modified: Saturday 18 June 2016
Experimental homemade aluminium and silicone PCBs
Why do I buy copper clad board to make my PCBs out of? Why can't I make my own out of what I have at hand?
It needs to:
- have a conductive layer
- be etchable
- survive soldering temperatures.
Ordinary aluminium foil and silicone caulk seem to fit the bill. As we'll find out later, things were not as simple as this (and the above list is missing several important items).
Can I etch aluminium foil?
I had a hunch that aluminium would be easier to etch than copper. Later on I find this to be true: I'm a lazy person that does my experiments back to front.
Can I print onto aluminium foil?
For my copper PCB etching I normally use a toner-transfer technique. With the right paper and preparation (namely cleaning the copper) it works wonders. But can I cut this whole step out and print directly onto aluminium foil?
I tried taping some aluminium foil to an ordinary piece of paper and printing on that:
Nope. The foil scrunches up. If my laser printer was a flat-bed design then this might work, but the paper and the aluminium foil expand and contact differently as they get curled around the various rollers in the machine.
I didn't try running plain aluminium foil directly into the printer. I was afraid it may distribute itself equally throughout the printer after getting jammed and shredded.
It's interesting to note that the aluminium foil caused no issues with the electrostatic toner-deposition system. I half suspected that I'd hear arcs or have a failed transfer, but everything seems to work well. Perhaps printer manufacturers test their printers with conductive or damp paper to make sure it does not cause any issues.
Can I toner transfer to aluminium foil?
This is what I traditionally do to get toner onto copper. It's a shame I didn't find a way to cut out this step: being able to print directly onto the final conductor would be amazing.
I print onto specially bought 'toner transfer paper' off eBay. It feels a lot like the waxy material that you peel off stickers. Infact I used to use the latter, along with IKEA catalog paper, but the specially bought paper is cheap and works better than anything else I have tried.
This is not how you should use your toner transfer paper:
When doing something abnormal 6 copies of the circuit is not enough. This is just a waste of the piece of paper. Later I learned my lesson:
Also make sure to mirror your circuit before printing it. I always lose a few sheets to getting this stuff wrong, and it's not as if paper grows from trees. It grows from eBay.
My first attempts at ironing the toner onto the aluminium did not go so well:
The curl due to cooling contraction is perfectly fine, but it's the creases that cause me issues. Aluminium foil is finnicky and you have to be very careful when ironing onto it.
After some practice I was able to make good transfers:
I used an old, cheap tube of clear silicone caulk from Bunnings:
It took a very long time to cure properly: days instead of hours. I think it was a bit too old.
Anyway, I tried applying it and smearing it around a few different ways:
No matter what method I tried I was unable to avoid air-bubbles from forming as I spread or applied the caulk:
I was concerned that the etchant might break through the aluminium and fill these voids, then etch random parts of the aluminium from behind.
Once set the final aluminium foil and caulk sandwich was very easy to cut using only a pair of scissors:
These pieces are what I continue with later in this post.
Caulk as an adhesive and insulator
I tried experimenting with adhering the aluminium foil to foamcore board using the silicone caulk:
Unfortunately the silicone didn't properly set after a couple of days, so the aluminium foil just peeled off. I'm not sure if this was mainly due to the bad silicone or lack of access to air.
I'll need to try this again another day, but with different silicone. What will be an interesting test is to see if the silicone is a good enough of an insulator to stop the foamboard melting when I try soldering some components.
Final steps and results
I had a mixture of specimens at this point: some of them had good toner transfers, whilst others were appalling:
... but when it came to etching, I was gobstruck:
Can you see my fingers through the board? It worked. It's beautiful. I should have taken a photo immediately after stripping the toner off too.
At this point I met a few more problems:
- Aluminium is a PITA to solder to
- The board is flexible but the aluminium is weak.
- Drilling into the board twists and catches the aluminium rather than cutting through it.
Have a look at what this beautiful board turned into after a bit of soldering, bending and drilling:
If you even gently bent the rubbery PCB then the aluminium would crease and shear in multiple locations. It turns into crackled and crinkly skin with broken traces. Next time I need to add a stiff layer to the sandwich.
Drilling component holes was a disaster, but I found a really good way to get around this problem. Many of the pads self-destructed when I drilled them (shown above). Perhaps a better caulk would provide better adhesion and a stiffer base. Instead of drilling I realised that I could just use a pin and poke holes through the soft membrane of the board. Overall this is actually a great method: it's much simpler and easier to poke a board with a pin than it is to bugger around with lots of little drill bits that break, and you don't have to worry if the hole size is different to component leg size.
Let's not talk about the soldering.
Aluminium very quickly obtains a mostly impenetrable layer of oxide on its surface when exposed to normal air. When heated to above 250C it does so even more quickly. No amount of abrasion can fix this problem.
I managed to get three good joints soldered (shown above) after lots of effort by soldering under large puddles of flux. This process completely ruined my soldering iron tip: I suspect that the aluminium alloyed itself into my tip, so now it looks heavily oxidised and can only stick to solder in few select corners. I've had this happen before when soldering aluminium. NB I use cheap eBay 900-M tips.
Even an aluminium tip is not a problem if I can somehow stop oxygen getting to the soldering region. I did some quick Wikipedia research and apparently ghee (clarified butter) has a decently high boiling point. To the fridge!
Nope. It boiled away and did not help with sticking the solder to the aluminium at all (see the balls of solder in the last picture). Adding vinegar (a mild acid) didn't help either, but it did emit some really powerful vapours that would get right up your nose.
Perhaps ghee and vinegar are a new flavour for chips. Mmmm.
This'll need more work another day!
I was amazed by the success I had up until the point of soldering. Hopefully I can find ways to work around the last few problems.
One idea I have is to electroplate the aluminium with copper using the solution that's in my old etchant jar. I'll have to do this before etching the copper, as I need to electrically connect to the whole surface. My etchant is relatively mild, so I'll probably be able to get away with just dipping the whole board/sheet in with some electrodes.
A few people have mentioned to me that I can buy copper foil. Unfortunately I can also just buy copper clad board. I think this defeats my motivation: it's much more interesting to make a copper clad PCB substitute using ingredients you can get from most supermarkets and hardwares in the world (or just find lying around home).