Slow motion look at chip formation on a restored CNC mill

Nov 10, 2026 Last reply: 1 day ago 4 Replies
Jump to replies (4)

When everything is dialed in, this rig films chip formation as a 3/4 inch cutter bites into metal, captured at 1,500 and 6,000 frames per second. The host machine is a Tree Journeyman 325, and the camera can be fired from an external input, which suits destructive tests. Better still, the camera itself was designed and built by the maker.



Getting there took real work. The mill was walked across the workshop floor on pry bars and floor jacks, then set down on a rolling platform the maker assembled from painted steel base sections. The Olsun Corporation transformer got a new home on a bracket under the cabinet.



The restoration details stand out. A cooling fan was found mounted backwards, its rotor and blades fallen off after the retaining snap ring slipped down the shaft. The maker lubricated the sleeve bearing and reinstalled the ring, and sorted a green ground wire whose insulation had been chewed by a fan blade. A new pneumatic filter, regulator and lubricator went in, and the DynaPath System 20 CRT booted and cleared its diagnostic checks.



One fix worth flagging: spindle speed was drifting by 200 RPM, and jumper settings plus potentiometer procedures took care of it. Footage comes out at 1920x1080. The open question is storage. Whether any device can write 1.2 terabytes per second is still unknown, and that seems to be the wall standing in the way of pushing the frame rate further.



Project by tesla500. Watch it on their YouTube channel:

formatting link


That spindle drift bugs me, 200 RPM is a lot on a cutter this size. Did the maker trace it to the drive electronics, or was it something mechanical like the feedback? Curious what the actual correction process looked like on this machine.

The handling of the move is the clever part in my eyes. Walking the mill across the floor with pry bars and floor jacks, then parking it on a rolling platform, means the machine can be repositioned again later without repeating that fight. The painted steel base sections look like they keep it steady, too.

If the maker ever wants finer detail inside the chip itself, a 60,000fps system would be the obvious step up from 6,000. The catch is data. At 1080p the storage burden gets massive fast, and writing 1.2 terabytes per second is beyond any device I could name.

The drift was sorted by adjusting jumper settings and working through potentiometer procedures on the drive. As context, the setup also uses a Fanuc spindle drive manual, and the machine had been showing a 200 RPM wander before the adjustment brought it back in line.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required