Machining Drink Coasters on a $100,000 CNC

[@TheYerko42]

You're right about starting with a hobby CNC where the sandbox is much more forgiving. I bought an X-carve in 2016 and it seemed like a huge complicated machine at the time however I really appreciate all of the lessons I got out of it. I eventually got confident enough to buy a Haas. The learning curve from 0 to X-carve was tough for sure. Going from an X-carve to 3+axis mill was pure joy.

[@bubblewhip382]

Man, I was having this problem this morning on my 100k CNC. Glad this came out in time.

[@iwannaratrod]

I 100% agree that desktop can teach you what the industrial environment may not. Messing around with my SO3 and learning adaptive put me ahead of aerospace programmers that only learned it once they got multi-million dollar machines a few years later. :) Since I was blocked from learning by elder machinists...... my Shapeokos were largely responsible for me getting the job I have now. Being free-er to "eff around and find out" taught me more than 7 years as an aerospace machine operator...... not that I didn't learn a lot there. So thanks Carbide 3D, Winston, Lars, and NYCNC for getting me where I am now. And I'm happy to see Winston where he is now.

[@gizmobowen]

So great to have another video from you. Your videos are entertaining, educational and well done. I really enjoy learning about your CNC journey and am glad you're so generous with your time. Thank you.

[@patregan]

This was awesome, Winston! I enjoyed seeing a CNC more than an order of magnitude more expensive than anything I have used from the perspective of someone who uses machines like mine. I know that you are infinitely more experienced than I am, but you have definitely given me hope that if something like that Datron Neo landed in my garage that I would be able to puzzle out how to get something useful done with it!

[@215Xavier]

Winston absolutely love your content, you taught me 90% I know in cncing. You’ve really made a significant impact on my life! Love seeing your journey, keep it up!

[@cradd00]

Good video! I recently made the jump from a hobby CNC to an industrial CNC router at a new job. I went from operating a CNC shark to a Biesse Rover K 1531 with a vacuum table and automatic loader and and unloading sweeper conveyor. It has a 40,00 rpm spindle with a 10 tool changer and 10 bit drill block. Also learning Biesse B-Solid and nesting software. I'm starting to feel more comfortable now, but wow what a learning curve! It's a dream job, and I'm very lucky to have landed it!

[@Andre-pe9mm]

Only used CNC at work Scm Morbidelli m90 with Maestro software not bad to learn but there are still things I’m stuck with.

[@MaltandMake]

I really enjoyed this format. Useful information and while I doubt I'll buy a Datron it's great to know the reality of using one rather than just 'more expensive, more better'.

[@OhHeyTrevorFlowers]

I enjoy machining videos on a variety of machines but I also like that you cover tools for home machinists.

[@BryanHoward]

I see the collet fan in there. Collet fan for the win!

[@johncheek07]

These videos are always so informative and free of gimmicky clickbait. I wish I could do more than just press "like". I feel bad that I can't always watch them on day one, because I know that gets better ratings.

[@tomwagemans1872]

Just discovered your channel. You explain things just the right way. Subcribed!

[@AudreyRobinel]

Hello there! Nice video, it gives me motivation to tame my desktop CNC :) About your closing question : should you do more of this : my answer is more of anything :)

[@MetalMetz88]

you probably could use the tab method for first op with a full sheet instead of buying blanks, then snip em free and face the backsides off. buy a sheet the whole size of the table and the warp should suck down with all the vacuum ports contacting.

[@penn9363]

I saw the thumbnail and was hoping you were visiting Eddie in Texas! Love the video regardless.

[@xnadave]

Great video - I'm glad you're back!

[@piccilos]

Didn't Henry Ford say something like, "If you need a machine and you don't buy it, eventually you'll find out you'll have have paid for it but don't have it."

[@That0neGeye]

These coasters are the best ever. If you can buy them do it. I love them

[@kw2519]

Better hope he anodizes them, or they’ll be corroded in a month.

[@guntersg]

Lowly, professional hobby machinist... yeah right! You're the go to guy for all things cnc!

[@PedrinbeepOriginal]

More of this please! Is awesome!

[@simcru933]

For small parts you can use their adhesive vacuum paper.

[@WinstonMakes]

That's a good point, I've seen it used before to great effect. Only thing that that I don't like as a cheapskate is that once aluminum chips get on it, it's basically used up. You can cut pieces as needed, but you'll go through it as a consumable much faster. When we machine touch probe bodies here, we leave the same vacucard sheet in place, since the cuts will not change. You can just keep cutting into the vacucard in the exact same spots without compromising vacuum.

[@RemarkXer]

Great video, that answered quite a few questions I had about the neo, but more importantly it exposed something about the nomad. The neo spindle is 17X more powerful and machined the coasters 7X faster, could probably do 10X, but maybe not 17X. It seems that the nomad is very efficient and appropriately powered.

[@BryanO82]

You’ll probably also get more than 250hrs out of the Nomad vs the IMT spindle on the Neo (now allegedly superseded)

[@WinstonMakes]

2000 hours is the supposed expected life of the IMT spindles, not sure about the allegedly superseded spindle... We got more than 500, I know people with 800-1400. There's a 1 year warranty on the spindle, so really the goal is to just keep the machine busy and see if you can find any defects early. And with a machine that expensive, you're losing money if it's idle most of the time anyway.

[@WinstonMakes]

One other factor from an efficiency perspective is the accelerations. I think the Neo is over 20x what the Nomad is set to. This makes a big difference in small adaptive toolpaths, where you can get up to programmed feedrate within a few mm of motion. Keeps your chipload right where you expect.

[@spikeypineapple552]

@@WinstonMakes Bruh you can't defend a spindle that costs 5$/hour to run.

[@automan1223]

Datron wont even reply to any business inquiries unless you send them wads of cash first.

[@nils1953]

That company is quintessential German Business. Good quality but horrid customer support and prices are only revealed after you singed the contract.

[@davidrahn9903]

My dealings with Datron Germany have been great. It's only Datron Dynamics, the American importer and servicer that have been less than stellar to deal with. And even they have been leagues better than DMG and thier resellers.

[@lukewallace3933]

@@davidrahn9903 That was my exact experience too. I had a Neo for a year and it was the worst business decision of my life.

[@jamesm3268]

Glad I went with mikron. Datron would have been nice but the customer service in sales was dog 💩.

[@bytesizedengineering]

I can totally relate to thumbing through the Haas machine configuration page late at night 😅 while also looking at the low balance in my bank account

[@espresso-shot]

I'm sure you can't answer this as an employee of Carbide3D, but it sounds like there's a few options that could make life easier for Nomad users - like an ATC, probe, or mist coolant sprayer. Now that I say it, seems similar to a Carvera, though they don't have nearly the community or customer service record as Carbide3D.

[@WinstonMakes]

Against my better judgment, I'm going to respond to this comment and say that I'd be lying if I didn't sketch out a Nomad with tool holders in my notebook. It's something I think all of us at the company would love to do, but there's a lot to balance and the planning and engineering decisions required are significant. Carvera uses a direct shank clamping system and custom collets, not too dissimilar to the Datron Neo approach. There are pros and cons to this, and it limits how much torque you can transmit. Any dust getting on the tools can also really hurt this, which is why the Neo's magazine is covered. Even then, we had one scary issue of tool pull-out which led to us being more proactive about cleaning the inside of the Neo's collet. PocketNC's V2-50 CHB system has clamping strength issues too which is why they offer the CHK version now. Using tool holders is stronger, but takes up more space. And if you pick one that skews towards being compact, it may be less common and more costly. If you have too few pockets, then you're probably going to want to implement tool groups in software so you can load sets at a time... etc. The requirements start creeping really fast. Basically, there's a really complex decision tree for a feature like this, so we're not rushing into it. I'm sure one day we'll get there though...

[@espresso-shot]

@@WinstonMakes Makes a lot of sense, what seems like a simple feature add turns out to have big implications, safety risks, and commits you to supporting that specific implementation for years - since the investment in tool-holders is quite pricy and you wouldn't want to upset users by changing the design again. I'd also love to see integrated support for advanced touch-probing (though it looks like the community is working on some macros). I imagine updates to the motion control also require a ton of validation and engineering time, and the "free" nature of Carbide Motion means it's harder to commit those resources to its development. A "Carbide Motion Pro" similar to create pro could be a way to get the funding, but subscriptions to "unlock" existing features tend to annoy makers, and bifurcates the development/validation. While you're here.. thoughts on the Penta Solo?

[@WinstonMakes]

So, Carbide Motion has a "quick actions" feature that isn't very well documented. I think it might be possible to build probing macros in there, though I'd need to confer with our lead (only) programmer. Given how our machines (N3, S5P) seem to be leaning towards having an accessible plug for a BitZero probe that makes it easy to disconnect when not needed, plugging in a touch probe might be possible in the future. But until we offer one of our own, and can define a nominal user experience/workflow, it's hard to support a more generalized workflow. I think the Solo is super cool and I'm rooting for the team, but without the final specs and pricing, I'm concerned the case for it will be difficult to justify. I know what the spindle costs (I think they should go Jaeger instead of IMT), and that means it won't cost less than $30,000. There are plenty of small 4-th axis capable mills you can get for that price. And if my worsts price fears come true (closer to $60-70k), I don't know how I'd justify it vs something like a Haas CM-1 w/ trunnion. I really like the engineering I'm seeing and I hope they can make a competitive machine that stands out without leaning too hard on a niche market, but I'm also a little worried for them.

[@davewithhemp]

Awesome work. Curious to know if you've tried these coasters on a Shapeoko 3.

[@WinstonMakes]

Not yet, just the Datron Neo and Nomad. Once I get my modded Shapeoko 3 back together, I might run some coasters at home.

[@mofosheee01]

Hello Winson.........would you please point me in the direction where you acquired your alcohol mist system. By the way nice work!

[@WinstonMakes]

The one in this video is built into the machine. You can buy cheap versions online though, like Amazon. But they are not pressurized, so you need to constantly fiddle with valves to get the desired quantity of mist. For most CNC routers though, simply a steady stream of air will help immensely.

[@xmanual]

Very well presented. How is the alcohol Vapor dealt with to avoid risk of fire?

[@WinstonMakes]

The enclosure is exhausted via fans at the top, and the consumption of alcohol is quite low, all things considered. It can be adjusted so that it's less than a mist system where droplets have a chance to collect. The use of the ethanol coolant system is also not recommended for ferrous metals (risk of sparks), so that kind of work would need to be done dry.

[@myk3l9675]

Hey, hes back!

[@statorbinhson9031]

hi what software do you use for Datron?

[@legionjames1822]

workholding tape nice.

[@nikita.gazarov]

I feel guilty asking about the smallest little thing in this industrial-cnc video, but how do you like your bit / collet fan on the nomad? I'm wondering if it's a good replacement for air blast when milling aluminum on shapeokos. Is there an STL file somewhere I could try with carbide's router?

[@WinstonMakes]

The fan works good for pockets and relatively shallow cuts (1-2x cutter diameter) but it won't effectively get chips out of deep slots well. That requires compressed air. But it's way better than nothing. The chip fan here ships with the Nomad, wouldn't work with a router though because the collet size is different. I think you'd probably have to design your own, not sure if one exists.

[@Thys00nPL]

+1 for Hans Zimmer on apple watch :P

[@PCBWay]

Impressive! 👍👍👍

[@angrydragonslayer]

Man, i love these videos It just feels so weird when someone says 2 kw is their max spindle power The millturn i'm in front of right now gets jittery at or below that wattage with tools weighing less than 1.5 kg in the milling head :)

[@angrydragonslayer]

Jittery as in the PID derivative over/under-shooting the RPM goal, making the RPM vary 3-4 RPM

[@Duraltia]

Kinda the Apple of CNC Machines, huh? 😁 Maybe I should consider making a DIY Vacuum Holding Solution like that for my X-Carve 🤔

[@WinstonMakes]

Vacuum table on any CNC router is a lot of fun. But unless you use a gasketted fixture, you're going to need a beefy pump that's going to run up your electric bill quite a bit.

[@maadmaxx123]

I had the luck of selecting a CNC machine for work. As much as i wanted the Neo but even though I could have probably fought for it and got this machine, it wasn't justifiable for me. Simply too expensive.

[@graveurgraveur2691]

Thanks!

[@spikeypineapple552]

Absolutely insane to me as a professional that you left the outside as a lasercut finish and used a plastic template on a professional machine, to save like 50p. Rather than using the machine the way it was designed, by slapping down a piece of 400x500 and getting at least 12 off at a time(costing under £4 per piece in material)

[@ScuffedEngineer]

Why would someone buy this over say the HAAS CM-1? For less it has higher capability with probing, 36 tool capacity on the ATC, 50k rpm, 32 or 64GB of memory, larger tools, faster and or deeper cuts can be made. Or it can be significantly cheaper with more torque allowing for again, faster and deeper cuts and better for the larger tooling. As far as the vacuum table being nice for it's ease, you can use superglue and painters tape for small parts and toe clamps and a fixture plate for large parts. The HAAS would also give you the ability of 4th and 5th axis if you only need smaller parts and probably still be in the $100k range.

[@russellsmith5278]

.......wait - nomad 4 - 22:44 - when is it out?

[@WinstonMakes]

"Nomad *for*..." But I appreciate the enthusiasm and excitement. If I were considering ideas to improve the Nomad though, what would you want to see change from the Nomad 3?

[@russellsmith5278]

@@WinstonMakes I'm a hobbyist, enjoying your content - I'm currently using a Genmistu 4030; I'd like to upgrade to something in the

[@WinstonMakes]

A lot of the pre-built desktop machines are going to be a step down in terms of working area compared to that (ex. Nomad, Evo, ETC) but you'd get more rigidity and potentially some quality of life stuff like tool length probe. Are you beholden to a particular size or form factor? Because in the sub $6k range, you can get bigger ball-screw driven,gantry style machines with more capability all around.

[@russellsmith5278]

@@WinstonMakes currently no - and, I think the bigger stuff I want to (but haven't) make is mostly out of wood. I'm likely to keep & expand the 4030 for that, and get something better at doing smaller aluminum things. I think Nomad would be a good size; the Evo is "better" workspace wise but is obviously $5k not $3k. Not sure what else to look at

[@WinstonMakes]

I think Penta/PocketNC is a good company, the people there have certainly been very responsive to questions and support. But the V2 is a machine that's hard to justify unless you already have ideas for things to make with it. And ideally some of those projects would generate revenue to help pay for the machine too. The small CNC market's in a weird place right now with supply chain and inflation right now, the scales we're dealing with mean labor and material costs have a significant impact on overall margins. Hard to point you in a particular direction. My mind did jump to the Tormach 440, but I just looked on the site and the prices shocked me. Same with a rooster company I hesitate to name. If you're not in a rush, I'd say have a little patience, and see how the market evolves over the next few quarters...

[@subthousandoaks]

Ty 👍

[@andychrest8876]

Why not make coasters on a 220v HDM?

[@WinstonMakes]

It would be a better experience than the Nomad for sure, but I'd still have to deal with double sided tape and tool changes.

[@toddmolloy311]

@@WinstonMakes Introduction to palletizing or makeing a pallet system could help people out. Utilizing tiger clamps maby setting up pallet's whyle tending to the machine? I notice you're videos introducing more of the industrial work flow do well. Take it easy, look forward to the next one 👋

[@AlwaysCensored-xp1be]

Only ten times faster at hundred times the price, mind you quality is ten times better? Just goes to show how cheap and capable hobby machines are now. Great beginner things

[@kylewarren69]

first ATCs spoil you and you can never go back to a machine without them, then you get a machine with a pallet changer and every non pallet changer machine is just annoying.

[@bobweiram6321]

I think you're better off metal casting for volume.

[@NorroTaku]

this sounds like it's really really not worth investing in something with the same flaws (work holding, work holding and power) as a desk top cnc that alcohol cooling sounds neat tho but why isn't it industry standard if it's so cool? it can't be cost machining is very tolerant of one off cost is that stuff significantly more expensive then good old oil and water?

[@WinstonMakes]

You can't use MQL like this on ferrous metals or anything that can spark. So the Datron really excels at aluminum, and micromachining anything else where you wouldn't need flood coolant to begin with.

[@tommays56]

As soon as I see all the OPEN ballscrews it’s over for me

[@Stephen8454]

Yup... I have been there lol

[@evanbarnes9984]

Awesome overview! I really don't see why someone would go for this tool over something like a Tormach 770, though, since the Tormach is like 10% the price. It just seems like the vacuum work holding doesn't offer a huge advantage and has quite a few pitfalls, whereas traditional mill work holding is very reliable. I guess like you said, it really seems aimed at small prototyping of high precision parts. Super cool either way! Have you considered diving into a PrintNC project? It could be a cool step up to a larger form factor that doesn't cost $100k

[@WinstonMakes]

For the majority of machinists, I agree that a Tormach, or even a small Haas type VMC would be a logical first step for someone venturing out into job shop type work. The biggest advantage of the vacuum workholding IMO is how it reduces setup time to near zero. For prototyping, where you have to switch between parts rapidly, you just clear the table, slap down material. And probe it in. If you need an Op 2, maybe do soft jaws on a pallet. Oh, mold work is another application where the spindle speed really helps, though I've also seen a fair number of air spindles on VMCs on Instagram. I think the PrintNC stuff is pretty cool, but because I have access to all the spare Shapeoko parts I want basically, my free CNC design time mostly goes into thinking about how to repurpose legacy parts into a "Franken-oko" and designing for manufacturing in case there's a market for older owners to upgrade.

[@poetac15]

Comparing a Tormach to this machine is apples to oranges. Completely different level of accuracy, rigidity, reliability, etc.

[@SPDLTD]

At first i was thinking the same thing, but if this machine truly is in the Micron level of accuracy it's a totally different beast from a Tormach or normal VMC. This would be excellent for very precise mold work done with very small tooling or machining very small parts (think things like components inside a Rolex). That being said, this is a very niche market, most machining doesn't need anywhere near this level of accuracy, nor is this accuracy even relevant or achievable on many things you would machine just from material de-stressing or thermal expansion from handling. I'm curious how much this machine weighs

[@WinstonMakes]

@@SPDLTD about 1500 lbs. The Y axis is built into an epoxy granite base.

[@evanbarnes9984]

@@WinstonMakes Oh that's right! I totally forgot you have all the Shapeoko parts you need! Man, what a cool resource. Well, I can't wait to see what you do to make a Frankoko

[@kw2519]

That’s a pretty expensive router 😂

[@harriesadam]

"A higher-order programming language like C" Quiet everyone, don't tell him about functional programming.

[@WinstonMakes]

I'm on the cusp of dating myself by saying one of my engineering professors made my class learn FORTRAN...

[@harriesadam]

@@WinstonMakes don’t worry, FORTRAN is still concerningly widely used!

[@gaiustacitus4242]

Why spend $100,000+ on a CNC when you can build a very capable DIY CNC that can do the job for less than $5,000 or go wild with features and spend up to $10,000?

[@3d8d]

Isn't that 100,000 CNC is quite small 🙂

[@WinstonMakes]

In the grand scheme of things, yes. But it's also much more accurate than our other machines. More accurate than we need to be honest. It's a tradeoff. But the versatility of the workholding is also a big plus.

[@st3phanb]

More vids

[@davebergstrom1819]

'Promo sm' ☹️

[@Thewaldo12345]

Aluminum is not the best choice for a coaster. Gonna make hot drinks cold and cold drinks not so cold.

[@WinstonMakes]

This is true. But the motivation/compromise was that I wanted an aerospace-style aluminum isogrid. And the negative thermal characteristics are mitigated by minimizing the heatsink effect/mass (by machining it into an isogrid) and putting cork on the bottom to prevent additional conduction losses from coaster to table.

[@Thewaldo12345]

@@WinstonMakes I’m sure it’ll be fine. I just repeated something I saw on Titans video lol.

[@Jessersadler]

aluminum coasters? Doesn't make sense

[@ATXAdventure]

This datron neo sounds terrible. The work holding is a nightmare, and the spindle is underpowered.

[@daynosdr]

that little piece of junk its 100k?!!?