Etching PCBs.

[cheesehead]

I need to etch myself some PCB's tomorrow. I've got ferric chloride, drills, several square feet of motherboard-grade copper-clad, and all sorts of other goodies in my Art Metal class.
So, anyone got any tips? Exporting the file into something I can use from EaglePCB would be great, if nothing else.

 

[jmcoreymv]

I don't remember how to manually etch boards, but 4pcb.com has specials where you can get a double sided board with silkscreen and mask for $33 (up to 60 sq inches). Not sure if that would help you or not.

 

[Mark R]

Heh. Never bothered manually etching stuff - I just send the files to olimex.com. (At $33 for a single double sided board with silkscreen and 2 masks, I'm not going to bother looking for anyone else - and as they take eagle files directly, I don't have to screw around exporting drills and crap).

How are you going to get the design onto the boards? UV photoresist, toner transfer, manual drawing with a etch-resist pen? That, from what I've gathered, is the most difficult and most important bit to get right.

 

[WildHorse]

I've never done that, but I do know that there are etching chemicals that display directional preference.

They preferentially etch in the downward direction, and tend to not etch sideways.

Thus with the correct chemicals / temperatures, times, etc. your resulting etched traceways are "square" bottomed, and the canyon walls do not spread out to undercut the surface.

IF you're near any companies that do this, it'd be worth contacting them & trying to speak with a manufacturing engineer there, who may help you with good suggestions. Or maybe chemistry professor at local college could help you with info?

 

[Jeff7]

I've used ExpressPCB's standard service before. Very nice quality boards.

Their software is also quite nice, and is free. Unfortunately, it uses a proprietary format. Still good though for trying out trace patterns though.

ExpressPCB is also making lead-free products available now.

 

[patentman]

I've manually etched a lot of PCB's, here are a couple of tips:

1. If you are trying to get very straight edges (e.g. a straight line), make sure the master you create for developing your photomask is as dark as possible and try to "adhere" it as closely to the photomask during the exposure step. You don't want any light leaking in around the edges of the lines of the master during the exposure step.

2. Your working with Ferric Chloride, which means you will be working entirely under yellow light. Ferric Chloride itself is yellow, and it tends to get everywhere. Wear a labcoat and clothes you don;t mind ruining, because while you'll look clean under yellow light, you'll come out looking like you and your brother just had a food fight where the main even was French Yellow mustard.

3. Timing is crucial. Make sure you know how long you photomask is supposed to be exposed and cured, and adhere to those times! Same goes for the etching and rinse steps. Overetching and/or over exposure -> crappy results. If you are using a U.V. cured photomask, note that exposure time depends on how efficient you light source is, which in turn depends on a. the age of your source, and b. the make and model of the bulb you are usuing. If your light source is not uniform or is too weak, you will have a tough time getting the photmask to cure uniformly, which guarantees crappy results.

4. Make sure the ferric chloride you are using is reasonably fresh. As it etches away uncured photomask and copper from the PCB, it will become less effective and etching time will increase.

5. When you submerge your PCB in the etchant, depending on the result you want, you should either "plunge" it in the ferric chloride or dip it slowly. The "plunge" technique can be a bit messy, but it allows the board to be etched more uniformly then the slow dip (which results in graded etch over the length of the board).

Hope this helps. Let me know if you have questions. As I said before, I've done a LOT of manual etching, although not for art. I was mainly etching stuff for antennas and RCSR (radar cross section reduction).

 

[Markbnj]

Originally posted by: scott
I've never done that, but I do know that there are etching chemicals that display directional preference.

They preferentially etch in the downward direction, and tend to not etch sideways.

Thus with the correct chemicals / temperatures, times, etc. your resulting etched traceways are "square" bottomed, and the canyon walls do not spread out to undercut the surface.

IF you're near any companies that do this, it'd be worth contacting them & trying to speak with a manufacturing engineer there, who may help you with good suggestions. Or maybe chemistry professor at local college could help you with info?

I used to sell etching/stripping equipment to the PCB industry (circa 1988 - 1992 or so). The directionality of the etching actually has little to do with the chemistry, and everything to do with the spray system and pressure controls.

A word of warning if you are planning to etch boards at home: the byproduct chemistry is highly toxic and you may have a hard time getting rid of it. I would use a contract etcher as others have mentioned in the thread.

 

[patentman]

Originally posted by: Markbnj
A word of warning if you are planning to etch boards at home: the byproduct chemistry is highly toxic and you may have a hard time getting rid of it. I would use a contract etcher as others have mentioned in the thread.

Highly toxic? Really? The typical byproduct of a ferric chloride etch of a copper pcb would be a solution of Ferric chloride, Cuprous and Cupric chloride, water, and dissolved photoresist. While you wouldn't want to drink it, I wouldn't call it "highly" toxic. Heck, there is a lot more toxic stuff in most homes, e.g. bleach, windex, any bathrrom or over cleaner etc. Heck, even shampoo is toxic if you drink it.

As for getting rid of the byproduct, I agree, it can be tough. You need to to use an approved chemical disposal service. While you could always pour the crap outside in you backyard, that is: a) environmentally irresponsible; and b) most likely against the law.

 

[Markbnj]

Yeah, highly toxic was the wrong word. What I really meant was: highly difficult to get rid of. It's got a crapload of copper in it, and you can't dispose of it anywhere, really.

Not especially poisonous though.

--Mark

 

[Googer]

Originally posted by: patentman

Originally posted by: Markbnj
A word of warning if you are planning to etch boards at home: the byproduct chemistry is highly toxic and you may have a hard time getting rid of it. I would use a contract etcher as others have mentioned in the thread.

Highly toxic? Really? The typical byproduct of a ferric chloride etch of a copper pcb would be a solution of Ferric chloride, Cuprous and Cupric chloride, water, and dissolved photoresist. While you wouldn't want to drink it, I wouldn't call it "highly" toxic. Heck, there is a lot more toxic stuff in most homes, e.g. bleach, windex, any bathrrom or over cleaner etc. Heck, even shampoo is toxic if you drink it.

As for getting rid of the byproduct, I agree, it can be tough. You need to to use an approved chemical disposal service. While you could always pour the crap outside in you backyard, that is: a) environmentally irresponsible; and b) most likely against the law.

Windex is just Alcohol, Water, and Blue Tint but sometimes has Ammonia.

 

[patentman]

Originally posted by: Googer

Originally posted by: patentman

Originally posted by: Markbnj
A word of warning if you are planning to etch boards at home: the byproduct chemistry is highly toxic and you may have a hard time getting rid of it. I would use a contract etcher as others have mentioned in the thread.

Highly toxic? Really? The typical byproduct of a ferric chloride etch of a copper pcb would be a solution of Ferric chloride, Cuprous and Cupric chloride, water, and dissolved photoresist. While you wouldn't want to drink it, I wouldn't call it "highly" toxic. Heck, there is a lot more toxic stuff in most homes, e.g. bleach, windex, any bathrrom or over cleaner etc. Heck, even shampoo is toxic if you drink it.

As for getting rid of the byproduct, I agree, it can be tough. You need to to use an approved chemical disposal service. While you could always pour the crap outside in you backyard, that is: a) environmentally irresponsible; and b) most likely against the law.

Windex is just Alcohol, Water, and Blue Tint but sometimes has Ammonia.

I know, the alcohol in windex, however, is not pure ethanol (it contains some methanol, which can cause blindness) and ammonia is toxic if you drink it.

 

[Markbnj]

I don't think you'd want to drink cuprous chloride or ferrous chloride either 🙂.

 

[patentman]

Originally posted by: Markbnj
I don't think you'd want to drink cuprous chloride or ferrous chloride either 🙂.

I agree (and I did say this in my prior post), its just that the stuff is not like cyanide or something where by the time you actually smell it your already pretty much dead 🙂

 

[wwswimming]

in the mid '80s i worked for a company named Wiltron, since sold to Anritsu.

started by a very smart engineer & businessman named Bill Jarvis. with another Bill,
Bill Oldfield, in the lab.
http://www.mwrf.com/legends/Index.cfm?fuseaction=Vote

i had the privilege of watching Bill SHRED schedules. he was the kind of engineer
who did everything. if he didn't like the answer he got for doing some drafting ( i was
the mechanical design supervisor ), he would do it himself. during a 2 year period
i watched him out-machine the machinists, etc.

one day Bill needed to make a circuit board for a microwave switch. he put some
tape down on some duroid circuit board material, and cut it with a knife. then he
etched it in some kind of acid on top of one of his workbenches. he added
connectors to it and did some microwave measurements. he had printed out
the predicted performance on some clear acetate, using a scale of maybe 1/2 inch
= 10 dB, for example. he pasted the predicted performance on top of the tube
output showing the measured performance. 98%+ correlation. another schedule
shredded.

and, to cut the tape, which he was using as a photoresist, he used - an X-acto
knife.

another time i had some circuits for my own self, i was trying to build an
oscilloscope from scratch, about 1993. it was a single sided board. i found
a shop in Sil. Valley named "Brothers International" which was a really cool
place. i walked in there with a 12 pack of Henry Weinhard's & $100 and
begged them to etch the circuit for me. They did !

anyway ... just out of curiosity - can you post the artwork ? i'm not going
to steal it. i'm just curious, you know, what size traces & spaces are we
talking ?

 

[blahblah99]

Depends on what equipment you have access to, and whether you are doing double sided or single side.

If you have access to a uv light system, I'd recommend printing the gerber files on transparency on a laser printer (darkest setting) and exposing the pcb to the uv light.

If you don't have access to it and have to do it the old fashion way, you can get the press-n-peel toner transfer sheets where you can print on those sheets and iron it on the pcb.

Both methods require you to etch afterwards, which you will have to agitate manually or using a bubble machine.


OR you can do it the easy way and get a few prototypes done by Chinese companies (search on google, there are plenty) for under $100.

 

[f95toli]

You can also print on plain paper (instead of transparancy), in my experience it works just as well and if you are using a laser printer the resolution is slightly better.

 

[cheesehead]

Sorry for not getting back to the thread sooner!
The PCB's a prototype for a small amplifier I'm building. It's called a gainclone; it's a knockoff of an amplifier called the "Gaincard" which retails for about 3000$. I can build one much more cheaply.

The PCB's something of a prototype; I'm going to try toner transfer first. If it works, Olimex's prices are quite reasonable; however, I'm not about to spend 30$ on something that won't work.

I've worked with Ferric chloride before; I use a tiny amount in a sealed frosting container, as the PCB's only about 1.2 x 1.4". (This keeps down the cost to me; I can get them for under 5$ each from Olimex!) I don't have any press-and-peel, but I've heard that some glossy paper works well; tomorrow, I'm going to try transferring it over.

Hmm....
If only I could feed this stuff through a laser printer!