My first steps into home CNC

[BuggyDad]

1 hour ago, Wystan Withers said:

Lots of details is an understatement. Not to mention CF is not cheap if you goof it up. I've been getting my sheets from a vendor named ARRIS on ebay, dunno if they also post on the UK side. Only under VERY close inspection can I tell a difference between their stiff and Tamiya's supplier... the Tamiya stuff might be a tiny, tiny bit glossier, but you really have to look closely and be super picky to even get there. Currently I'm running at 100mm second linear travel, 9k rpm, and 0.3mm depth per pass. I can likely speed that up a bit but I've cost myself enough 'whoops' CF at this point I dialed it way back to "better safe than sorry" territory. Takes 4-5 hours to cut a chassis baseplate at these speeds.

Funnily enough, I was thinking I'd come back here to check the speeds you cut at for safety. So thanks! You read my mind! 😉

[BuggyDad]

I've had a bit of progress today. 

I've got enough 3d prints of water bath sections done to be reasonably confident I have the fit right, my design is strong enough and my holes should seal OK with recesses for o-rings. It'll be a while before I get the rest done though. I'm not going to be near the printer enough to finish the job in the next couple of weeks. 

And size wise I should be able to get up to 300x400mm sheet stock in and out OK. 

And then, about 150 YouTube videos in, I have got far enough with the software to turn one of my existing part models into a real life thing. 

Kinda.

If this tool path was in 3mm cf sheet over a sacrificial base I'd have a shock tower fixed into its stock sheet with just two little tabs. In principle, all good. As long as I got the speeds right and didn't snap my bit. I won't try cutting CF or GRP til I've got the water bath though, for the dust, so I think I'll leave it there now except for simulations. 

That's using Fusion for the tool path stuff and just opening it in UGS to send to the machine. Took me ages to work that out, and it may take me a while to repeat it I think. It's not very intuitive. I don't know whether there are more variables than 3d printer slicing or whether I just find the software harder to navigate. 

One thing I notice - travel speed when the spindle is up out of the piece is set to same as cutting speed - dead slow. That could do with being sped up. 

[Wystan Withers]

So jealous of that larger table  Those speeds were with a 2 flute ball nose bit. I think you could safely go faster with a 4 flute, but don't quote me on that, it's all still pretty new to me too.

[BuggyDad]

8 hours ago, Wystan Withers said:

So jealous of that larger table  Those speeds were with a 2 flute ball nose bit. I think you could safely go faster with a 4 flute, but don't quote me on that, it's all still pretty new to me too.

The bit I used is an end mill. It has two cutting edges at the base but I'm not sure it's fluted as such. Looks like this:

Or are they lots of little flutes? I was thinking the way carbon cuts (and using a cutting depth of only about 0.3mm) it wouldn't really need to worry too much about lifting the material up and away, more the multiple cutting edges would do their job and the spinning bit would keep the dust out of the way well enough. 

I was cutting at near enough the same settings as you except 200mm/min, so overall twice as fast. So once I get some CF in the mill I'll halve the speed. I know it was a completely different material I was cutting but mainly I was trying to practice run a carbon tool path and settings. 

[Wystan Withers]

Looks like a 4 flute to my eye. Think of it like a candy cane- from the side it looks like lots of stripes, but viewed from the end it tunes out it's just two green and two red strands all twisted up. I agree that with the water bath material removal isn't an issue, also it stays nice and cool so the tools should last longer. I've been using the 2 flute because it was the only ball nose I had but I'm about to talk myself into one of those fancy $20 for a single just because.

[BuggyDad]

25 minutes ago, Wystan Withers said:

Looks like a 4 flute to my eye. Think of it like a candy cane- from the side it looks like lots of stripes, but viewed from the end it tunes out it's just two green and two red strands all twisted up. I agree that with the water bath material removal isn't an issue, also it stays nice and cool so the tools should last longer. I've been using the 2 flute because it was the only ball nose I had but I'm about to talk myself into one of those fancy $20 for a single just because.

Is a ball nose better than a flat end mill for CF? Better finish? Faster? I guess it might be less likely to chip? 

[yogi-bear]

1 hour ago, BuggyDad said:

Is a ball nose better than a flat end mill for CF? Better finish? Faster? I guess it might be less likely to chip? 

bull nose cutters are great for 3D profiling. I would stick with an end mill for flattening and just cutting plates. And there are different types of end mill too. Some aren't designed to plunge straight down as there is no centre part on the end of the mill that can cut.

[BuggyDad]

1 hour ago, yogi-bear said:

bull nose cutters are great for 3D profiling. I would stick with an end mill for flattening and just cutting plates. And there are different types of end mill too. Some aren't designed to plunge straight down as there is no centre part on the end of the mill that can cut.

Have a feeling my cheap bits are like this. Does a 0.3mm drop into a flat face to start cutting count as a plunge? Is that a Mercedes for vacuum forming? 

[yogi-bear]

11 minutes ago, BuggyDad said:

Have a feeling my cheap bits are like this. Does a 0.3mm drop into a flat face to start cutting count as a plunge? Is that a Mercedes for vacuum forming? 

I don't think so, and probably depends on what your machining too. And not a Mercedes, although I'd like to do body one day. It's from a still unfinished Volvo build, the 3D model I found online somewhere, but it had a couple of errors.

[markbt73]

Watching with interest... I used bigger versions of these machines for many years in the sign industry. Seriously considering picking one up when I have a place to put it.

Feeds and speeds, cutter shape, and flute count are all sort of interconnected; mostly it's about getting the chips up and out of the work area as quickly and easily as possible. Lots of resources are available online to help you figure out what you need. But be warned - it is a BIG can of worms.

As for travel speed above the material, that should be an adjustment you can make in the software. Not sure where it will be, but somewhere in the toolpath generation. It will give options for X/Y travel speed, Z (plunge) speed, and no-load or up-travel or something. You should be able to max that out so it just zips over to wherever the start of the next cut is.

[Wystan Withers]

My thought on the ball nose was that it would make the countersunk holes cleaner. Maybe it did, maybe I changed some other setting at the same time, dunno. It's a lot like tuning a carburetor- if you change 5 things at once you never know which one fixed your problem, but it takes forever to just tweak one little thing at a time when you're having to walk back and forth between your office and the garage for each change.

Bugged- if you're using the 'Bore' command and not the 'drill' command in Fusion it enters in a corkscrew motion rather than a straight plunge so technically all the cutting is being done on the outermost point of the bit, not the center like with a drill bit. If you click the box for 'Finishing Pass' it will then do a second pass using the sides of the cuter to clean up the hole after it had been bored out.

As always, big grains of salt here, still noobish at this stuff.

[BuggyDad]

Printed all the parts for my water bath, and the hole positions look OK*, so I'm a little further down the track. 

Glue, sealing, screws, sacrificial top sheet and testing still to go.

*After a "measure twice, cut once" moment that I only caught because I have to print two of everything and so I could make up the mistake by adjusting the second copies

[BuggyDad]

Glued and epoxy sealed the water bath. Now to test whether it's watertight... 😬

It looks OK but I shall leave it a few hours and see if there are any drips. I could give it a second coat of epoxy. 

Thats 12mm deep. Hopefully that's enough. There's room for a little more but I'd be at infinity pool levels of spillability. No hiding a not quite level workbench! I can see myself making a right mess trying to get dust-filled water out of this. 

Next step is to buy and then drill holes to clear the screw heads in a sheet of acrylic. Low profile screw heads so 5mm acrylic will be enough. I might also grab a couple of sacrificial sheets at say 2mm which can be spot glued on top with something like shoe goo so they can be pulled off periodically and replaced easily. I wonder if there's a way to use the machine to check how level the bed is? Because I could also mill the top of the first sheet to perfectly level before going further, if needs be. 

[yogi-bear]

even if the spindle isn't trammed yet, you can still use a end mill or routing bit to machine the top surface and level the bed, just take small step overs. Yes it will take awhile, but you shouldn't need to cut too deep, you are really only skimming the surface. That will make the overall surface the spindle can reach flat. If you want to check before hand, you can use a z-probe, which is basically a simple circuit from the spindle to a metal plate on the surface (ideally you know the exact height of the plate, then you can also zero the end mill height). (just google cnc z-probe on YouTube too see how they work) The plate and spindle are connected to the z-limit circuit on the board, so when you lower the end mill, as soon as it touches, you will activate the z-limit circuit and it'll stop. Zero out one spot and test others to see the difference.

One option for holding down a work piece or item is to use painters tape, apply to surface and apply to the base of the object you are holding down. Use superglue (cyanoacrylate) to glue the two together, holds things down very well and because painters tape can be peeled off, can make objects to remove.

[BuggyDad]

I got round to drilling, glueing and fitting an Acrylic top surface today finally. 

It looks a little messy but I only put a relatively small amount of epoxy down, my theory being that I didn't want more than just enough to stick it, so that extra epoxy volume didn't result in some areas raised. Weighed it down fairly heavily to dry. 

I've been shying away from doing the computer work required to get further though, I'm just not feeling that at the moment.

On 6/4/2024 at 9:41 PM, yogi-bear said:

If you want to check before hand, you can use a z-probe, which is basically a simple circuit from the spindle to a metal plate on the surface (ideally you know the exact height of the plate, then you can also zero the end mill height). (just google cnc z-probe on YouTube too see how they work) The plate and spindle are connected to the z-limit circuit on the board, so when you lower the end mill, as soon as it touches, you will activate the z-limit circuit and it'll stop. Zero out one spot and test others to see the difference.

Yes. Thank you. Mine came with a z probe and I intend to do just this. With all the manual steps, glueing and whatnot, I must say I'm not hugely confident in it being all that level. 

[Tamiyastef]

54 minutes ago, BuggyDad said:

shying away

You're a brave man IMO. There are several "skills" I believe to be highly desirable for our RC hobby. CNC, 3d printing, airbrush,... 

I can't bring myself to get into one of them. By gathering information on them I realised they are all hobbies (or jobs) on their own and not just another aspect of RC cars.

[BuggyDad]

I finally got around to the final steps needed to make my first part. Fixing down a sacrificial sheet, reacquainting myself with software via about 10 videos when the pertinent information was in about 30 seconds of one of them, setting speeds*, soldering back together the power cable for my laptop that my mother-in-law's dog chewed through, persuading the computer that the machine really is plugged in etc etc. Tasks relating to software took a great deal of (largely unnecessary) time and focus for this luddite. Gluing stuff to other stuff and pouring water about, not so much. 

And then the final bit, pressing the "send" button and holding my breath for 24 minutes, was a piece of cake! 

Lo and behold, Super Falcon has a shock tower**

Really nice finish too. I could give the edges a very light sand but overall this is genuinely finished to the standard you'd expect from a bought part. 

*plagiarised from the marvellous Mr @Wystan Withers

**not quite, I used the wrong thickness carbon, d'oh! 

[Wystan Withers]

Looks awesome brother. Hopefully your experience engenders the use of far fewer explicatives than I end up using 

[Elbowloh]

Is your rig big enough to do a lower deck? Looking for a TB Evo 6 deck...

[BuggyDad]

2 hours ago, Elbowloh said:

Is your rig big enough to do a lower deck? Looking for a TB Evo 6 deck...

Well, it is big enough but I've only made one part so far. I may be able to do you one one day but it won't be very soon and I don't see myself going into business! Have you got a CAD drawing of it?

This does remind me actually, there are a few rare old items I'd like to make and I'll be hunting for patterns/measurements/drawings. I think I'll start another thread on that..... 

[Elbowloh]

1 hour ago, BuggyDad said:

Well, it is big enough but I've only made one part so far. I may be able to do you one one day but it won't be very soon and I don't see myself going into business! Have you got a CAD drawing of it?

This does remind me actually, there are a few rare old items I'd like to make and I'll be hunting for patterns/measurements/drawings. I think I'll start another thread on that..... 

No. I have the alu deck, so could probably measure and model from that (I used to do a lot of 3d CAD modelling back in the day and am looking to relearn for my 3d printer.

[Wystan Withers]

If you're able to do the CAD work, I've used a company called CNCMadness in the past to cut CF for me and they do high quality work for very reasonable prices. They're in Canadia so shipping is a thing, but if you're confident in your measurements it's another path for you. I got my machine in no small part because I don't have confidence in my 'hunt and peck' method of measuring a deck by hand with a pair of calipers without being able to do the test runs myself. I'd imagine there's a similar hobbyist/company closer to you if you look into the RC Drone Racing guys... they do all sorts of cool custom stuff.

https://cncmadness.com/

 

[BuggyDad]

3 hours ago, Elbowloh said:

No. I have the alu deck, so could probably measure and model from that (I used to do a lot of 3d CAD modelling back in the day and am looking to relearn for my 3d printer.

Then I guess I may be able to cut it for you in the future. Give me a few weeks/months to get better at this though first. I'm very much a beginner. 

Yeah I think if I was to run off a CNC part for someone else then it would be necessary either for them to do the CAD themselves, so the design/measurements are theirs alone, or for it to be a rerun of something I've done for myself and am happy with. 

[BuggyDad]

@Wystan Withers more questions if I may.....

Are you sticking your work piece down with the masking tape and CA glue method and finding that's OK in a water bath? That's what I did for my test piece and it did work OK, but masking tape and underwater would not normally be happy bedfellows. 

And then... Are you arranging a large number of parts in Fusion 360 manually to make full use of your sheet material? This I've just been doing. but very manually.

I have thought about making my model a little deeper than the sheet material, to make sure I get a full through cut. Is this a thing? Or do you just model a 2mm thick part for a cut in 2mm CF? 

Finally, I'd like to have it cut all the holes before it does the perimeter, because then it'd be cutting the holes with the work piece at its most fixed and solid, but mine did the perimeter first last time. Do you care about this and if so have you found an easy way? 

[Wystan Withers]

 I'm using double sided silicone mounting tape, like the kind you might use to hang a poster on your wall but don't want to leave a mark. Sometimes sold as 'removable' or 'outdoor' Been working through several brands, trying to find something good that isn't as expensive gen-u-wine 3m products, but so far they have been the best. Some of the cheap ones have been a bit 'jelly-like' and allow a little lateral wobble, depends on the substraight they use. It's a mater of finding that happy medium between sticks like mad, underwater, but can still be lifted by just sliding a plastic wedge or scrap styrene sheet.

I 3d printed a L-shaped curb for the rear left corner of the lexan which stands a bit proud, then cut into it a bit with the mill to tram the vertical edge plumb to the movement axes of the table, and so when I place a chunk of CF down it sits squarely each time. That way the long axis of my cuts is as close to perfectly aligned with the weave of the CF as the supplier sent it to me. I trust their cuts more then my own. I also printed some small sacrificial rectangles about 1/2" x 1" I super glue directly to the lexan on the outside edges abutting the edges of the CF to lock it in place laterally so the CF doesn't vibrate too much (the current tape I'm using is a little mushy so these squares really help to lock it in place).

I bore my holes first, then my perimeter, then any internal profiling (for example the two squares that get cut out of the TA02 top plates) as I found that cutting the center out before the perimeter reduced the surface area which was held by the adhesive too much and the part would walk. Cutting the squares out last lets them remain adhered under their own power as it were.

Since my CF ends up 'hovering' about .5 -1mm above the surface of the lexan, I set up my cuts to go about .5mm below the bottom surface of the model to make sure it is cutting all the way through. if you cut into the lexan too much it picks up melted bits of it on the bit and gums up the works.

At the moment, since most of what I'm doing are the TA02 parts so far, their dimensions work out perfectly for me to order a 200x300 sheet and cut it into one strip 50mm wide for the top plate, and 2 100mmx250mm chunks for the lower plates. Mostly this is a function of my smaller work surface. In general though I do each of my files individually that way I can manually place the zero point on each cut, specially when cutting a smaller part out of a larger scrap which might have been a mistake, like cutting a damper tower out of a messed up lower chassis and I need to avoid holes which might have been already cut into the chunk of CF.