I've been trying this idea on an experimental board. I purchased some MG Chemicals Silver Conductive Epoxy and have used it to connect the layers on a small, two-sided PCB design with 29 vias.
I applied the epoxy by mixing a small amount and brushing it on the drilled top copper while the bottom rested on a piece of paper. I then used a plastic paint spreader to make one smooth wipe across the board to remove the bulk of the excess epoxy. At this point you can see that the vias are filled. (There are some light streaks of epoxy left on the surface too, but this residue gets removed later.)
This epoxy can cure at room temperature, but I opted to use a hot plate to cut the curing time down to 15m. I pre-heated the plate to 65C per the manufacturer's instructions and after the 15m let the plate cool down with the board resting in place.
Next I carefully peeled away the bottom paper, being careful not to pull directly upwards on any of the filled vias. Then I removed all traces of the epoxy left on the top and bottom copper surfaces using fine grit metal finishing sandpaper stretched over a sanding block.
The board was then ready for CNC milling. (I'm not sure how this would fit into a chemical etching process. Maybe someone here has an idea?) After milling the top and bottom traces I went over each one with a multimeter to see how the process worked.
The good news is that all of my vias made connections. The not-so-good news is that the silver-laden epoxy is of course less conductive than solid copper, so all of my vias introduce small resistances into the circuit. My Radio Shack multimeter measured resistances ranging from .3 to .7 ohms, with an average trace passing through two vias end-to-end picking up .7 ohms of resistance total (so on average 3.5 ohms for each epoxy via).
My next step is to solder on the components using the hot-plate technique and see how the epoxy tolerates the heat. The specs give a maximum operating temperature of 150C so I will be trying a "low-temperature" solder paste (liquidus at 138C).
I am calling the experiment a qualified success... Probably not great for analog designs but it meets my criteria for quick, reliable and relatively low-mess / easy clean-up. The small amount of hardened waste epoxy can simply be dropped in the trash, as all it contains is silver. No chemicals baths or subsequent liquid disposal required!
--- In Homebrew_PCBs@yahoogroups.com, "ohmware" <kaelin@...> wrote:
>
> That link points to a process that's only intended for hole wall activation. You still have to electroplate afterwards to have permanent connections between the layers. What is described in the book is the use of a conductive glue that completely fills the vias (and leaves little bumps on the surface of the PCB) when it sets up / cures.
>
> The author does not offer any insight as to why a commercial manufacturer might go for this method over electroplating, as his book is just about repair. He basically points these out only to say that he's seen problems with them-- and they can't really be fixed in an electronics shop.
>
> I've no idea what glue they'd use, unless it's the silver-bearing epoxy stuff. The usual complaint about that stuff is that it's expensive
But for hand-made PCBs it's probably not out of line.
>
> I'm interested because I've been investigating electroplating, and while it seems straightforward it does involve liquids that have to be disposed of, and fumes, and I'm trying real hard to develop a complete process that I can use in a lab that sadly doesn't have a sink or good ventilation.
>
> --- In Homebrew_PCBs@yahoogroups.com, "studleylee" <indigo_red@> wrote:
> >
> > goto http://www.thinktink.com/stack/volumes/volvi/condink.htm
> >
> >
> > --- In Homebrew_PCBs@yahoogroups.com, "ohmware" <kaelin@> wrote:
> > >
> > > I just read something that intrigued me in the book "How to Diagnose and Fix Everything Electronic" by Michael Jay Geier. He talks about some newer PCBs in commercially manufactured produces using conductive glue in through-holes rather than plating. He goes on to say that he hates seeing these, as he has found them to be a s
> >
>