for my current project I want to engrave a slightly curved board.
I can’t mill the board straight because 1) it’s bigger than the mill (It overhangs the sides) and 2) I would mill away the discoloration due to aging.
Now I know mesh bed leveling from my 3D printers.
That would also be the right approach here, but the classic BLTouch sensor is out, because the cutters can always be different lengths and I do not always want to adjust the sensor to the right length.
Currently, my Z-axis is homed by contacting a metal plate.
I wanted to keep this principle.
My idea was to attach a metal contact below the cutter so that leveling is possible everywhere. So that the cutter can get contact to the metal contact, I attached the contact with the help of springs. These allow that the cutter can get contact, but the contact is immediately separated again as soon as the cutter moves up again.
I haven’t tried mesh bed leveling with it yet, but homing seems to be pretty reliable.
Here you can see some photos and a video of my approach.
The only problem I see with that is there may be errors introduced by friction on the spring slides as well as tilting in the sliding mechanism. There are other trade to print mechanisms that do what you are after but use a break open style contact, like this one:
I would look into something like that instead. The break style probe will have far less errors. The main error with those style probes is repeatability of the rested position. There are some designs I have seen on tv that do well to address that. This would be the or area that would make it worth designing your own if you have better ideas to make it recenter more reliably.
Tilting and friction are indeed problems that I have also noticed. I am currently trying to minimize the friction with lubricant. This works so far also so far that my measurements are all in a range of 0.05mm.
For the tilting I have not yet found a good solution.
The problem with the 3D probe approach is that I always have to remove the milling bit and then measure how far the milling bit protrudes from the holder.
Since my spindle has no quick-change function, this method is rather unsuitable for me, or do I see it wrong?
It seems you would have to do the same regardless unless you can ensure consistent z offset from the bit within your precision goals. In the grand scheme, most parts require a tool change to complete anyways. So having a tool probe is a requirement, and then operating the initial probe and first tool probing are not a huge setback. In this case for a one off job with one tool only, perhaps you could make do with just the tool and some tin foil to avoid the tool probing part?
