Sorry if this is the wrong place for this, I couldn’t find a more relevant Forum. (proposals for future development belongs here, so this might…)
What are your pipe dreams for a “perfect” version of Maslow, if time, effort, money, etc were no object? (But based on the Maslow base, not a commercial CNC :P)
I’ll start: I used to work with a guy who did computer vision programming, so I know this is possible. It would be awesome to have a camera set up to monitor the sled’s position on the 4x8, and do an auto-stop if it wasn’t where it should be. So far most of my serious problems have happened due to chain fouling, and this always is first noticeable (unless you’re watching the motor at the time) because the sled takes off in a random direction…
Any other large improvement/addition projects that will likely never happen?
How about your computer vision idea to actually position the router…
Have you seen the “Shaper Origin” - it uses a camera to get positioning data from “dot strips”… wonder if it would be possible to have a camera hanging out above the workspace and get absolute position from it. Combined with the motors counting their encoder locations, I would think you could get even more accuracy (and as you say, you could e-stop if things are going off-path for whatever reason).
Another one - somebody else suggested an auto z-index zeroing with a piece of tinfoil/other conductor on the cutting surface, and a current detection from the spindle.
What if we had a metallic backboard, and used a similar system to prevent cutting deeper than we need to?
This is a pretty minor thing, as I don’t mind engraving my backboard, it’s replaceable, but it is a pain having to create a whole different project to cut the same shapes out of different thicknesses of ply…
The auto zero is already there. You actually want to cut a little way into the spoil board to get a complete cut through the workpiece, so the spoil board is a consumable. I’m using 1/4 OSB, the cheapest stuff the local lumber yard sells.
me too, like I said, it’s less about saving the spoil board, and more about not having to set depths to cut similar pieces from different materials. Having the machine know it’s gone far enough would be cool
Off the wall idea… Cover the backer board with aluminum foil then hook a wire to it and a wire to the router bit bearing. Put 5 volts volts to the wire going to the router bit and hook Aluminum foil to one of the IO lines of the MAslow CNC to sense a contact then you know you went to far.
I swear when I wrote my post that yours was not there but it clearly shows the time line of you having posted the idea first. You must have planted the idea in my mind… Good job.
One is ground, the other is connected to AUX4 which provides the voltage. It’s probably better to have the foil grounded. I couldn’t find a way to connect to the bit except directly, the body of the router is isolated on mine.
I am guessing we would have to hack into the router case and add a screw coming out from touching the front or rear bearing housing. The actual bearing would then be the electrical path to the shaft holding the bit. Then we could add a wire from the screw to the I/O pin.
At 20K (+/-/whatever) rpm that brush is going to be very electrically noisy and prone to constant nuisance trips.
@TomTheWhittler I’d be very, very hesitant to do something like that from an electrical safety standpoint. Plus it would void the UL/CSA/etc approval, which means if your house burns down your insurance company can laugh and leave without passing on a check. Also you’d have the same noise issue; rolling bearings aren’t a great way to get a good electrical connection, what with that lubricant film making and breaking the connection.
As an editorial comment/IMHO, I’ve been dinking around with little CNC machines for must be 10 years now, and never seen or read about anybody worrying about this issue. Use your cheap import digital caliper and measure the material thickness, and cut just enough into your spoil board (a few thousandths with my belt driven Zenbot, admittedly it was easy to clamp parts down since there wasn’t a sled) to make sure you’re always through it.
I don’t think this is a workable solution. You’ll notice that no insanity priced CNC uses this method. They use a metal block ( of exactly know thickness and known max possible distance of travel in the z axis ) that’s either placed on the material and the computer then determines the exact thickness of the material, or that have a metal surround that bit touches to ground out. This is because as you cut the spoil your going to loose areas that have been cut off from the rest of the circuit. Constantly replacing that foil is going to be extremely tiresome. I mean you’ll have to use some sort of glue on the foil. Over time that’s going to become a headache. You’ll have to constantly apply new foil. Headache. Any time you make a full cut out anything inside the cutout is going to not register anymore.
The updated X-Carve has a prob that clips to the spindle collet and has a metal disk of exact known thickness. Spindle is lowered until it touches the disk. Computer then knows material thickness and prob is removed and you start cutting. Something like that is workable.
I honestly think it would be easier to set a pressure switch or proximity sensor that was placed inside the fixed base of the router. That could judge the distance between it and the collet. Though exact measurement of the exposed but would be required. Which only reinforces the X-Carve method of a prob that you attach lower bit till it touches the metal disk and Ground Control factors in the material thickness. Then sets a depth limit just past that thickness.
remember that the sled moves, you can lift it away from the workpiece and sliip
a 1/4" thick plate between them and let the maslow check the bit and then remove
the plate.
I use an 8"x12" piece of aluminum flashing. Almost any conductive material would work so long as it is flat and spans the hole in the sole of the sled.
Now that I’ve got my Maslow up and running, I think its ideal use is for something like the Shaper Origin. The drawback to the origin is that you still have to pilot it around the workspace. The drawback to the Maslow (at least so far) is the accuracy. If you put the two together replacing the regular router, you theoretically have a high accuracy (they claim 0.01") vertical 4x8 CNC machine for still less than $2500. Just hate the concept of buying the special tape all the time.
Neat idea… I hadn’t really thought of combining the two…
I would think there would be some way to use the technology of Shaper (micromovements of the router) to key off the fixed points of the frame (say, put two reflectors next to the motors and have lasers or something) - that way you wouldn’t need the tape.
