Any ideas for a cool touch plate?
Since anything works maybe make some pcb ones, or have you seen any that are for sale that are cool? I am not a fan of the really tall ones, a small foot print is preferred. I am trying to think of some instructions and maybe this would be a cool first milling project? A PCB plate?
Good topic Ryan. I am interested to see what comes of this thread. As soon as I get a free weekend I hope to get the cnc up on the new table and make some chips. Would like to explore a touch plate for bit changes eventually to make things a little easier.
I like the idea of a spatula. The diner ones that are bend and have a wooden handle. They are constant thickness, and you can flip them when you’re bored. Plus, if you glued your work down, you’ve got a pry bar.
That not even mentioning how great they are in the kitchen! Does Ikea have one?
I would like to buy something like this https://www.inventables.com/projects/zero-and-corner-finder as I don’t think milling it myself would give the right precision. That’s also the reason why I did not make a PCB touch plate as I believe most PCBs are somewhat bent. And as I like to mill PCBs I want the precision to be < 0,1mm. That’s why I bought one of these cheap Aliexpress touch plates:
Yes, it’s a bit high (19mm), but it’s very even and gives me excellent results.
I am thinking a small pbc, about the size of a quarter, for no warping but easy to hit. Z accuracy is much much easier than XY on this machine.
The corner finder is thing for us would not work well without some serious software upgrades. If we had software to actually utilize that correctly it is better to use a touch probe. That only locates one corner and gives no information about rotation, if it probed it multiple times it could but it would not be nearly as accurate as a touch probe and we don’t have cam control software to utilize that information anyway. So for now it is best to stick with locating pins for anything that needs edge accuracy on a second operation.
Spatula would be cool if it was small like a metal kids one.
I don’t really understand what you mean. Why could it just locate one corner?
It serves as a regular touch plate and an edge finder (for 90 degree angled work pieces). I know Marlin does not support this at the moment but it’s perfectly possible to use it as an edge finder for X, Y and Z with bCNC. You just position the endmill inside the circle and bCNC will probe the corners and find the center spot. Then it can move up and a bit outside, search for Z and apply the known thickness of the corner finder as Z offset.
Finding Z after XY is not automatic at the moment (I think). I have never looked into this as I don’t have such a corner finder but I don’t see a reason why it should not be possible to script this. But in the worst case you can do Z manually after finding XY.
I have some short pieces of my stainless steel conduit left and have used it to have a fixed XY between different runs and repeatability was better than with my microswitch limit switches (those were the ones with levers which reduced precision). I think my current hall effect limit switches are more precise but I have not run a test to compare the two.
What happens when you mount the work pieces rotated 5 degrees around the Z axis? The only way to do this is probe the corner in multiple places. With the software we have at our disposal in the price range we all like, the corner is is just a ball park, locating pins are easier and more accurate.
I am not familiar with bcnc.
If our software could correct for this a touch probe is a better choice as you can probe the piece itself in multiple places. As far as I know we have no easy way to alter the gcode to correct for this on the fly. Some things are going into Marlin right now but it is very much at the Alpha stages.
It depends. If you are cutting a contour with some distance to the edges of the workpiece it obviously does not matter if it is skewed a bit. Then the corner finder is just a probably more precise alternative to limit switches and a touch plate combined. If you need the edges to be precise you obviously need more fix points if the work piece is rotated. I’m sorry, I don’t want to shove it down your throat, but bCNC can actually handle that case as well although it gets a bit more complex. You need at least three reference points in reality and in your Gcode-coordinates. That would be the case if you had a workpiece of a known size. You locate (at least) three corners and tell bCNC what coordinates those are in your Gcode space. It can then rotate the Gcode coordinates to match reality.
This is perfectly doable. At the moment I don’t do this with a corner finder but with a small el-cheapo webcam mounted next to my spindle. bCNC knows the offset between the endmill and the camera center point and can even project targeting lines and a circle of a know size into the picture to help you find the right position. So if you have a hole you drilled before with a diameter of e.g. 3mm you can set the circle size to 3mm and this helps you to position the camera right on top of the center of the hole.
I also don’t see that locating pins and the corner finder attack the same problem. If I understand it correctly what “locating pins” are you have to drill holes in your work piece and your scrap board and fix it with a wooden dowel. This is great for flipping parts over (if you don’t loose (micro)steps in the process in case the steppers are turned off) but it does not help in finding the edge of a work piece. And it does not help with repeated positioning between runs (with turned off motors). This is where limit switches and/or this corner finder could help. Or are locating pins something different? (I’m sorry, English is not my native language.)
I’m sorry if I’m aggravating by mentioning bCNC so often. But I just can’t keep quiet if I know that those things are possible (within our budget) and you say they aren’t. The GRBL maintainer even has a (relatively new) grbl-Mega fork (GitHub - gnea/grbl-Mega: An open source, embedded, high performance g-code-parser and CNC milling controller written in optimized C that will run on an Arduino Mega2560) and merged RAMPS support not too far ago although I have never tried it. If I understand it correctly performance is not quite as good as with the regular pinout as the step pins are not connected to the same Arduino ports on the RAMPS board so the firmware (including Marlin) has to do more work to control the steppers and cannot do it quite in sync (but close enough by all means). But it should be possible to try it out without changing electronics and cabling.
The big thing is a corner find can’t correct the axis skew like we can with the dual endstops, so to implement dual endstops and a edge finder in software I have never used is not something I can commit to right now. If you can make it works that would be amazing. People like you with the knowledge to make things like this work don’t need my help.
The people I am working the most with and spend the most time with, don’t know how to attach pictures to there emails, have a hard time using the forums. We have some older and some younger generation that are not on that level. For them the machine is actually the easy part, the software is the hard part. Guys my age and older know when a cutter doesn’t sound right in wood , younger don’t seem to have spent much time using wood tools.
You guys that understand the crazy stuff, teach me in the forums constantly. I am always asking questions but right now dedicating time to bcnc is not nearly as worth is as making the directions more clear, getting more endmill section (most people are using router bits).
I’m not saying my way is better but it is the way I understand 100% so I can teach it to others that don’t understand it at all.
I am trying to be easy without being limiting I guess. This why I am looking for a simple z probe.
I’m sorry if I came of not the way I intended. I would never think of telling you what you should do or what you should commit to. I applaud the level of commitment you show for the project and the time you invest into it.
I did not know that you were looking for a touch plate for the general user. It sounded more like a curious inquiry what would be nice to have and see what possibilities are out there. I would not even have mentioned the china touch plate if I had known that you looked for an “official” one for the project. I still think it’s more precise than a PCB one which makes up for the increase in thickness, but I would not want to keep more things from questionable sources in stock either, so I completely understand if selling this stuff is out of the question for you (although mentioning it in a side note as an alternative as the one you will be settling on might be nice for other people who are looking for the best precision for the cheapest price).
If you ever have the time and want to look around and see what else is out to experience how other systems work and what might be nice and have some problems getting GRBL/bCNC to work properly, drop me a line and I will gladly help as far as I’m able to. (I have never used a CNC mill before the MPCNC and have never used anything else either. So I’m no expert by any means.)
And BTW you are absolutely right there is no other system out there (that I know of) that supports auto-squaring. I suggested it as an enhancement to the grbl-Mega maintainer as an Github issue a couple of days but he has not responded yet and don’t know if he will as it is probably very MPCNC specific. I ordered a Arduino Mega recently and might look into making a pull request for grbl-Mega, but I did not have a closer look at the source code to see if it would be possible with a reasonable amount of work. And as I’m short on time as well I don’t know if or when this will happen. (So much more reason to appreciate all the time you spend here, helping people and improving you creations.)
I’ve made mine with a little bit of wood and aluminum.
The wood can be milled with the MPCNC, the aluminum can be cut and filed by had or milled, whatever.
It works nicely. I can share the design if anyone’s interested
Why don’t you just use the bed as your touch plate?
I bought these for my MP3DP and use the smaller one extensively. The long one would work well for a thin spatula, though I’d want to either remove the handle or bend it such that it sits fully flat on the surface.
I like those Bill!
I was thinking something more like this: https://www.amazon.com/Winco-TN719-Hamburger-Turner-6-Inch/dp/B001VZAKGM/ref=sr_1_4?ie=UTF8&qid=1510266937&sr=8-4
I guess if it’s smaller, then there’s less chance that a chunk of dust will raise it up, but if it’s smaller, there’s a chance you could push it down into the material some. So maybe the size of a quarter is about right.
This looks like a cool little project, I’ll try milling one out of some scrap once I work my way up to aluminum with the new build.
In the past when I’ve needed to get the XY accurately positioned on a corner, I would get the Z very close and then turn the end mill (only works with flat mills) so a flute lines up with the edge for the axis I’m trying to align and bump the XY around until it feels smooth between the cutter and workpiece, then offset by the tool radius and repeat for the other axis, and a sheet of paper to get the Z lined up. Definitely a bit of work/more room for error, but it works well enough to do every once in a while. Considering we use gages at work that require you to feel a step of a few thousandths for go/nogo, you should be able to get within a several thousandths with this method. May not be adequate when working with a semi finished workpiece.
What is the ultimate goal of the touch plate? I have been wanting to add this to my setup for accurate starting z heights. I was planning to start with foil tape and alligator clips. Are there other use cases I am missing?
Just make sure your foil is on the material. I think the best thing to use is actually a feeler gauge if you have one, even a super cheapy. I think it needs a known thickness and just enough heft and give. A feeler gauge or a thin strip of spring steel even. Too thick or too big and I think it will be easy to screw it up, too thin and it might not touch the surface properly.
Just my $0.02.
