When placing by hand became unbearable, I looked for a small Pick and Place machine. This post recaps some of the rationale behind the decisions I made.

Make or buy?

This evaluation was not a well-organized, directed process but rather chaotic meandering through the options. As I gathered bits and pieces of information, my decision to make or buy took complete 180° turns many times. Not a good sign.

But finally it seemed to me that (at the time) no commercially available machine was really well suited to prototyping and small runs. These machines tried to be underbudget production machines, really and judging from all the problem reports on the internet, they mostly tried but failed. More about that, in the EEVblog post.

The Benchmark: OpenPNP

It was OpenPNP, elegant GUI, code quality, documentation, videos, tone in the community that quickly set the benchmark for all other solutions. My PCBs are complex. I need a large number of different parts, each only in small quantity. Many short strips quickly laid out on the table, even loose parts, all recognized by computer vision: OpenPNP clearly embraced the kind of prototyping I was looking for.

The commercial machines had none or very little of that. They are overloaded with (sometimes questionable) automated tape reel feeders instead. Some users successfully adapted commercial machines to OpenPNP but none of the affordable machines had the amount of free table space I needed.

Liteplacer – the middle way

As a kind of middle way I decided to buy the Open Source Liteplacer kit. It actually comes with its own Open Source software, but OpenPNP is also embraced and often used in the community:

The only concern I had was speed. My last board had 400 parts (100 different ones) and when I did some rough calculations with the P&P times observed in videos, this really added up. So I wanted to keep an eye out for optimizations. This influenced many of the decisions down the road.