Single side THT pads and soldermask

I am trying to make single sided THT pads (on B.Cu), but I’m having difficulties getting the correct solder mask expansion for the top side.

If I select both B.Cu and F.Cu, and both B.Mask and F.Mask, the soldermask is expanded for both side pads (as expected).

If i select only B.Cu and B.Mask, the bottom pad has the correct soldermask expansion, but on the front there is soldermask covering the hole (expected, as there is no soldermask expansion defined for the front layer). Didn’t check the Gerbers though, but I see no reason for them to be different as no F.Mask is defined.

If I select B.Cu and both B.Mask and F.Mask, the bottom pad is still correct as expected, but the soldermask expansion for the front is the same size as on the bottom layer (i.e. the size of the B.Cu pad).

This is where things are unexpected for me, and perhaps undefined for KiCad. I would like the front solder mask expansion to be the size of the drill hole and the bottom solder mask expansion to be the size of the bottom copper pad.

Can someone verify this behaviour, and perhaps suggest a viable solution?

Edit: I expected to be able to draw a filled circle / polygon on the F.Mask layer the size of my drill hole (which works), but there are no tools for a proper geometric circle, and I can’t be bothered trying to approximate a circle with line segments on a sub-mil grid to try and make it look nice.

On mobile atm so I can’t upload any pictures.

People mostly don’t use one sided boards anymore, so your use case is rare wich may explain why nobody has cared about it before. On the other hand KiCad can define only one size for pads for all layers (in v5) anyways, so this is expected.

If you need different sizes you have to create more pads and put them on top of each other. For the mask layer you would create an aperture pad with only the desired mask layer selected. Don’t add a pad number to an aperture pad.

I wouldn’t worry about it, pretty sure your fab will add those apertures themselves if needed.

That seems to do the trick - thanks! Never quite noticed the aperture style pad before, I’ll have to dig into the docs to figure out the difference with smd pads.

And yes I know my use case is rather esoteric, but it’s for a faithful reproduction of a 1979 pcb :slight_smile:

1 Like

Maybe yes, but explicit is better than implicit :wink:

It’s just a special type for non-copper pads, some properties are disabled (like net and pad number). You can get the same effect with a normal pad - just set copper layers to None and set the properties accordingly - but “Aperture” makes it easier to not make mistakes with the properties.

Surely there is no such thing as single sided THT, plating the THT pad requires an electrical connection on both sides.
To recreate an old board, I would use a conventional pad and just not track on the top side. The THT pad soldering is much stronger and the pad does not lift so easily.

pretty sure that’s not true.

besides, you could just have NPTH and copper on the underside, which is what lots of old boards were.

Exactly, they just etched the board and then drilled it. There was no plating stage.
A plated board starts off with thinner copper layer (1/2oz for 1oz finished), the board is etched and drilled. Then it is plated, creating vias and the PTH barrels, while also thickening the surface copper to the final value. If there is no copper on the top side, the PTH barrel will be badly controlled.

My early precision hand pcb designs were drilled first which left a tiny copper ridge around the hole.

After that with tools like this, and that tiny copper ridge as a shape guide allowed to draw perfect circular pads every time. Follwed by drawing the actual tracks. A thoughtfully laid out design did not need vias at all.

There existed some special (acid resistend) transparent reddish varnish like fluid. The viscosity in combination with the surface tension of the tube was just right to flow freely through an about 0.8 mm tiny opening. The fluid also did not run off of its drawn shape due to the steel tube ‘engraving’ an ever so very slight rim along its path.

The resulting pcb quality was amazingly precise. Actually way better than some of today’s cnc rough grinds.

And that (I understand one side) PCB had soldermasks at bottom and top?

1 Like

There is copying an old pcb to replace a damaged board, close enough to work the same. I have no problems with this.
And there is copying the old board to the point that you are making a fake, which I feel goes too far, not easy though as modern materials look different

Some collectors, restorers, and museum-like organizations are quite demanding about replicating an original down to the most minor detail. I once met a man who worked for a company that made replacement parts for Model-T automobiles, working as closely as possible to original Ford Motor Company drawings. As I recall, he had put a lot of effort into finding vendors for fabric-insulated wires (cotton, silk, etc) rather than the plastic insulation we use today.

Dale

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.