ETCHANT Question

[janoc]

don't forget how much the chinese government likes to steal designs.. good reason to make your own if you are doing prototyping prior to mass production where you can get a market following.

Yeah right. The Chinese have an army of people poring over and reverse engineering every prototype gerbers that get sent to their fabs. Especially the small orders from hobbyists, in case it was something worth stealing. Come on. Paranoid much? 

I guess you have no idea how many boards fabs like JLPCB and similar process daily and how much work/time does it take to get the gerbers processed, without even trying to figure out what the board could be for. Naked PCB without matching schematics and firmware is pretty much worthless, not even the Chinese have so much free time and labor on their hands to try to copy your product idea just from that.

If you order assembly service, that is a bit different (there have been cases where the assembler "ran away" with someone's product and started to make copies)  but we are talking bare boards, not assembly where the factory has everything necessary to manufacture the product, including any firmware, BOM, etc. But then nobody forces you to do assembly in China and/or you should have a proper contract and means to enforce it (lawyer, etc.).

There are many reasons why etching boards at home may be worth the time and effort but this would certainly not be one. Also, what would you do if your product needed a multilayer board or something denser than 8/8 which about the limit of what is doable at home?

[nctnico]

lots of good info.

the cupric chloride method sounds interesting...and relatively cheap to try, though requires a bit more maintenance and knowledge than just using a throwaway etchant.
I might give that a go to learn about the process.

You can also start a cupric chloride process using ferric chloride. Once the ferric chloride is saturated it has turned into cupric chloride. Add a little bit of HCl (hydrochloric acid) from the farmacy or grocery store and you can continu etching. This way you can use the same etchant for decades (been there, done that). With FeCl/CuCl you will also want a spray etching machine. These work just brilliant (and fast).

Still consider the use of cheap services like Seeedstudio etc because they can make much better boards than you can.

[janoc]

Spray etcher is a bit of a luxury for a home gamer. They certainly do work well but are quite an overkill unless making a lot of boards.

You can get the same result without, it just goes possibly a little slower. Make sure to agitate the solution every once in a while to get a fresh solution to the copper. Or buy an aquarium pump and pump air in it. Works just fine.

Some people etch using a sponge full of the etchant and rubbing it over the board (the resist will survive it), that goes really quick but I don't like the idea of sticking my fingers into a rather aggressive chemical, gloves or no gloves (gloves tear easily).

[CopperCone]

don't forget how much the chinese government likes to steal designs.. good reason to make your own if you are doing prototyping prior to mass production where you can get a market following.

Yeah right. The Chinese have an army of people poring over and reverse engineering every prototype gerbers that get sent to their fabs. Especially the small orders from hobbyists, in case it was something worth stealing. Come on. Paranoid much? 

I guess you have no idea how many boards fabs like JLPCB and similar process daily and how much work/time does it take to get the gerbers processed, without even trying to figure out what the board could be for. Naked PCB without matching schematics and firmware is pretty much worthless, not even the Chinese have so much free time and labor on their hands to try to copy your product idea just from that.

If you order assembly service, that is a bit different (there have been cases where the assembler "ran away" with someone's product and started to make copies)  but we are talking bare boards, not assembly where the factory has everything necessary to manufacture the product, including any firmware, BOM, etc. But then nobody forces you to do assembly in China and/or you should have a proper contract and means to enforce it (lawyer, etc.).

There are many reasons why etching boards at home may be worth the time and effort but this would certainly not be one. Also, what would you do if your product needed a multilayer board or something denser than 8/8 which about the limit of what is doable at home?

yea you can totally prevent someone from taking a picture or downloading a data file on a unique project that you might release 12 months from making a alpha prototype.  Because you can't look at it at home on your own time
you know, you might want to like make a alpha prototype that looks nice and shit to help gather investors and maybe work on software?
 

what does throughput on a digital file matter? All it has to do is look interesting enough to investigate further. When things are labeled with a silkscreen its even easier to reverse engineer. Someone can easily figure out something they are looking at might be a profitable arduino shield or such.

[james_s]

I used to use ferric chloride, but then I discovered HCl + Peroxide. It's cheaper and easier to get, works faster, it's transparent (at the start) and once it turns into cupric chloride you can keep on using it.

Funny thing is I remember finding instructions to dissolve copper wire in the mixture to produce cupric chloride, but that's just silly since you can just etch some boards. When you finish it's easy to neutralize any excess with baking soda.

[james_s]

^I wonder, now. Maybe the corrosion is due to the bottle of muriatic I keep by the tank. Any unpainted steel within a few feet is toast. I often used ferric indoors without any problem.

That is a problem. I store my bottle of HCl outside next to the garage, sealed in a plastic bag. I do my etching out on the deck after laying down a garbage bag with a sheet of newspaper over it to catch any small spills. I keep a jug of water saturated with baking soda nearby.

[KL27x]

Funny thing is I remember finding instructions to dissolve copper wire in the mixture to produce cupric chloride, but that's just silly since you can just etch some boards. When you finish it's easy to neutralize any excess with baking soda.

It's funny that the reason half the world thinks this is due to a runaway popular Instructable by "the Real Elliot," about ten years ago. Especially since he clearly stated that he made his batch of cupric by deliberately dissolving copper wire. He stated you could just use muriatic and peroxide until it eventually turned into cupric, but he had never even done this.

In fact, in many countries the cost of 30% peroxide is high enough to be a deal breaker. And with 3%, you will not make useable cupric simply by "just etching some boards."

I used to use ferric chloride, but then I discovered HCl + Peroxide. It's cheaper and easier to get, works faster, it's transparent (at the start) and once it turns into cupric chloride you can keep on using it.

So you discovered this muriatic and peroxide, and you're still using it. Why are not using cupric chloride which you have automatically made by now? Surely, you have "simply etched a few boards" by now? Are you repeating this recipe without ever having done it yourself? Just restating as fact, something someone else initially stated as fact? And with no experience anywhere to back this up as actually being a good idea?

Funniest thing about it. A year or two after the Real Elliot's Instructable, someone made a giant industrial spray etcher. And they mixed up several gallons of muriatic and 3% peroxide, citing the same thing you have just said as fact.... this several gallons of etchant with mere days of shelf life at most... to etch a COUPLE TEST BOARDS.  I'm sure he had fun trying to etch with his 0.1% cupric chloride in 10% muriatic solution the following week. (I hope he didn't burn his house down what from the O2 gas produced by all that peroxide wasting away.)

The problem here, is that peroxide is inherently unstable. You know how it bubbles when you put it on a cut? Same thing happens when you mix it up with cupric chloride. When you mix it up with the straight acid, it doesn't bubble, noticeably, but as soon as you start etching, the peroxide degrades faster. In fact, when you add even 3% peroxide to proper cupric or ferric, it will bubble vigorously... So only a fraction of what you put in there even oxidizes/recharges the cuprous to cupric. The rest just releases into the air.  After you put your used acid and peroxide etchant away, all your peroxide will be plain water by the next time you want to etch a board.

[Ian.M]

If you live somewhere where you cant easily buy *any* peroxide, its possible to get going with acid copper chloride etching without peroxide, although its a real PITA and takes a few weeks.   Even a 50ml bottle of 3% Hydrogen Peroxide will vastly speed this up if you feed the initial acid/peroxide mix as much fine copper wire as it will take.   

Without peroxide you need to oxidise the copper as much as possible - take fine copper wire and heat it with a reducing an oxidising gas flame to get a layer of black copper oxide, ball it up (over your container to catch any copper oxide that flakes off) then wet it with conc. hydrochloric acid, and leave it outside, sheltered from the weather in a wide open container^*, with a little pool of acid in the bottom, prodding it every few days to turn the ball of wire over in the pool of acid so it gets wetted again.   Once it starts to go green, it speeds up quite quickly and the copper will start dissolving overnight.  You can then feed it more copper and acid, and when it wont take any more copper without adding acid, add just enough acid to redissolve any white crystals that have precipitated out, check the density to estimate copper concentration, dilute to get it within the typical range for the etching process, then titrate for acid concentration so you can bring that into range.

* Open containers of conc. acid are *DANGEROUS* - make sure its adequately secured against kids, pets etc, cant blow over or get knocked over, and that nothing can fall into it.  

Edit: thanks Kl27x for the flame correction.

[KL27x]

As a lay person who once stayed in a Holiday Inn, I would think you want an oxidizing flame?

Anyhoo, it's very easy to make the cupric with just 3% and a bubbler in the bottom of a jar or bottle. Took me just a matter of 3-4 hours, maybe, but this is temp dependent, and I live in hell. I suspect your method would also be a bit speedier with a bubbler.

BTW, I know the Real Elliot left out a few things because I followed his advice. And I ended up with a bottle of useless greenish piss water and had to start over.

Once it starts to go green, it speeds up quite quickly and the copper will start dissolving overnight.

My experience jives. There's a trip point of concentration where the stuff actually works as etchant and warp drive kicks in. If you don't reach that within hours of mixing up your acid peroxide, you will have a hard time to reach the end point with just 3% peroxide to add in order revive your already diluted acid mix. Unless you are ok with "etching boards" by leaving them in the tank for several days and getting unusable result.

[Ian.M]

My bad - yes a blue oxidising flame is correct as you *dont* want the copper contaminated with soot and un-burnt hydrocarbons that would be produced by a yellow reducing flame.

When you are getting started without peroxide, you cant start off with enough acid to use a bubbler unless you can get copper oxide powder, as starting with oxidised fine wire, if you simply immerse it in acid, the cuprous ion concentration will be too low to dissolve copper metal at a useable rate.  Also, due to the conc. acid, the fumes are too noxious if you use a bubbler.   

Without *any* peroxide, it takes a very long time to get enough copper chloride for the reaction to 'kick off'.   Basically one is weathering the copper to get a chloride patina, then washing that down into the little acid, and repeating until the cupric ion concentration in the acid is high enough that the bottom of the ball of wire starts dissolving at a noticeable rate.  The cuprous to cupric regeneration is purely depending on diffusion of atmospheric oxygen into the acid pool, which is why it needs to be shallow and in a wide open container to get started.   

Once you've got enough to run a bubbler and have got the concentrations into the acceptable range, you can feed it copper + diluted acid quite rapidly till you've got enough for your etch tank.

I'd expect it to take weeks without peroxide to get 100ml of conc coper chloride to start off, then a few days to 'grow' that to enough litres to fill your etch tank.

Its not something one should consider if one can buy peroxide, even in small quantities and low concentration, without getting on any terror watch lists.

[sleemanj]

If you are in a REAL bind, I have in the past, out of curiosity and with no records or for that matter much though in the process, successfully etched a test PCB with using HCl, water, and... sodium percabonate (laundry whitening powder, or it might have been sold as deck cleaner or something), percabonate (aka peroxyhydrate) plus water gets you hydrogen peroxide, and sodium carbonate, the problem of course is that the sodium carbonate will work to neutralise the hydrochloric, and gas and foam and the hydrogen peroxide you're producing is also killed  pretty quickly, but, purely out of curiosity, I did get it to work once.

I think I mixed up the percabonate and water until it was saturated, then basically added that to the HCl (which may have been 33% at the time come to think of it so that might have helped).

[KL27x]

When you are getting started without peroxide, you cant start off with enough acid to use a bubbler unless you can get copper oxide powder, as starting with oxidised fine wire, if you simply immerse it in acid, the cuprous ion concentration will be too low to dissolve copper metal at a useable rate.

Yes, I totaly understand. But I figure using a bubbler will be advantageous, here. You want to make the copper "rust." And you want that rust to be dissolved in acid. The bursting bubbles at the surface will splash and wet the exposed copper, speeding up the rusting process and also dissolving any oxidized copper.   

Also, since it does take weeks, might as well put that little bit of copper in there to work, best as can. Fish tank bubbler is what I used back then. They are made to run for lots of weeks.

then basically added that to the HCl (which may have been 33%

Not like you oughtta? 

[janoc]

The problem here, is that peroxide is inherently unstable. You know how it bubbles when you put it on a cut? Same thing happens when you mix it up with cupric chloride. When you mix it up with the straight acid, it doesn't bubble, noticeably, but as soon as you start etching, the peroxide degrades faster. In fact, when you add even 3% peroxide to proper cupric or ferric, it will bubble vigorously... So only a fraction of what you put in there even oxidizes/recharges the cuprous to cupric. The rest just releases into the air.  After you put your used acid and peroxide etchant away, all your peroxide will be plain water by the next time you want to etch a board.

Yep, basically you end up only diluting your etchant with water (which is what peroxide degrades to). This type of etchant is not storable/reusable, it needs to be mixed fresh every time.

Also, a 3% peroxide is not very useful for etching, you will need a lot of it like that because it gets used up rapidly. If your only source of peroxide is the pharmacy 3% one for disinfecting wounds, that will get expensive pretty quickly. And getting a 10% or 30% one is not always easy - especially the 30% one. It is a dangerous chemical and a precursor for making "cool" stuff like TATP (the European terrorist improvised explosive of choice - e.g. the London and Brussels bombings used that) and pervitin (methamphetamine - illegal drug) so the peroxide sales tend to be closely watched/regulated in a lot of countries.

[AE7OO]

Throwaway HCl+H2O2 is more tricky to get right

In a glass bread pan, I just pour in 3% H2O2 to a level of about 2 inches, and then add 4 caps of HCl, no water.  No fumes, no fuss.

I just swish my board(I have duct tape handle on it) right side up until it is etched.  I then let it evap, then rinse with a little water into an old wine bottle next time I etch; If I ever decide to go the cupric route, I'll have a copper solution to get me started.  I do all of my etching outside and I'm in Arizona so it evaporates fast.

I buy the H2O2 at the dollar store in a big bottle, and the gallon of HCl I bought at Home Depot for about $10 ages ago.

[sleemanj]

Not like you oughtta? 

Well, I was almost certainly in bare feet, possibly without gloves and likely working in a dimly lit garage with poor ventilation and everything precariously balanced on piles of junk, so adding acid to water wouldn't make a lot of difference to the chance of disaster ;-)

[Siwastaja]

The problem here, is that peroxide is inherently unstable. You know how it bubbles when you put it on a cut? Same thing happens when you mix it up with cupric chloride. When you mix it up with the straight acid, it doesn't bubble, noticeably, but as soon as you start etching, the peroxide degrades faster. In fact, when you add even 3% peroxide to proper cupric or ferric, it will bubble vigorously... So only a fraction of what you put in there even oxidizes/recharges the cuprous to cupric. The rest just releases into the air.  After you put your used acid and peroxide etchant away, all your peroxide will be plain water by the next time you want to etch a board.

This is a good point and worth mentioning. When converting from HCl+H2O2 to cupric chloride "by etching a few boards" - which is exactly what we succcessfully did, although we also did put in some scrap copper -, you don't do it by adding the H2O2 to the massive tank. That would be hugely expensive, since you need to add fresh H2O2 every time and need a certain concentration.

So, you do it by taking a bit out of your big tank, and then proceed like you did with single use - add H2O2 to this small single-use container. Then you pour the "waste" back to the big tank.

As the time goes by, you can use less and less H2O2, and at some point, you can just start etching in the big tank. Although, you need a bubbler for effective etching speeds. Spray etching would be even better, but bubbling is easier.

We get 1 hour etcing time without bubbling, around 10-15 minutes with bubbling, in a large tank.

The bubbling also ensures the self-regeneration process to curpic chloride.

The amount of etching required to get necessary cupric concentration in our 5-10 liter tank was surprisingly low. I think we were fully converted in about a month IIRC, and as said earlier, almost zero maintenance after that, for more than 5 years now. HCl has been added maybe about five times, and water to top off the surface level (due to evaporation).

[nctnico]

Spray etcher is a bit of a luxury for a home gamer. They certainly do work well but are quite an overkill unless making a lot of boards.

They are also available in the hobbyist realm. I got mine second hand for a low price. They are not difficult to make either.

https://www.instructables.com/id/The-Etchinator-low-cost-spray-etcher/

[Psi]

I've used ammonium persulfate and it's very clean working and IMO better than ferric chloride.

I could never get that stuff to work unless it was over 80degC
It would always cool down too quick to use.
Even if you put the container of it in a bucket and poured boiling water around it, it would still cool down and stop working before you had finished etching.

I guess if you had direct heating of the solution while etching it might work.
But it doesnt work cold at all.

HCL peroxide or Ferric chloride will work at room temp, just not as quickly.
ammonium persulfate doesnt work at all.

[tautech]

And yet I've never had a problem with a fresh mix of ammonium persulfate sourced from Jaycar.

What put me off it was it couldn't be stored or rejuvenated to be used again.
I used it hot but not so hot that it would scald and with a bubbler. ~5 minute etches.

Because of its cost I built a thin upright etching tank so to use less solution however for even a medium size PCB I still needed a 500ml brew.

HCL has worked out much cheaper however sleemanj's new experimental brew looks very interesting.
Please keep us posted on further developments. 

[CopperCone]

How much better is a spray etcher then bubble etch?

I have some problem doing thin packages like TDFN-10 with etching on large PCB's (like 6x10 inches).. but I think its more of an exposure problem then a etching problem.. I need a light box, I failed to get a perfect board in 4 attempts once.. finally managed it with only a single transistor being compromised. I think the nonlinear light distribution of the desk lamp I use is the cause.

[james_s]

So you discovered this muriatic and peroxide, and you're still using it. Why are not using cupric chloride which you have automatically made by now? Surely, you have "simply etched a few boards" by now? Are you repeating this recipe without ever having done it yourself? Just restating as fact, something someone else initially stated as fact? And with no experience anywhere to back this up as actually being a good idea?

I've been using the same dark green liquid now for about 5 years. When I'm done etching, I pour it back into the storage bottle and dispose of any excess. I don't know precisely what the chemical composition is but it's dark green in the bottle and turns a very light green when I add a capful of 3% peroxide. As I etch a board it gets darker and darker green until it becomes almost brown, adding a capful of peroxide turns it light green again. It etches copper, what more could I ask of it?

So is it a good idea? Well like I said, I've been using the same stuff for several years now, I've etched many dozens of boards, haven't had any trouble with it. Why wouldn't it be a good idea?

[james_s]

I've used ammonium persulfate and it's very clean working and IMO better than ferric chloride.

I could never get that stuff to work unless it was over 80degC
It would always cool down too quick to use.
Even if you put the container of it in a bucket and poured boiling water around it, it would still cool down and stop working before you had finished etching.

I guess if you had direct heating of the solution while etching it might work.
But it doesnt work cold at all.

HCL peroxide or Ferric chloride will work at room temp, just not as quickly.
ammonium persulfate doesnt work at all.

I use a $5 aquarium heater in my bubbler tank. I've never tried persulfate but I suspect it would work fine in the same tank.

[janoc]

I've used ammonium persulfate and it's very clean working and IMO better than ferric chloride.

I could never get that stuff to work unless it was over 80degC
It would always cool down too quick to use.
Even if you put the container of it in a bucket and poured boiling water around it, it would still cool down and stop working before you had finished etching.

Persulfate works well and it is certainly less messy than the alternatives but you need a heated tank for it or at least a water bath. Normally around 40 degrees C should be enough but I have seen some vendor instruction sheets specifying up to 75 degrees.

If you had to heat it more to get it going then your solution was likely stale/decomposed - keep in mind that persulfate is not chemically stable and decomposes over time releasing oxygen. Heating it will also speed the decomposition up. So you both must not keep it in an airtight bottle (it will burst or eject the stopper) and don't keep a solution around for long (beyond a few days) because it will go bad. Ideally you should make a fresh batch every time.

This is the main reason why this is not a good/economical etchant if you only make one board every once in a while - it is really wasteful and gets expensive. OTOH, if you are etching a larger amount of boards or working for a hackerspace/club/etc. where people need to make boards all the time, this is a good option.

[innkeeper]

I'm going to attempt to make Cupric Chloride - my tank is 1.5 gallons 6 liters but id likely never fill it up completely...
though, I am going to target making 1.5 gallons of cupric chloride.

I've read quite a bit. but, still, I would like some validation on what I am about to do.
anyone who knows about this who can check my reverse engineering to see if this holds or not? i am far from benign a chemist and so i am just trying to back into Jim's and Adams notes.
Also, anyone can comment on the approach, the levels I am trying to attain or anything else, please chime in.

my target concentration will be 150g/l Cu++  that seems to be in the center rage of the recommended concentration of 125-175g/l CU++ ikn Adam Seychell's paper also the target used  by Jim at his site jimlaurwilliams.org
though I am wondering if I should target a lower concentration as over time the concentration will increase with use.

I will start with 1 gallon / 4L of Muriatic acid at a concentration of about 30%   900g copper  (6l * 150g)

I will put this into a 5-gallon bucket for the conversion process and to that, I will add hydrogen peroxide (concentration dependent on what I can get) in small doses, just enough to get a reaction but not enough to start any kind of foaming or highly vigorous action. (the lower the concentration of hydrogen peroxide the more I can add)

[WARNING: READ JIMS SITE, ADDING TO MUCH HYDROGEN PEROXIDE AT ONCE WILL LEAD TO A VIGOROUS REACTION, CREATING LOTS OF HEAT AND FOAM ETC..SO IF YOUR USING PLASTIC CONTAINERS, IT WILL BASICALLY MELTDOWN]

if my hydrogen peroxide concentration isn't high enough to finish the reaction once I reach 6l I'll add a bubbler and let it finish over time.

once the copper is dissolved, I will make up the remaining volume by adding water. to total out 6 liters.

HOW MUCH H2O2
well.. this was an interesting exercise in reverse engineering notes.

2.4 moles of h202 for 2.4 mole of coper according to jim so
basicly 1 to 1 on weight at 100% concentration
with 30% peroxide.. i need 3.33 times the weight to be equivelent
so 900g * 3.33 = 2997  = 3kg = 3l

4l muric acid plus 3l hydrogen peroxide is gonna lead to 7l too much. the specific gravity will be too low.
doing some reverse math, the minimum concentration I can use is about 40%

900 * 2.5 = 2.25 - 2.25l

which gives 6.25l and would allow for some evaporation.

So, in general, anything less than 40% hydrogen peroxide will need aeration to complete the conversion. and since I am doubtful I can get 40% hydrogen peroxide easily I will likely need to end the process with aeration unless i target a lower concentration of CU++at 125g /l i can get away with 30% hydrogen peroxide and hit my 6l goal.

comments, questions or observations are highly appreciated.

[james_s]

I would suggest experimenting with small quantities before you mix up a large batch.

My observation was that the ratios are not at all critical. The best etchant was the raw HCl + peroxide, it would etch a board very quickly without causing any pinholes, fantastic results and fairly clear for the first couple uses. After using it some it starts to pick up copper from the boards you etch and turns into cupric chloride which works well but is much slower and requires a bubbler to work well. If I want a very good quality board I'll sometimes still mix up some single use HCl + peroxide, otherwise I use the batch of cupric chloride I've been using for years.