner transfer and go to dry film.

And I had just finally modded my laminator too to boost the temperature!

Richard.

[carloscuev]

A lot of great information there. Thanks. I will give the process a go.

So it looks like for double sided boards, I will have to give up the toner transfer and go to dry film.

And I had just finally modded my laminator too to boost the temperature!

Richard.

Well, I've tried toner transfer and it worked, but now there's a new problem, not only the alignment of the 2 layers is important, now the new problem becomes the alignment of the 2 layers and the PCB, take a look at the youtube video, the guy still uses toner trasnfer

[chickenHeadKnob]

hope all of you can find this large post useful

Yes! very useful, muchas gracias. You are saving us all time and frustration. I have been looking for phosphor-copper alloy and there seems to be multiple types: ones with relatively low phosphor around 0.02% upto 0.06% and just copper which is intended for electroplating and then higher mixtures plus other metals which are intended for brazing rods, ect. I am thinking that the problems with just using regular copper sheet or bars isn't so much the lack of phosphor but rather the alloying metals like zinc interfere and pollute the electrolyte. If I find a good source for anodes I will let the forum know. Searching right now just in Canada/U.S. to start.

[carloscuev]

I am thinking that the problems with just using regular copper sheet or bars isn't so much the lack of phosphor but rather the alloying metals like zinc interfere and pollute the electrolyte.

Yes, I think so too. I was trying at first with spirals of nude grouding wire, awful results. I got better results with plumbing copper pipe. I think this wire made just for grouding is made of a nasty mixture of copper, zinc, carbon and what not. But definitively Phosphor Copper works miracles. I don't know the content of phosphor on this one, all I know is that it's called Plus 4 Phosphorus-Copper (http://www.galvanolyte.com.mx/indmeta.htm) maybe that 4 means 0.04% ? I'll call and ask

[peter.mitchell]

Im curious on this, has anyone tried plugging vias in DIY boards with solder paste, eg;

Drill out holes for vias
Get a griddle/electric frying pan/small cooktop
Place a large, very flat block of Aluminium on top of it.
Using an accurate thermocouple, get the temp to 190~ degrees
Place board on plate, wait for board to reach temperature.
Using a syringe, carefully fill the holes with 63/37 eutectic solder paste (smallest particle size preferably, with no clean flux)  until it just comes out the top and wets the surface on the top.
Turn off the heat, wait for temp to drop to 175~ degrees
Flip the board over, turn the heat back on
Fill vias from the other side in same manner.
Turn off heat, wait for temp to drop to 175~ and remove board to cool.

Using an aluminium plate would be nice, keeping an even temperature across the board. Additionally, 63/37 won't stick to it, and its fairly soft, so easy to mill flat.
Flatness would be important, this would stop the solder from flowing out the other side of the board, and would apply even heat to the board, meaning minimal temperature needed.
Eutectic solder would be good, because you wouldn't need the board to fully cool between turning it over and it is very easy to tell when it is a safe temperature to flip.
Small particle paste would be nice too, to allow it to fill the voids easier.
No clean flux in the paste may be necessary; if flux residue got caught inside the via you wouldn't want it to corrode your board.

[krivx]

How do you stop the solder from failing out of the vias when you solder the parts?

[GiskardReventlov]

@flodins, nice job!  This post started almost 2 years ago, how has your process changed since then?

[peter.mitchell]

How do you stop the solder from failing out of the vias when you solder the parts?

Surface tension should hold it in there just fine.