Using an X-Carve Pro 4x4. Is anyone else lost when it comes to getting the spindle trammed? I can’t get a . 001 feeler gauge under every mating surface on the y axis between the gantry and bearings and this thing still leaves visible lines that you can feel. 10k for a machine and it’s not what I would call precise in this regard. Any help would be appreciated.
Just because its put together tight, doesn’t mean that its trammed in.
You need to do that with a dial indicator on a bar in the collet.
Google “machine tramming” and you will find what you need.
this is the process i use to verify tram is in or out of spec:
Surface the wasteboard (or a decent sized piece of MDF that’s securely mounted)
Then using a device in the spindle, rotate and check the variances front to back and left to right.
This process is shown in this video from Peter P. : Is your Router on the Level? - CNCnutz Episode 131 - YouTube
However since he has a different model cnc his adjustments are much simpler to make versus the Xcarve Pro, so you may need to shim certain brackets if adjustments are needed. . . BUT the point I must make is that tram needs to be set to a surface that’s Parallel to the Gantry’s movement and the BEST way to achieve this is by surfacing the area that you’re using to measure tram.
You can make a tramming gage quite easily,
An old dull 1/4" straight router bit, a piece of wood about 1/2 X 1 X by 6 or 8 inches long and a 1.5 drywall screw.
An inch or so from one end of the wood use the router bit to drill a hole through the 1" dimension of the wood, Hot glue, ca, or white glue it so that the collet end sticks up enough to mount in the collet, flush on other end. Put the drywall screw through the other end of the wood so the point sticks down. Any errors in manufacturing will be self cancelling.so
Mount the device int the collet and lower the spindle motor so that the tip of the screw just clears the spoilboard parallel to the X axis, turn the spindle by hand 180 degrees and check the clearance at the tip of the screw. You now know the amount of error if any in the X axis,
Repeat in parallel for the Y axis.
Here’s a thread discussing tramming on the XCP and a response from Inventables. It seems they don’t offer a good solution because it shouldn’t need tramming (from what they say). Just wanted to throw this out there to give you some more info.