Small CNC Engraving / Milling Machine - variable material thickness?

[hotponyshoes]

I'm looking into the idea of putting some text onto some horseshoes.

A bit like this sort of thing.. but a bit tidier..


These shoes will be made from relatively soft materials like aluminium, copper or brass.

So I'm after recommendations on a suitable machine to do this job.
Capacity of the machine won't need to be any bigger than 200x200x20mm (and the engraved area much smaller)

I'm not sure if laser is much use at making a decent mark on those materials?
If not, I'm thinking machining with a 1mm ball cutter to a depth of 0.5mm will give me the look I'm after..

They can all be engraved in the same location so I'd think a couple of locating pegs on the bed and some over-center clamps should do for quick loading?

My first concern is these shoes are all handmade so the thickness is going to vary a bit.
How do I account for that?
Any desk-top size machines that come with a probe to recognise the height of the inserted material?
Or twin heads (or tool changer) so I could machine a flat area then engrave the text into that area?

I'll probably run the top surface over a belt sander first. Maybe the bottom surface as well but I don't really want to be measuring each one to get them 0.1mm perfect so need some way of coping with the variation if the text is only 0.5mm deep.

 

[Erie Fred]

If they vary in thickness, don't locate your part on the bottom.

Pull it up with clamps to a plate, has a window in it for the engraving cutter.
(and to miss those pyramids sticking up)
That way, you'll reference the machine on the correct surface.

 

[hotponyshoes]

If they vary in thickness, don't locate your part on the bottom.

Pull it up with clamps to a plate, has a window in it for the engraving cutter.
(and to miss those pyramids sticking up)
That way, you'll reference the machine on the correct surface.

No pyramids on the ones I'll be engaging, top will be flat (ish)

But pulling it up to clamp it makes perfect sense!
Got to be the way forward

 

[Erie Fred]

No pyramids on the ones I'll be engaging, top will be flat (ish)

But pulling it up to clamp it makes perfect sense!
Got to be the way forward

When you do tool & machine design like I have for too many years, it's easier
to see the problem....

In this case at least.....

I hope I have instilled a sense of looking at things from a different way.

I find there is a trick or angle to every job, just need to think a bit to find that key
to unlock that door.

 

[Agroshield]

I was to say a similar thing. Set the top (to-be-engraved) surface parallel to the bed of the machine and let the base be wherever it ends up.

The fixture does not need to be permanently attached to the machine bed. Imagine something like a tripod as used in your chemistry class at school. Fix the parts to some of those offline (off the machine)* and then just have something on the machine into which the three legs are inserted and clamped.

Having an interchangeable fixture allows you to prepare the next one while the current one is cutting and spindle off time (part changeover) is minimised.

If you were doing this for production, you might put more than one on the bed at once and even interlock the U-shapes to maximise the number within the machine envelope (the machine don't care if it engraves alternate ones upside down).

* doing it offline also allows you to invert the fixture and drop the shoe into it with the engraving side downwards. Gravity holds it while you clamp it. Having the fixture on the machine, the shoe is always trying to fall as you are clamping it.

 

[Dev]

Desktop fibre laser would do that no bother and not worry about slightly different heights.

 

[GaryS65]

Pantograph engraver?

 

[julianf]

Depth only really matters if youre using a v cutter. Within reason.

If youre just going to mill a channel in it, so long as the varying load is acceptable, what does it matter if some of the cut is deeper than other areas.


To laser metal you need to either spend a whole load of time, or proper money. A mechanical device will be way cheaper for the same result.


Detailed engraving is always done in two passes, but then things like backlash are important. If youre milling a groove, and its not like you need to be able to engrave a unicorns head on the cheek of the queen on a 1 penny coin, just get, or make, a simple cnc mill.

Your spend goes up a load with repeatability, but, really, there are always degrees to all these things, and, for a 1 pass cut on somthing big, a bit of backlash (which you can compensate out in software to some degree) wont be an issue.

Personally, id just make somthing, rather than buy a kit. It will cost you more, but youll know how to fix it and you can do things to spec, rather than just buying a jack of all, master of none.

 

[sako243]

I've got a cheapo Genmitsu 3018 CNC for milling out PCBs and occasionally I use it for machine out holes in aluminium which it can do slowly (usually doing a 0.5 or 1mm depth of cut etc but I'd need to double check).

It comes with a little zeroing touch probe that you attach one lead to the mill bit and it descends until it touches electrically. You can dismantle the probe and attach the other lead to the conductive material. This technique can also be used to contour map the surface of what you're milling so if it's not flat it can maintain a consistent profile.

You can just spy the touch probe back left with the red Croc clip, silver pad and blue housing. Basically the software knows how tall the probe base is so once it's touched down you can zero things automatically.

That's roughly an A5 panel on the to give you an idea of bed size but it can be extended. They do laser modules too but not tried them.

 

[sako243]

Here's a slightly better photo of the setup above - which is this kit - now down to ~£205 and pretty much has everything you want in it apart from bits (actually it does come with some I think).

The "touch probe" is basically checking for a short between the crocodile clip and the metal plate on the blue "button" thing in the back corner. The USB port just appears as a Serial port so I actually have my own Python code chatting to the GRBL controller (you basically just send it G code via the serial port) but I'm doing more maths and interactive stuff hence why it's attached to a PC / similar. My issue is that of alignment - I'm etching the PCBs becuase it offers far better quality but given the tolerances needs to be drilled pretty accurately.

What I do in this instance is the code prompts to install a drill bit and them moves to where it thinks "pin 1" is on the bottom left PCB, I jog it manually using the keyboard until it's centred, then the software goes to the top-right PCB and does the same thing. Given those two fixed points I can work out the translation and rotation matricies required to align everything virtually in software without having to move the PCB panel.

When it installs a new tool into the collet it zips over to the touch probe in the back right corner, prompts you to attach the croc-clip and then it touches off on the pad. Then I know the pad is ~17mm tall so can calibrate the zero offset, I'm not too fussed it being supremely accurate in this instance because I'm always plunging straight through the board so as long as I'm deep enough to go through it's fine.

The main issue I've had to overcome is fixturing things because a lot of the "space" between the boards gets milled out by a 2mm end mill to act as little mouse-bites so I can break out the PCBs after they've been populated. Consequently as it's only 0.8mm thick FR4 it flexes quite a lot so I've found the best approach is to apply the blue 3M masking tape to the back side of the panel (it's only single sided) and then some double sided tape, some broken drills serve as rough alignment holes into the MDF backing board and those holes are drilled manually first. The tape combined with a relatively slow XY traversal speed at full-depth means the tape holds on well enough and is easy to strip off afterwards.

Just had a look through the GRBL files I had for milling out some aluminium panels and it was doing a 0.25mm depth of cut but faster than I could do at full depth through some 3mm aluminium boxes. I found that the best combination given fixturing that particular extruded aluminium panel was challenging. Again I was just milling out holes for an LCD and buttons so full-depth through the panel so mapping out the contours weren't important.

For your particular application (it's actually guided through online) but basically you dismantle the probe in the background, the "button" bit can be attached to the conductive metal part you're milling and you attach the crocodile clip to the milling cutter (without it spinning obviously). The software on the PC driving it can then basically iterate through a whole grid of points mapping out the height so it can be used to calibrate the Z depth per point. I have once played with this but haven't had a need to use it so I'm afraid I've forgotten what software I used. It's used a lot for milling of PCBs rather than etching as they're typically milled using a V shaped cutter so depth of cut is critical into getting the correct width of tracks etc.

I'll be honest for the £2-300 I've spent on this thing it's been great for what it is - yes it's light cuts but it's a fully functioning CNC that's by and large mostly open source so easily modifiyable to improve if you want (it runs a GRBL controller). I was going to build one but then thought by the time I'd messed about I might as well see how this one does and for what I want it's been fine.

 

[Erie Fred]

Using a touch probe may well in fact work, depending on how close and how many touches you do.
However, it's good to get the basics of tool & fixture design down, before resorting to hi-tech solutions.

What am I trying to do ? where is the machine referencing (spindle) from ? and other details.

 

[Misterg]

The software on the PC driving it can then basically iterate through a whole grid of points mapping out the height so it can be used to calibrate the Z depth per point.

bCNC can do that without having to pay for something.

I have to say I'm sceptical about 3018 machines for metal work, but I guess you won't know unless you try.

The next step up is a big hike in price: £2k ish for a Nomad:

Nomad 3 - Desktop CNC Mill

This product is expected to ship within 7 business days. The Nomad 3 Desktop CNC Mill is the third major revision of the Nomad 883 that launched Carbide 3D. Overview Specs What's Included FAQ Warranty and Support Turn your Desktop into a Workshop™ Here are the big changes for the Nomad 3: New...

shop.carbide3d.com

 

[Dev]

If you’re going to go down the desktop CNC route there’s a bit of development in that area just now with things like

https://www.kickstarter.com/projects/makera-inc/makera-z1-entry-level-cnc-pro-level-power?ref=f21vma

And

https://www.kickstarter.com/projects/959480926/nestworks-c500-next-gen-smartest-cnc-with-industrial-power?ref=profile_created&category_id=335

 

[m_c]

It mostly depends on how much money you're willing to spend on a machine.

3018s (and larger versions) can be OK.
I'd certainly take one over the likes of the Nomad or the other options, as when you've seen inside most of those machines, you're essentially getting a similarly constructed machine in a shiny case (and in the case of the various kickstarters, lots of added hype too), that you can achieve with a 3018 machine, and a few select improvements to iron out the common issues with the basic machines for less money.

If you're willing to throw some money at it, then an XTool F2 Fibre laser could tick the boxes for you.

 

[Dev]

That’d be my choice too.
If you can stomach the price they are a very versatile tool for general craftiness without being as noisy or messy as a rotary tool.

 

[hotponyshoes]

I've been looking at the Xtool F2...

There seem to be a couple of versions / generations?

What sort of power would I need to engrave 1mm wide text to a depth of 0.5mm on aluminium/brass/copper?

Will a desktop laser like that actually engrave? Or just etch/mark?

 

[Agroshield]

No pyramids on the ones I'll be engaging

If there are no pyramids, cold you set up two abrasive wheels (or belt sanders) with a gap between them, poke the item in and assure constant thickness.

Even a rotating drum sander with a fence and push through the gap.

 

[Erie Fred]

If there are no pyramids, cold you set up two abrasive wheels (or belt sanders) with a gap between them, poke the item in and assure constant thickness.

Even a rotating drum sander with a fence and push through the gap.

A cheap double disc grinder !

 

[hotponyshoes]

I'm not sure about running them through a belt sander as they are all hand made and I think putting a flat surface on them will ruin the hammer finished look.

I think clamping them from below should be fine then the engraved surface will be mounted in the same place for each one.

I'm looking into the laser idea as that won't really need any clamping, just holding in place. I'm thinking a plate with a hole fixed to the machine base (so the underside of the plate is set once as top of object) then a simple spring-loaded platform under that so I can just push it down, insert the shoe, let it raise. If I can make the upper plate out of perspex I can manually jog the shoe to the correct position while it's lightly clamped by spring pressure.

Apparently the x2 can emboss metal to a depth of 1-2mm?
I'm not sure how well it will work on a section of something like 30x15mm of material that will conduct heat quickly?
Or is the heat from the laser so localised the metal doesn't really heat up much?