Insights

Despite preheating, the panel delaminated. And because the PCBs were still connected to each other, we could no longer process the remaining ones separately.

On Thursday I am at FUSE in Enschede. Plenty of startups, plenty of ideas and probably plenty of MVPs that are not quite finished. With hardware that can go faster.

Sample 2 is usually exactly the reason you built sample 1. On why "get it right the first time" is rarely the fastest route in hardware.

With a smoothing broach from the jewellery trade we brought the holes in a PCB neatly to size, without damaging the plating. So the prototype PCBs could simply be used.

The answer is often not in the technology. Almost any EMS company can assemble a good board. The real differences come from the way you work together.

Don't aim for a perfect first version, but learn from the first version as fast as possible. A post by Bart van Arnhem (Orbisk) got me thinking about how electronics is making the same leap as software.

Some people see a technical problem and think: uh-oh. Jochem thinks: fun! Meet the new lab manager of our PCBAProtoLab at Kennispark Twente.

The real question isn't how to save €1000 on a protoPCBA, but what it costs if you can only test a week later. Why we look at the Total Cost of Development.

Misplacements out of nowhere — no software, no nozzle, no component changed. How two shiny stainless-steel bolts and bright summer light confused Buddy.

Sometimes improving a process isn't about a more complex machine, but about a simpler and more controllable one. How we get certain prototypes spotlessly clean.

A prototype isn't a phase but an umbrella term. Which risk are you trying to remove in which TRL stage? An overview of the types of prototypes in electronics and PCBA development.

A small series with a low yield and a request to x-ray it. Why a second opinion sometimes confirms your first hunch — and sometimes proves an engineer's gut feeling right.

A prototype has only one goal: to learn as fast as possible. Why a parallel track saves hardware teams weeks.

A single proto says "it works". A proto series says "it keeps working". Why we expanded our process with extra pick & place capacity.