Depth Variations: Is It the Code?

Inconsistent machining is probably a mechanical issue, not a code or controller fault. December 10, 2007

Question
I have encountered a problem processing shapes to G-code using Microvellum. The products in the Microvellum library process and cut fine, but when I draw a shape using AutoCAD and process the code, using Microvellum's post processor, the cut is deeper into the spoilboard, although the code is the same. Any ideas?

Forum Responses
(CAD Forum)
From contributor S:
Open G-code file in the notepad, click edit ->find, type in Z- and hit enter. It will show you all negative Zs. Click Edit ->Replace, type your negative Z-s in find what box and type Z.0s in replace with. What CNC router are you using? How you probe?



From the original questioner:
We use a Weeke BHP20, WoodWop and AutoCAD/Microvellum, and we run a test program to set tool heights. The code generated is showing the Z at the right position. Ideally we cut .2mm into the spoilboard. The code, in both situations, reflects this (Z-0.2), but the actual depth of cut from one to the other is different, a lot different. Could it be an offset issue? Could it be that programs created solely with Microvellum use different position offsets (G54's) than a program created just using Microvellum's post-processor?


From contributor D:
I do know that there are two places that Microvellum checks the router clearance. One is in the globals under "machining general settings/general machining/router clearance." The other option is in the tool file; under the settings tab there's "spoil board penetration." Make sure those two settings are set the same.

As far as machining calibration, do you mean you use the BHP-200? Make sure into the soft keys that you have the proper selection for "spoilboard or pods."

I have found that Microvellum is usually not the culprit when it comes to depth errors or panel sizes slightly off. Most of our problems stemmed from the router calibrations.



From contributor J:
I agree with contributor D. If the code is the same, then you have a controller (not very likely) or a spoilboard flatness issue. This is a hard number business, so we need to know how much variation is "a lot." I have had up to 40 thou variation creep up at times with moisture, heat, thickness of spoilboard, etc. all being possibilities. I am not a Microvellum user, but the code tells the machining story.