As I've read through hundreds of messages in these forums written by people who are trying to diagnose problems or are trying to evaluate their machines, it has occured to me that it would be great to have a "CNC Obstacle Course."

Think of the obstacle course as a standard set of g-codes. There would probably be more than one of them, with increasing levels of difficulty, or perhaps designed to test different attributes of a machine. One could take one of these codes and run it on a new machine to produce a part. Obviously the "smoke test" is just to see if the machine can cut the tool paths. But a finer granularity would involve taking measurements off the part with sensitive instruments to see how close it came to "ideal".

Lathes have their test bars to check for unwanted taper, bed warping, and the like. The Obstacle Course could test for many more kinds of things. It could for example have feeds and speeds of increasing difficulty designed to be done on a standardized alloy so you could check surface finish and chatter.

One could ultimately envision running contests just for fun or at least posting the result you got after making a change. I'd love to see things like the difference in ground versus rolled ballscrew conversions on the same mill. It would be exciting to see once and for all the difference on the CNC obstacle course for an Industrial Hobbies mill out of the box versus one that followed Aaron's prescribed (and highly controversial) way lapping. Can the obstacle course uncover a difference between servos and steppers? And yes, we should be measuring not just how accurately the part is produced but also how quickly.

Does anything like this exist already? Maybe one of the posters who is so adept at suggesting ways of diagnosing and who has access to some high end CAD/CAM could create one and post it.

Just an idea!

BW