Centroid Community CNC Support Forum

Alright folks, I thought I'd share this set of macros I've built.

On my ShopBot I have a spoilboard that I plane down with the machine from time to time. Most jobs I cut into the table a bit so I get clean through whatever sheet good I'm cutting.

Over the years, it seems every time I replace my spoilboard with a brand new one, that's the time I set up my CAM software the wrong way and gouge into the table. That, and when I've been doing demos jogging the tool around with the pendant, I've nailed the spoilboard a few times.

I figured now that I have real machine coordinates (ShopBot doesn't have this) I could set it up so that I could never smash my Z into the table, or deeply gouge the table.

To do this I set my jog distance so that it's the exact length of whatever tool I have installed. This distance is set whenever I zero out a new tool. So no matter how fast I jog towards the table, the bit will always stop right on top of the table. The one problem with this is that if I send a job to the machine that calls for cutting into the table a little bit (just about every job) the code gets kicked out because of exceeding the Z travel limits.

So what I did was setup a series of macros that take care of all of this. When I home my machine, it automatically sets the max Z travel for the bit that's in the spindle.

When I start a job and a tool is called, the Z travel is extended .04 so I can cut into my table a little bit. When M30 is run (when a job is over) the travel is set back so that a tool can only go to the table top.

Here's a video of it in action:

That's cool! Nice video too. I really like that you were able to solve a problem in an innovative way.

Do you always z-zero to the spoil board? I have found that there are certain processes that just make sense to set z-zero to the top of the workpiece. I do quite a bit of work that requires cutting t-slots on the back in order to hang on the wall or v-carve text. Because the material my customer supplies vary quite a bit I would have to painstakingly measure the thickness of each board. I finally figured out that I could save a bunch of time by just zeroing to the top of the workpiece and cut the t-slot or v-carve the text.

In the g-code header information from Fusion360, it tells me the minimum Z value for each tool. I will look into using that information to do something similar to what you are doing, but still be able to z-zero to the top. It would be really nice to not accidentally tell my machine to do something stupid.

I work like Clint does- z zero to the top of the work piece. I really like your idea, Toaster. Sounds like something I need to sort out.

So I still zero to the top of my material a lot, but I don't... Let me explain...

First it's worth noting that I have a prox switch at the top of my Z, and that touch off plate. It's also worth noting that I change tools pretty regularly and they're just clamped into the collet. So every time a tool is put in, it's a different "length" than the last one.

Every time I change a tool the length of that tool is measured from the top of the Z travel, to the spoilboard top by way of that touch off plate. So a tool that sticks out of my collet 1" is probably a 4.5" tool... That 4.5" means that from the top of the Z travel it'll take 4.5" for that tool to hit the top of the spoilboard.

The beauty of this system is that once a tool is measured, even if I power off the machine and power back on, my Z zero is remembered. Make sense so far?

Now the obvious use case for this workflow is cutting stuff like plywood where zeroing to the table is what you'd normally do. I use this same method for V Carving too.

So what I'd do for a VCarve is caliper my material and set that thickness in my CAM software. I leave the Z origin at the table. When I switch tools I use that same, very accurate touch off plate. Have you ever done a VCarve with a big clearance tool, and come back with the V bit and had the Z be off a little? If you're using a metal touch plate you can get different zero heights with different shaped tools. With that mechanical/inductive plate I have, that doesn't happen.

So you can still "zero to the top of your material" but I just zero to my table via the plate, and tell my CAM software where the top of the material is.

If that doesn't make sense, don't read on until it does because I'll confuse you further...

There ARE cases where I want to zero to the top of the material, they're rare, but they do happen. Let's say I have some irregular material that I just can't reach with a caliper, or otherwise measure properly. For this I created this script:

https://youtu.be/nW_7iHxc0QM

This lets me measure with a touch off plate. Now I know what you're thinking "He just said those metal touch off plates aren't as accurate as that inductive plate he's got" You're right, I did say that. What's happening here is that I'm using that touch off plate to to offset the current Work Coordinate System in Z. I only have to do this one time with one tool because it raises the Z up off the table to wherever I measured with that plate. So every tool change after that uses the touch off plate for accuracy. To repeat myself: This script just raises the current WCS Z UP by whatever the thickness of the material is, and you only have to do it once. Make sense?

So you can still "zero to the top of your material" but I just zero to my table via the plate, and tell my CAM software where the top of the material is.

I used to do exactly what you do, however, I found that that was a huge PITA when material thicknesses changed unexpectedly.

Let me explain: I had a decently large production run for retirement and service recognition plaques. On the front of each plaque, I cut out the outline of the years of service and cut out the perimeter (Zero to the Spoil-board). On the back, there was a t-slot to hang on the wall and a recess for a small easel to set in to be able to display the plaque on a table or desk. The depth of the recess was critical 0.275 +- 0.02 inches. I would Zero to the Spoil-board and used the CAD software to compensate for the thickness.

The customer supplied the material that was a glue-up of Walnut, Maple, and Cherry. Initially, the glue-ups were extremely consistent in thickness, then they started to vary. Initially, the parts were exactly 0.75" But over the course of the run they varied from 0.625" to 0.875". This posed a HUGE issue trying to maintain 0.275 +- 0.02 inches for the recess. Rather than using a set of calipers on each board and using the CAD software to modifying the g-code for each board I simply kept the TOP's z-zero as the Spoil-board but the Bottom's Z-Zero was the top of the part. Problem solved the remainder of the job went perfect and the depth stayed at 0.275 +- 0.02 inches.

The overall thickness of the part is not a critical dimension for 98% of what I do.

Because of this experience when I need to cut thru a part I Z-Zero the spoil board. When I do anything that the distance is relative to the top of the part I Z-Zero to the top of the part. It simplified my production setups, speed, and improved the overall quality of the parts.

Ahh, so I see how you solved it, you essentially tricked the CAM software into it...

That obviously worked fine for you. You could do it with this method too...

You'd essentially set the CAM software for Z zero at the spoilboard and save out all of the toolpaths that you'd cut that way, so the paths that would actually cut out the plaques.

For the ones that you need to zero to the top to cut out, you'd just set your CAM software so that your Z zero is at the top. Then you'd use this plate to zero to the top of the material and run the job.

With that G52 offset you're essentially raising the zero plate from the table to the top of the material, so you wouldn't need to re-zero your tools, and you could even come back to a job if you had to power down in the middle of one. All you'd need to do is use a zero plate to automatically raise up your G52 Z whatever the thickness of the material is.

If you wanted too get really slick you could mod your post processor to ask you if you need to zero to the top of the material and than have it go and do it for you like this:

https://youtu.be/nW_7iHxc0QM?t=302

After you zero to the top it could go off and start cutting the job. That might be a good way to do this in production...

Toaster wrote: ↑Tue Dec 11, 2018 12:02 pm With that mechanical/inductive plate I have..

Would you be up for sharing the details on this?

First, full credit to Gary Campbell on this one.I took his design and re-worked it into something that is 3D printable. I 3D print the base and top, and then I drill and tap if for the threads and the bolts.

The idea is essentially to use an inductive prox switch and slide a bolt in sideways in front of it when the top is depressed. The inductive switches are very accurate, and because you're touching off on a flat surface the bit size/material doesn't matter.
http://a360.co/2jqx7Nn

You can download the design with that link if you want. I set it up for some rare earth magnets in the bottom so I can stick it to my machine.

This is really clever! Would you mind walking me through how to implement the code?

Also do you mind sharing the code using the touch off plate? I love feature you added to touch the bit with the plate to verify it's working and having that initiate the sequence!

One thing that isn't clear to me is the tool touch off in relation to your spoiler board. I assume there is an offset between the two every time you plane down the spoiler board. Can you explain your procedure when it's time for a resurfacing?

Thanks for all he helpful info, I've been browsing your youtube channel as well and you pointed out some very useful things!

All of the code is in the first post. It's commented pretty well so you should be able to step through it. If you're not sure what something does, grab the Centroid manual and look up the codes and numbers and you'll see what they correspond to. You'll probably have to alter a few things to suit your setup, in particular the input numbers for your prox switches.

You are correct about the offset difference between the switch and the spoilboard. It does change every time I plane down the spoilboard. For that I have a simple script that measures the distance between the two:
This is the script that let's you use the touch plate anywhere: