Second test recording using the indicator mount. Still some vibration but it's less. You can definitely see a lot better when I stop the spindle, but swarf usually blocks the view a bit during the cut anyway so I think this is probably good enough for livestreaming purposes? I'll be switching to the other microscope for high mag imaging and soldering etc anyway.

I need to move the mic even closer to my mouth or make a conscious effort to speak up, a lot, to be intelligible over the background noise.

This is also postprocessed in Audacity, the raw recording has more noise but might be a bit clearer actually? I'll need to experiment a bit.

https://youtu.be/4dWEbmjVN7g

Top down view of this test cut seen on the Leica. This is the ground plane one layer below the QFN thermal pad. I'm not sure why the board designer put five big vias in the pad for heatsinking and then added reliefs to the plane, but... you'd have to ask them why the plane was on the wrong net first (one of the reasons this board was scrapped years ago lol).

But hey, good practice.

@gsuberland I can't remember what EDA they used, this was from a contract like a decade ago.

I told the prospective client I didn't have time to do the board, but offered to advise their engineering folks and do some bringup testing since I had expertise in the specific application domain.

Big mistake.

I gave them my signoff checklist, the other contractor they brought in didn't do a full signoff review despite my urging.

When I got boards the first sign of trouble was one of the SOICs being the wrong size and not fitting on the footprint. It got worse from there, while I did salvage a working prototype I probably spent more time debugging and reworking than if I had just done the design myself to begin with.