PCB etching frustration

[Fred 453]

I've spent a lot of today experimenting with PCB etching as it's not something I've done before. To be honest the day has been pretty much a waste of time. I've CNC milled PCBs in the past without a problem but etching has been winding me up.

 

I'm using pre-sensitised UV resist board and I've had some success. It turns out that sticking the board on top of my van under some fluorescents works better than UV exposure though.

 

However, it's the etching that killing me. I tried a Ferric Chloride and sponge technique and it hardly etched at all. I tried some Sodium Persulfate that came with a starter kit and very little there. In desperation I heated it up by putting the tray (incl. small PCB) in the microwave. Yeah - I know, but my wife's away for the weekend so she'll won't find out. The etch was very slow and hence mostly dissolved the traces too. I'm going to try a larger tray of Ferric Chloride next.

 

Anyway, I know there are lots of guides out there so I'm not really after any advice (unless you'd like to). I'm just venting my frustration. Does anyone else bother or do you all just get them made properly? Prices aren't great in the UK and I (normally) like doing stuff myself.

[spirilis 1,265]

Well I have a good bit of boards under my belt now, so I'll share my thoughts-

 

I got started by following these guys and using their kit-- http://www.pcbfx.com/

 

Their Dextrin-coated paper I'd use (process was to print the gerber on normal paper, cut a piece of dextrin paper to match the size + adequate clearance around, use some 3M painter's tape to secure it and feed that through the printer again with heavy toner settings) with my Samsung multifunction laser printer and the exact laminator they recommended.  Got the laminator for $55 off amazon at the time.  I rarely use the dextrin-coated paper now, there is a glossy color-laser paper sold at Staples (office supply store) I use that works fantastic and I don't have to bother printing + cutting the dextrin paper + taping it on + printing again, etc, I just print it on the paper and waste the rest of it since it's much cheaper.

 

I do not use UV sensitized boards, just straight FR-4 1oz stock, although their kit comes with 0.5oz stock and they highly recommend it (it does etch a lot faster).  I think somewhere on their site they mentioned commercial board fabs sometimes use thinner stock like 0.5, etch it, then put it through some kind of electroless copper process to bolster the thickness of the copper, or maybe that's involved in the creation of vias/plated-through holes but whatever.  I use 1oz and it does take ~35min to fully etch.

 

I ALWAYS include a ground plane on my copper layers so there is less to etch--and less likelihood of erasing the traces via under-cutting (acid etching the sidewalls of the traces underneath the etch-resist toner).

 

That website also mentioned the "Edinburgh Etch" -- Ferric Chloride with citric acid added.  I have a 1 gallon container of ferric chloride and a smaller pint container (from the local RadioShack); after I used up the pint, I reloaded it by adding a tablespoon of citric acid + a little water to help dissolve it, then filled the rest with ferric chloride, capped it & shook to mix.  It definitely works better.

[Fred 453]

Thanks for the tips. I did take a look at the pcbfx stuff a while ago. I don't have a laser printer and didn't really want to get one just for PCBs. It's not the cost because I'm sure you can get a cheap one on eBay but the space. I've already got an inkjet, transparencies and UV light for doing soldermask following this - www.instructables.com/id/Dry-Film-Solder-Mask/ The soldermask worked well so thought I'd use the same principle for the traces. I had heard that photoresist was generally better than toner transfer too.

 

After I posted earlier I tried a different brand of sensitized board and got much better results on the resist. I also managed to etch after I went with more Ferric Chloride and warmed it up. I went with the standard "sloshing the tray about" technique and dabbed it a bit with the sponge. It was passable on some text and numbers which I'd printed to help test exposure length. I haven't tried it on proper traces yet as I ran out of time. I'll save that for when I next have a board to etch. I'll also try the citric acid but didn't have any today.

 

I'm using Ferric Chloride hexahydrate granules and am only mixing what I need but maybe it'd be easier to mix a bit more. I just saw one post saying that you should let it sit for a couple of days, but if I only see one post on something as widely discussed as etching I tend to discount it as background chatter.

[spirilis 1,265]

Ah, right... My own technique is to use ziplock bags, 3 of them to be precise (a small thin sandwich bag for the board; it always gets poked/leaks but it's good enough) inside a larger heavy-duty ziplock which is inside another of the same size, that combination works well and it's less messy IMO; I can entomb my cleanup towels & latex gloves in the bag and throw it in the trash.  Nice thing is I don't need gloves until the board is etched, and I just sit it on my workbench and massage the bubble of acid back and forth over the board.

 

I was ready for a new printer when I bought that laser anyhow, my inkjet kept going through cartridges from them drying out when I wasn't using it... my wife has an inkjet too that seems better on that particular issue so I ditched mine and we got the laser.  Definitely understand trying not to have too much crap in your place!

[Fred 453]

As usual I'm inspired to crack the problem rather than put off by poor results. I'm going to order some citric acid, some better photoresist developer (sodium metasilicate pentahydrate rather than sodium hydroxide), some quality photoresist boards and some sodium bicarbonate to neutralise the ferric chloride before disposal. I'll definitely try your ziplock bag technique too.

[roadrunner84 466]

A technique I recently find popular is using H2O2+HCl. It's basically using copperchloride instead of ferricchloride. It's more environmentally friendly and does work fine at room temperature (ferricchloride seems to work best around 45 degrees Celsius). It also tends to stain a lot less.

[Fred 453]

I have seen a few recommendations for H2O2+HCl - and a few who say FeCl3 is better. I may give it a try later but I'll stick to one etchant and get that right for now. I only tried the Sodium Persulfate as it was included in a kit I bought.

[zeke 693]

I've etched about 640 boards in my time using this technique and the biggest thing that I was annoyed with was drilling all the holes afterwards!

 

Nowadays, I would rather have someone else deal with the fiberglass shards.

 

I would consider taking a peek at the pcb maker thread and see what you think of the various board houses not located in the UK.  ;-)

[Fred 453]

The drilling is the easy bit as far as I'm concerned. I've previously CNC milled some small boards so that can go straight into drilling and the cutting out any shape I like afterwards with perfect alignment.

 

The problem is that it's difficult to get a consistent shallow cut depth. That makes SMD tricky and the solder mask too thin over the tracks where it flows into the milled channels. There are ways I could improve my milling (height probing, etc.) but thought I'd give etching a try.

 

I've seen the PCB thread and I'm sure I'll give one a try at some point, but I just like the challenge of a home made board. Isn't that why we do most of what we do? Just to see if we can. 90% of what we make we could just buy quicker and easier but that's no fun.

[spirilis 1,265]

Yeah not having any CNC equipment, the drilling is a pain in the ass.  I'm starting to wise up and figure out bending PTH pins to make them surface mount, etc...

[demir 0]

Doing yourself isn't bad at all if it

[Fred 453]

I just finished my first etched board - after lots of demands on my time dragging me away from it. It's been a learning experience and I'm not 100% happy with the result but it looks like TSSOP and 0805 is acheivable at least. I'm sure I can do better next time.

 

This is the board after etching. I kept to single-sided so I have a couple of large vias that will accommodate jumper wires.

 

This is it with A UV cured solder mask film.

 

Next steps are drilling, cutting out and soldering. Oh - and of course letting you all give some constructive criticism of my first proper attempts at PCB design.

 

You're right that the effort probably isn't worth it - unless you're doing it for the fun of it.

[spirilis 1,265]

That soldermask is gorgeous. Good job!

 

Sent from my Galaxy Note II with Tapatalk

 

 

[yyrkoon 250]

I'd have to agree with spirilis. Those look very nice to me. I do not have any advice, but have tried toner transfer etching( twice ) and my results were far worse.

 

Good job

[TheDirty 33]

There has to be something wrong with your ferric chloride or possibly exposure.  Warm FC should etch a board in a few minutes.  I've used all the standard home etchants, and FC has given me the best etch.

 

Quite frankly solder mask is a big waste of time for me.  I just don't have much use for it.  People mention how nice it is to separate the solder, but I find soldering just fine without it.

 

No comment on the PCB layout.  Looks nice to me.  I have the (possibly bad) habit of trying to mash in everything as close together as possible and making the smallest board I can.