Hobby CAD, CNC machining, and resin casting (2015) (opens in new tab)

Right. This article is (2015).

3D printing with PLA has improved in the intervening decade. You can usually get a good print on good modern printers. The first generation of those things had poor extruders, and filament formulation has reportedly improved. There's complicated heat transfer going on in those things. You're welding a hot thing to a cold thing, which is inherently troublesome. I'm told that works better now.

Machining resin molds is straightforward, because you start with a block of something and machine only its top. So there's no work-holding problem. Trying to figure out how to clamp something that needs to be machined on several sides is usually hard.

Not sure what's going on in tiny mills today. I've used Tormachs, the whole range of Shopbots, and some strange one-off machines that TechShop somehow obtained. (Never did use the beautiful little Pocket 5-axis machine. TheShop had one just before they went bankrupt.)

These also had discussions:

https://news.ycombinator.com/item?id=4679939

https://news.ycombinator.com/item?id=34339459

"GatorCAM for CNC"

Instructions to download: https://youtu.be/hqyPzCKGUQc?si=VJ0KRhQOl_-d7Bmm

Download here: https://sites.google.com/view/gatorcam/home

Very user friendly. Tab support, v-bit, ATC, sorts toolpaths for faster carving.

awesome dump, thanks for sharing. solvespace is cool but until they iron out the chamfer/fillet situation and parametric input I ll have to continue using freecad..

> It is not comparable with 3d printing at all.

Not just not comparable, but totally opposite. Subtractive vs Additive. CNC is like chiseling the David out of single block of stone by removing bits. 3d printing is starting with nothing and adding material.

It is definitely more expensive than 3d printing in terms of consumables but china/aliexpress has really good hobby-quality equipment now - if you don't mind the country of origin. Many of their coated end-mills are decent and < $5 depending on diameter.

Maybe not stuff you would want to run professionally but it works quite well in a hobbyist setting and a mistake won't cost you $50.

I also looked into the Millenium Milo, only downside for me and kinda why I decided to design my own Voron V0-ish sized mini CNC machine was the really large footprint for about 3x the work volume (work _area_ is ~2x afair). An enclosed Milo would take up about the same volume as two 350mm Voron's side-by-side. So footprint of a small desk.

Imo besides the price, the other big factor is just how much less forgiving CNC machining is than 3d printing - so many mistakes you can make, zeroing the WCS, wrong WCS, mounting the work different than you had it in CAM, ... bam, at least the work is ruined. That's kind of another longer-term goal with my CNC machine, reducing some of these errors if possible with a web-based UI and maybe some computer vision. But that's far off, I'm currently playing with using a camera for work-probing/WCS-zeroing and it's sloooow progress :')

I'd just like to point out how stereotypical of a modern mechanical engineer it is to want to print everything out of plastic and not appreciate the greater versatility of a CNC router.

I can't count the number of times I've converted gasketed joints into O-ringed joints, slotted holes, casting lines and welds I've smoothed out. Yes, there are better machines for performing these tasks in metal but I can just tell the CNC router to go nice and slow and come back hours later. In addition to that there's all buffing and polishing operations you can do with abrasives and the ability to cut cheap wood templates and test parts. The 3d printer can make complex things relatively quickly and easily but it can't work with existing things.

Any time I see a guide like this, everyone will go on and on about the types of cutters, toolpaths, speeds & feeds, etc. but will always gloss over work holding. For every action, there is an equal and opposite reaction. Without reaction forces, the best cutter in the world is useless. A sizeable chunk of your tooling will just be clamps and indexing tools needed to align and secure the thing you're CNCing, while also trying to keep them out of the work area. 80% of the end mills I've ever broken have been by crashing them into a clamp or vice by accident. This is another thing you tend not to see in 3D printing or laser cutting, or plasma cutting for that matter.

There's a much higher barrier to entry with CNC. I have a cheapo CNC - few hundred dollars - and really have only ever made one practical usable "thing" on it.

It's so slow and needs much trial and error to get a decent cut, I feel a more expensive machine would be much more versatile but cost 10x for the machine and consumables as you say.

With the similarly priced 3d printer I feel like I've actually saved a lot of money and hassle printing things. I can quickly download a model and print a perfectly useable "thing" in no time, often with no mistakes. Prusa slicer and a larger $5 nozzle has made things even better, even though I have an off brand 6 year old printer.

Lots of hobbies can get expensive though, especially when high quality tools are involved. I'd actually argue 3d printing is an exception to that.

Decent power tools and consumables rack up and don't even look at the price of hand tools from the likes of Lie Nielsen.

A reasonable quality plasma-cutter CNC table offers the speed and low operational cost most fabricators can manage well.

Having a CNC Router table in a non-industrial zoned area will not work for most people. Services like sendcutsend.com makes life so much easier... =3

The Millennium Machines Milo does look interesting. I wish there was an inexpensive cast iron or epoxy granite equivalent to aluminum extrusion.

Have you tried metal 3D printing? Is that a thing for home hobbyists?

Have you looked at Shapeways or sendcutsend or xometry or 100k Garages?

That said, I have an especial interest in archery and multiple machines.

Contact me at: https://community.carbide3d.com/u/willadams and we can at least work out doing a couple of prototypes.

Find a hobby machinist.

People who aren't trying to expand their business don't want to deal in non-gravy projects from customers who don't seem likely to shovel them a bunch of gravy projects later.

Machining is a slowly dying industry in the west so there's far more people not growing their business at any one time.

Do you have a complete technical drawing?

My guess form rejection is either:

1. No drawing and vague specs.

2. Too time consuming and they don't want to quote the crazy price it'd take.

Machine shop time is very expensive.

Fully enclosed with a chip vac is good. Chips all over the place is no fun. Especially with coolant.

Don't expect people to precision-cut wood for the frame. The Liteplacer people tried that for their pick and place machine, and most people never got a working machine. If it needs plates with holes in them, make them in bulk and sell them. Waterjets are good for that. The holes will be where they are supposed to be.

(The Liteplacer was a really good idea - a pick and place machine for assembling prototype PC boards. Camera controlled, with the input parts in partitioned trays rather than reels, it was slow but did the job precisely. The PixiePlacer seems to be the next generation of this. But, as with the Liteplacer, you can't just order the metal parts. You have to make them or have them made. There are commercial machines, of course, but they're for production, feed parts from reels, and are more expensive.)

I uploaded a quick video of it cutting some alu recently if anyone's interested:

https://youtube.com/shorts/XUsj06iMbb0

What's the advantage of this over a 3018 or 3030 machine? With some basic upgrades (you mentioned needing to diy anyway) you can easily cut aluminum on those for $500 or less.

This would be a lot more interesting if it could cut steel and titanium. I guess that is more difficult.

Added: a dumb question, if the main cutting device is a router with a spinning bit, how do you cut angles? One thing I'd like to make is a 4mm hexagonal hole in a piece of steel, to turn little hex drive screwdriver bits and the like. Is it even possible to make that with a mill, or do you need a different machine?

that looks like a nice project, reminds me of the ghost gunner 3 machine. I like the safety note too. might hop into discord sometime next week. thanks for sharing!

You may also like lcamtuf's current thingy, a substack

https://lcamtuf.substack.com/

Flashlight body would probably want a lathe, but one of the newer flat ones would be easy in two halves.

For an example of a knife blade:

https://www.youtube.com/watch?v=G5TYla5x-Nk

small tool:

https://community.carbide3d.com/t/nomad-made-custom-titanium...

Current hobby 3d printing of metals usually requires a clay-metal type of filament and sintering, usually by heat or laser to harden the metal, and often needing compensation for shrinking and a lot of other inconveniences.

The resultant pieces are "metal" but not strong enough for most internal firearms parts that need to be metal but not necessarily pressure-bearing like a barrel: most notably bolts and carrier groups, fire control group parts like triggers, hammers, sears, disconnectors.

You can't easily make a high-quality firearm a hobby mill, chiefly because the barrel needs to be made out of good steel and needs to be rifled. But the quirk in the US is that federally, only the receiver is the regulated part, and many types of receivers can be made out of plastic or aluminum. You can certainly use cheap three-axis CNC with some fixturing to make AR-15 receivers, for example, and many people did. Cody Wilson / Defense Distributed had this whole thing where they were selling CNC mills for cranking out guns.

You can also definitely make junk single-use guns using either technology, just like the 3D-printable "Liberator".

You can do it but don't expect top notch results.

Routers don't have enough rigidity (this is why you don't see gantry mills except in huge sizes) for the kind of work you'll want to do. So much of your stuff will be one off that you're not going to have any speed advantage over a guy with a Bridgeport.

The most likely successor would be:

https://shapeokoenthusiasts.gitbook.io/shapeoko-cnc-a-to-z

(ob. discl., I work for Carbide 3D)

For one thing, IME with cheap hobbyist 3d printers (e.g., Ender printing PLA), the article is dead on about accuracy... people say 3d printers are good for 0.1mm but I don't know how long they spend tuning the machine. I mean, they are... kind of... just not reliably. CNC accuracy of 0.025mm is much easier to achieve (IME).

If you mean to tell me that I should just have a resin printer to start with, well I agree but I think we should be specific about what kind of 3D printing we are discussing.

Remember, the guide was written over a decade ago. Certainly 3d printing has gotten better, cheaper, and a lot more reliable. Resin is really strong, though.

It is possible to get good results with 3D printers, but that requires you to go to below than 0.1mm layer heights with 0.1mm nozzles and nobody has time for that. You can make a hundred resin parts in the time it takes for the printer to finish making one.

It depends on the accuracy you are trying to achieve, and with what technology. If you mean FDM, it's absolutely true. If you mean SLA/MSLA these days, i think you could get there.

On the first - i have an entirely ballscrew + linear rails driven FDM printer with closed loop servos and proper absolute encoders - I got bored and do lots of CNC retrofits/etc, so had lots of parts around. Think of it as insane version of the Pantheon HS30, if you didn't care at all about costs or practicality.

It can position, at speed (IE 300mm/s+), and repeat, the nozzle placement to within 0.0005 inches on all axes, easy.

But even with input shaping/etc, you will not get a better surface finish than CNC, and definitely not one good enough for casting.

Filaments are just too finicky, even with really really good hotends, sensors everywhere to monitor response and optimize, etc.

Now, i did this for giggles, and yeah, i never get misplaced layers, my parts are consistent, etc.

But 0.4mm is still a pretty big feature size, and that's the minimum if i don't want prints to take until the heat death of the sun.

So I could still machine the same thing on my metal mill and achieve castable surface finish, etc, in much less time and effort.

That was written before resin printers were affordably available or input shaping and so forth had made such a leap in print quality for fused filament printers --- that said, I get much better surface finishes on my CNC machines than my 3D printers, and casting will reproduce even the tiniest imperfection.