Unusual ground layer on the surface of PCBs from the 90s?

[rfmerrill]

(Image album: https://imgur.com/a/41cD9MM)

Hi all,

So I recently realized something strange was going on with some circuit boards I was repairing from game consoles in the 90s. So far I have found this construction in at least two versions of the Sega CD/Mega CD as well as the SNS-101 ("SNES Junior").

Here's a picture of one of the boards from an early revision Mega CD (Japanese):
(Click for big)


It appears that this starts with a fairly typical 2-layer design, and then on one or both sides a ground conductor is slapped on top and coated with more solder mask (or maybe the solder mask is only applied afterward?

You can see that the dark green area of this board has no vias or pads connecting to it--it has holes in it revealing the layer below wherever vias or pads are. You can also see traces running underneath it that poke out in places. Also in the bottom left corner of the board you can see a little square depression--I think this is likely a hole in the insulating layer where it makes the electrical connection with the ground on the copper layer below.

Here are pictures from the SNS-101. On these the topmost ground layer is colored brown. In this picture you can see two conductive areas which are not connected electrically--the upper one is ground and the one beneath it is the csync signal:


And here's the edge of that same board, showing that the upper conductor appears to be somewhat different (darker color) than the one below it. Because both are ground in this case there may not be an insulating layer between:


I'm honestly really curious what process this is because I've never seen anything quite like it before. The large cutouts around vias and pads leads me to believe that this outer layer was stamped and then applied to the board, as it's clearly not drilled together with the main part of the board. The Nintendo and Sega boards might have entirely different processes as well, because they do look subtly different.

I'm hoping someone on here has seen this kind of construction before and can tell me what it's called and how it actually was done? Also, are people still doing it? It seems that certainly nowadays you would just make a four-layer board, and that was probably the case even shortly after these boards were made.

[TimNJ]

As you mentioned, there seems to be vias and outlines of traces which are beneath the ground layer, but are separate from ground.

I've did some thinking and think I ruled out any kind of electrostatic spray plating or electroless plating process. I don't see how that would work with all the normal exposed pads on the board.

Maybe it is a screen-printed (stenciled) copper "paint" layer which is reflowed (sintered) by a reflow oven. Something like this...maybe: https://www.copprint.com/category/product/

Perhaps this separate layer is tied together to the "real" PCB ground layer at one of these squares or similar locations. Perhaps there is no soldermask/insulated under the square and the copper is deposited directly on top of the real PCB ground layer.

Wild guess (because I do not actually know).

[bdunham7]

It almost looks like a masked sputtering process, followed by a topcoat with the same mask.  The copper seems to come out underneath right to the very edge of that top insulating layer.  I would think that sputtering would be too expensive and laborious for this, so maybe some similar masked process.

[T3sl4co1l]

Huh interesting, I hadn't noticed that on anything before.

Also, is that a, phenolic board I think?  Sony made a shitton of boards like that, with riveted vias I believe (hence the different appearance), and probably others that I just don't know offhand.  Almost wonder if that's a film (possibly photographic) layer on top, then, Idunno just glued on copper foil, or maybe it can be activated and electroplated on somehow (in contrast to the rest of the thing, which can't -- epoxy boards can form vias, but not phenolic, that was always the problem AFAIK).

Kinda crazy they needed that, I guess for EMC, and that it was worth it over an epoxy (or imide) 4-layer board (more traditional plane construction?), and over redesigning it for actually-good performance by itself (which, I mean, that can take up more components and area, so maybe it wasn't practical).

These days, you might do a rigid-flex board something like that (not that you would, with the regular kind being so cheap), but I don't know that that's quite the process they used back then.

Tim

[rfmerrill]

I was thinking it could be something as simple as a stamped copper sheet on top of a stamped sheet of kapton (or maybe even just relying on the solder mask for insulation?).

[ajb]

Unless it's a very thin foil, I think the way the extra layer conforms to the topography of the board structures underneath rules out the application of a stamped copper sheet.  Although maybe those are less dramatic than they look in the photos, hard to tell with the focus/depth of field issues there. 

There are techniques used on heavy copper boards to selectively plate up areas of the board, perhaps they use a similar process where a mask is applied, maybe an initial conductive activation layer, then it goes into a plating bath to grow a solid copper layer? 

As to how the technique came about, high volume consumer electronics can end up with all kinds of weird solutions that are completely alien to those of us who are used to working in more modest quantities, just because the economies of scale are so different.  Something like an electroplated copper cladding layer may make perfect sense as a less expensive alternative to a multilayer board--the materials and processing are probably dirt cheap, so you just need enough volume to absorb the NRE of setting up and initially qualifying the process.  The generous clearances around the pads and vias shows that this was not a particularly high precision process, at least in terms of alignment.  The fact that it shows up on both Nintendo and Sega PCBs argues against it being someone at one of those companies just getting creative in remediating an EMC problem or whatever, so I wonder if this was a more widely established technique for consumer electronics in those days.  Would be interesting to find out if it also shows up in things like stereo equipment of the day.  I'm sure inner layers were a lot more expensive at the time, so something like this may have made more sense when you just need a third layer for shielding purposes. 

[rfmerrill]

The fact that it shows up on both Nintendo and Sega PCBs argues against it being someone at one of those companies just getting creative in remediating an EMC problem or whatever, so I wonder if this was a more widely established technique for consumer electronics in those days.  Would be interesting to find out if it also shows up in things like stereo equipment of the day.

If I've learned anything from working on these old game consoles it's that engineering practices were very regional back then. It could have been a process that was frequently used but only by Japanese companies.

All of the discretes on these products are japanese part numbers (e.g. 2SCxxx for transistors) and they used JIS capacitor codes (JB and CH).

[TimNJ]

The fact that it shows up on both Nintendo and Sega PCBs argues against it being someone at one of those companies just getting creative in remediating an EMC problem or whatever, so I wonder if this was a more widely established technique for consumer electronics in those days.  Would be interesting to find out if it also shows up in things like stereo equipment of the day.

If I've learned anything from working on these old game consoles it's that engineering practices were very regional back then. It could have been a process that was frequently used but only by Japanese companies.

All of the discretes on these products are japanese part numbers (e.g. 2SCxxx for transistors) and they used JIS capacitor codes (JB and CH).

I don't know if this is a over-generalization, but I have noticed some Japanese electronics (to this day) are quite a bit different in comparison to what others might think is "normal". By "normal" I mean...electronics design mentalities in the West and probably even China. etc. (And I know just because it's the West, doesn't make it "normal"...You know what I mean.)

I work with switch-mode power supplies and have noticed Japanese manufacturers like Murata and Cosel often make competing products which are noticeably different compared to what everyone else is making. I think it's quite interesting. Kind of reminds me of the different flavor of electronics and electronics components that came out of the USSR.

First attachment: Cosel (green PCB) and competing Delta product (black PCB)
Second attachment: Very interesting mechanical design of the Cosel.

[rfmerrill]

There are techniques used on heavy copper boards to selectively plate up areas of the board, perhaps they use a similar process where a mask is applied, maybe an initial conductive activation layer, then it goes into a plating bath to grow a solid copper layer? 

So this sounds almost like it could be the method, but then the question that comes to mind is wouldn't that plating grow the electrically connected pads on the rest of the board? I.e. you'd expect to see the ground pads of components looking a little different from other pads.

[T3sl4co1l]

Note that tin plating is normally used as an etch stop; maybe something similar works as plating stop (without having to completely mask an otherwise finished board).  I forget if there's any... compatibility issue with plating copper over any metals simultaneously, kind of thing that would be relevant here.

Tim

[rfmerrill]

There's a high res board scan of the SNN-CPU-01 here, with the depopulated board and then the traces underneath after sanding off the solder mask and ground layer. https://bitbuilt.net/forums/index.php?threads/snes-jr-snn-cpu-01-scans.3488/

This board appears to have that ground layer applied only to the bottom side instead of both sides like the Sega CD board.

[ajb]

There are techniques used on heavy copper boards to selectively plate up areas of the board, perhaps they use a similar process where a mask is applied, maybe an initial conductive activation layer, then it goes into a plating bath to grow a solid copper layer? 

So this sounds almost like it could be the method, but then the question that comes to mind is wouldn't that plating grow the electrically connected pads on the rest of the board? I.e. you'd expect to see the ground pads of components looking a little different from other pads.

I've seen PCBs with selective heavy copper on only some parts of a net, so there has to be some way of achieving that.  Random example photo attached.  But I haven't seen much information about those processes when I've looked just out of curiosity before.  Maybe it's enough to apply an insulating mask that can be removed later?  Maybe it's even the same layer that insulates that bonus copper layer from the conventional outer layer, then after plating they strip that layer where it's not covered by copper and then apply the final solder mask?