A couple of NOOB questions on Headers

1. Yes, PTH (plated through-holes) are frequently seen on production PCBs because they provide electrical conductivity between layers. NPTH (non-plated through-holes) are also seen, typically for mechanical-only connections to the board.

2. I’m not an Eagle user so I can’t speak to their default routing rules specifically, but I tend to see traces being put on the same side as the components. Yes, this works for through-holes components, such as pin headers, because of the PTHs. However, although a trace may be routed by default to the top layer, you always have final say on which layer your trace(s) appear on.

3. If you don’t want to use rivets, yes, you’ll have to modify your default routing settings to route traces on the bottom layer and manually change the layer of existing traces.

4. Is there a particular reason you’re reluctant to use rivets? Rivets are the standard solution for routing 2-layer boards with the V-One and I suspect that most people making 2-layer designs are using them. If someone chooses not to, it would be entirely reasonable to expect them to move traces to the bottom layer if they want to minimize PCB layout changes between prototyping and production.

Zooming out to the larger question you’re asking – and perhaps the Voltera team will chime in with their official view on this – a Voltera-printed PCB, with or without rivets, is still going to differ from a factory-manufactured board in a number of significant ways. I view a V-One PCB as a prototype of design intent that allows for fast iteration, but I would never feel comfortable sending the final revision of that prototype straight to ramp production without at least one intermediate EVT/DVT spin. Any time you’re changing production methods, you need to be prepared to change your layout to account for the difference in tooling and process.

@evanevery I totally understand your reluctance to manually add 64 rivets to your board. I do want to say that after you get the hang of riveting, it is actually quite a fast and low-effort process. 64 of the large size would take you maybe 10-15 minutes from start to finish, with the bulk of that time being spent inserting the rivets into their holes prior to riveting and then doing continuity tests (& possible re-riveting or soldering to fix opens) afterwards.

Beyond simpler PCB routing, the real advantage you get by taking the time to add rivets is dramatically increased mechanical strength and reliability. I’m not sure what the use case is for your headers, but if you plan on inserting & removing anything more than one single-pin connector at a time (i.e., jumper wires), you are going to be putting a nontrivial vertical load onto your header. Without rivets, the only mechanical connection your header has to your board is through the solder joint between the pin and a PCB trace (if one is routed to that particular pin). That is an incredibly weak connection and it doesn’t take much force to break the trace, which is fairly brittle when baked. You are welcome to experiment with rivetless headers for your application, but unless you are quite careful and use low-pin-count cable assemblies that allow you to minimize insertion & extraction force, I suspect you’ll start seeing continuity failures after using your board for a little while.

You could also try a compromise: Add rivets to holes strategically. E.g., one or two on each corner and one every other pin, or only on those pins that have traces. So you don’t have to do all 64 but you’re still getting some increase in mechanical reliability.

Edit: I just re-read your original post and saw that your use case is to attach this board to another board. I would strongly urge you to use rivets on every pin in that case. That is a huge insertion/extraction load if all 64 pins are involved simultaneously.