.50" Round Pocket yields .53" diameter

Anyone else experience an inaccurate circle pocket being cut when using the pocket operation? If so, how did you correct it?

Material: Pine, cherry, and walnut…didn’t matter, it happened to all 3 types of wood.

Using Carbide Create v7 beta with Carbide Motion v565

Steps:
a) created several .50" circles (.25" radius)
b) assigned the pocket tool path to them
c) specified .375" depth of cut to the pocket tool path (each pocket required 4 steps to compete the operation, so I don’t think I wasn’t “over-plunging”.

Results:
Since the dowel I planned on putting into the pockets seemed to be too small, I used a digital caliper to measure both the pockets and the dowel:
The pocket was .53" in diameter.
The dowel was .4980" in diameter.

I re-did the pocket operation using .47" diameter circles.
The resulting pocket was exactly .50" in diameter.
The dowel was a nice snug fit, as expected for a .50" hole.

All v-wheels were checked before the pocket operation and were all tight…no play or slop in any direction.

The router assembly (z-tram?) was also tight…no play or slop in any direction.

All belts were checked and found to be tight.

Questions:
Perhaps the #201 1/4" end mill isn’t as rigid as I expected, i.e.: can it be due to tool deflection?

That is one of the possible culprits, yes. Overview of the other ones here:
https://shapeokoenthusiasts.gitbook.io/shapeoko-cnc-a-to-z/x-y-z-calibration

1 Like

I would also check your steps per mm calibration.

This is an excellent post by @LiamN that goes into detail on calibration and squaring.

1 Like

Thanks Julien, I really appreciate your feedback…I’ll check out the article tonight.

Thanks Nick, I appreciate the feedback. Looks like I have serious homework to do tonight! :slight_smile:

Again, thanks for the suggestions and reference links.

Gleaning what I think was the best approaches from both sources (based on my skill level), I decided to create a ‘calibration board’ just 2" smaller in both directions than my machine’s usable work bed to determine if there was any x or y “run-out” across the entire table.

However, CC threw me a curve ball when it wouldn’t let me use a 1/4" end mill to cut a 1/4" wide x 28" long pocket (re: “empty tool path”)

I then measured the 1/4" end mill’s diameter and found it was actually .2480 and not the .2500 listed in the tool DB. I also measured over 6 different 1/4" end mills from several vendors and found this to be the norm across all vendors.

So, I changed the tool’s diameter in the DB to reflect the tool’s actual .2480 diameter.

CC then allowed me to use that same 1/4" end mill to cut a grid consisting of 3 parallel 1/4" wide horizontal (“x”) and three parallel 1/4" wide vertical (“y”) pockets evenly spaced across the entire board.

Result: All grid lines were perfectly true (measured to 1/128th of an inch) to their respective x or y base line reference points located on the machine’s frame, as well as parallel to each other on the board across the entire bed showing me the machine’s x and y gantrys (?) are both square to one another and to the table’s usable area and that the belts are equally “tensioned”.

Next, wanting to know how thick a pocket vector’s lines are in CC, I changed the tool’s diameter in the DB to .2499 and CC still accepted it (re: CC didn’t indicate the pocket was an empty too-path)

So it can be concluded that the usable space between pocket vector lines is reduced by .0001" (I’d say .00005/vector line, but I don’t believe the machine’s tolerance is that tight)

Lastly, creating a 1/4" diameter pocket using the 1/4" end mill listed with a .2480 diameter in the tool DB resulted in a 1/4" diameter round pocket…as expected.

Conclusion: I’m happy now that I have a better understanding regarding tool and machine tolerances and how those tolerances need to be factored together into designs when a higher level of precision is needed. Yeah, yeah, yeah…I realize this isn’t a commercial machine, but at least now my expectations are correctly aligned with what the machine can produce.

Again, thanks for everybody’s input. It helped make me think this through on several different levels that I would never have considered exploring on my own which helped me design a calibration verification test that I could understand and more importantly, conduct.

2 Likes

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.