The Home Manufacturing Thread

In the Industries Ripe for Disruption thread, I mentioned manufacturing, and the future of distributed, bespoke manufacturing at home.

I said,

I thought this would be a good thread to maintain to discuss issues around this including 3D printing, CNC machining at home, CAD/CAM design at home, programming and electronics design, services for producing products, etc. Pretty much anything involved with designing and creating products as a small or one-person shop with low-cost manufacturing devices.

In the other thread, this question came up:

It depends on your previous knowledge level, I guess. 3D printing is easy, but finicky. If you buy a decent printer (a Creality Ender 3 is a good and very cheap printer), you can be up and printing the aame day. There are thousands of 3D models you can download and print from places like Thingiverse.

If you want to do your own modelling, there are easy and hard programs out there. I decided to bite the bullet and learn Fusion 360. That does have a steep learning curve, unless you are already familiar with parametric modelling. I wasn’t, so it took me some time to get my head around it all. Luckily, there are lots of tutorials out there. But Fusion 360 is awesome. You start with sketches, it renders thm as 3D, you can build components and put them together, and then if you are happy you can either output files directly to your printer or CNC machine, or you can go the full CAM route and run stress simulations, range of motion tests for components that have joints, etc.

Since I’m now quasi-retired, I decided a good way to get my skills back up to speed and learn new tech would be to try to build a product all the way from concept to final prototype. So I also did some electronics work.

The product I decided to build was a lamp that is also an Internet of Things hub. To make all that happen I bought some LED strips, an arduino-compatible Wemos D1 clone using an ESP32 processor, and a few chips like a BME280 temp/pressure/humidity sensor, an NR24L01 radio chip, and a little Buck converter power supply. Total cost of all these parts was about $30.

Programming the processor was done in Microsoft Visual Studio in C++. If you aren’t a programmer, a much easier way to go is to use the Arduino IDE. Programming is really easy for these things, and you could get this whole setup running with about 2 pages of code. But if you aren’t a programmer and have never done any code, there’s definitely a learning curve there.

For the electronics, I breadboarded them first, then drew up the schematic. You can simply send the schematic to a service and they will send you back finished PC boards ready for soldering for just a couple of bucks each. There are even services out there that will build your PC board out, for about $20-$30 each plus parts if you don’t want to do the work. Or, you can find manufacturers who will build your entire product, design packaging, make the packaging, and send you however many you want, ready to ship.

Or, after you’ve developed your device, they will simply do everything else for you, including selling it to wholesalers, but you’ll only get a small cut of each unit.

The last home thing I’m trying is to use the CNC machine and copper-plated boards to actually carve the traces on a PC board so I can go from conception to an actual commercial-ready product just with the stuff I have at home.

This is the CNC router I got:

That also has a 5.5W laser that can cut plastic and wood and engrave all kinds of things.

The printer:
https://www.amazon.com/Creality-Printers-Motherboard-Carborundum-220x220x250mm/dp/B08BLFB9D4/ref=sr_1_1_sspa?crid=25MNEN9BDYGVN&dchild=1&keywords=creality+ender+3+v2&qid=1618452873&sprefix=Creality+Ender+3%2Caps%2C208&sr=8-1-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUEyODVOQkFGUFZJMFRIJmVuY3J5cHRlZElkPUEwODY1OTE3MzcxVE5RM0lURFhSVSZlbmNyeXB0ZWRBZElkPUExMDI1NDExMTVGSzA2NkJCOFQzMyZ3aWRnZXROYW1lPXNwX2F0ZiZhY3Rpb249Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=

ESP32 Processor:
https://www.amazon.com/Development-NodeMcu-ESP-WROOM-32-Internet-Compatible/dp/B08R885LN5/ref=sr_1_1_sspa?dchild=1&keywords=wemos+d1+mini+esp32&qid=1618452954&sr=8-1-spons&psc=1&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUFBV1YzTzFBQ0lHVUgmZW5jcnlwdGVkSWQ9QTA2ODM4MzExOUZVTktZU0lUMjJMJmVuY3J5cHRlZEFkSWQ9QTA2NDg0NzYxQ1VZTEFVQ1JJWVlUJndpZGdldE5hbWU9c3BfYXRmJmFjdGlvbj1jbGlja1JlZGlyZWN0JmRvTm90TG9nQ2xpY2s9dHJ1ZQ==

To be honest, I was shocked at how easily this stuff goes together. I had the processor running everything, a web server to control it, and a web service I could call and get the weather from the BME280 chip going, all in a couple of evenings. When I was doing electronics a couple of decades ago, this would have required a team of people to build and cost many thousands of dollars.

If you want to make something out of metal or an exotic material, you can design it at home, print out plastic versions to test things like fit, and then send the file off to a service like Shapeways and they will return it in the material of your choice for very reasonable money.

By the way, those little ESP32 processors have bluetooth, Wifi, a full TCP-IP stack, and a huge number of inputs and outputs. Amazing little things.

If you have any other questions, ask away.

Not to rain on your parade, but you do know that virtually none of this stuff meets safety standards, correct? 3D plastic parts generally fail to meet hazardous chemical requirements, fire tests, etc. It’s all fun and games until your house burns down because the 3D printed part catches fire…

you get what you pay for. My company spends a boatload of money on testing labs and to get certifications to be compliant with country regulatory laws. Avoid those at your own risk.

I’m fully aware of that, thanks. My project is an LED lamp powered by a CSA approved wall wart or a USB port. It’s all low powered.

As for proper engineering, there’s nothing about this that requires poor engineering. Plenty of professional engineers work at home building products - especially electrical engineers. Professionalism includes knowing which materials to ise for which applications, and the limits and strengths of 3D printing.

In addition, I was talking mostly about using all this for prototyping. These methods are way too slow and finicky for any kind of production process.

The services Immentioned earlier have engineera whomwill modify your protype for production and ensure that it meets whatever standards are required before they build the actual product.

That whole story is totally cool / amazing.

Not trying to pile on after @China_Guy, but this snip from your post in the other thread got me to thinking when I first read it there a couple days ago:

I’m wondering how you, or that shop, determined what material to make that part out of? And what sorts of treatments are needed on it in addition to it’s bulk outline? Surface hardening, heat treatments, annealing, etc.

It’s been years now, but I too have had to scour junkyards to find undamaged suspension components for older model cars. The ability to build a new part correctly would be huge. But the gotcha is the “correctly”. Compared to, say, the 1960s, a lot of parts of modern machines are more tightly engineered; it’s no longer Soviet-style cast iron “over-engineering by the pound.”

If the 3D manufacturer had access to the original parts’ complete design and production documents, or to detailed structural strength specs they could, at least in principle, produce a fully equivalent copy. Absent those things they’d be guessing about everything except the gross shape and the general sort of material (e.g. plastic vs aluminum vs steel but no better than that).

Or in the case of fly-by-night operators, making stuff out of the high tech equivalent of pot metal. And planning to be gone (or stay safely in e.g. China) before the inevitable failures occur.


I totally get that you aren’t planning to go into the business of making suspension components for out-of-production cars. But somebody is/did. Do you know anything about how they or their peers square this particular circle?

Is your home zoned for light manufacturing? Especially if you try to run a business and make it onsite. Or do you need to have a shop in the correct area?

My understanding was that the part came from China, and was spec’d properly. It wasn’t 3D printed. Possibly cnc machined from raw stock. It was something like an idler arm. The point was it was a bespoke item custom made using one-off manufacturing techniques that are koving down in price to the level where very small businesses can domit.

I’m familiar with the engineering issues with 3D printing. Inconsistency between parts, layer adhesion, etc. I worked in manufacturing improvement, OEE, etc.

Once again… the use of 3D printing at home is about prototyping, not kicking out manufactured parts for high stress situations. Or if you are selling things you make at home they are more likely to be things like figurines, or cosplay accessories, or toys, or something like that.

And let’s not get hung up on 3D printing. There are other home manufacturing methods, such as small six axis CNC mills that can mill parts out billet material, or home laser cutters that can make intricate small parts out of many materials.

And if I mill a part at home out of a block of 6061 aluminum, it’s going to be as good as any other part milled from 6061 aluminum.

Making prototypes on a 3d printer is not ‘light manufacturing’. If you are worried about your own situation, by all means knock yourself out and make sure you are completely compliant with local zoning laws and regulations before settinng up your desktop printer.

Thank you. Makes sense.

I really wasn’t trying to rain on the larger parade. You’re absolutely right to be stoked about the possibilities this opens up.

Totally agree. CNC machining is great if you’ve got the right specs and the right materials.

3D printing depends on the use. Even prototypes. They may look fine but contain banned chemicals in the US, or worse catch on fire and act as an accelerant. Will the toys and figurines be safe for home use. As in will it be safe if a child actually swallows one? Go back in time and lead figurines were discontinued because they could be deadly if swallowed.

Most desktop 3D printers, like the OP’s Creality, use PLA or ABS or PET or TPU - none of those are banned in the USA.

I’m hoping to spend most of June playing around with a 3018 + laser to see how much use I can get out of it with leatherwork, including laser-engraving the leather, as well as cutting out plastic embossing plates and wooden wet-forming moulds. So I guess my current comment is - watch this space.

If I could easily get one that could carve aluminium or brass, I would, but that seems beyond the spindle limits of what I can get locally.

I’m not so sure doing it at home will become a huge thing for most people- most people aren’t inclined to ix their own stuff anyway. What will happen though is that the 3D printing and machining equipment will get to the price point and ease of use point that your appliance service company will have some, and rather than order parts for your 1987 Speed Queen from some dusty warehouse in rural Maine, they’ll just print/machine one out for you and have their guy install it. Same thing will happen on a larger scale- if you want to do it yourself, you’ll be able to go to a storefront, and some guy will make you one and you’ll pick it up the next day.

Basically it’s going to replace some portion of parts and items that aren’t too complicated and that otherwise would need to be built or fabricated elsewhere. Where it chokes, IMO, is that a lot of parts still need finishing, assembly, etc… and for those, I’m not sure localized manufacturing is going to be better than doing it in some factory where labor’s cheap and quality control is institutionalized and high quality.

I see there being a real need for verification methods for the plans/templates used to print parts and the materials used for that. Nobody wants their car’s suspension to fail because the parts store was using a janky template to print a bracket or whatever. Or that your oven catches fire because the guy who made whatever gizmo at the appliance repair store used substandard materials to make your part.

Check this thing out:

https://pocketnc.com/

Too much money for a hobbyist at around $5,000 bucks, and it only produces small parts, but it’s incredibly accurate. One of their demos is a miniature CNC-milled complete aluminum engine block from a piece of stock.

While it’s too expensive for me as a current hobbyist, if I were setting up a small home business to make custom parts for people, $5,000 is nothing.

In a space as small as a bedroom you could build out a complete design/prototyping/small manufacturing shop. A 3D printer for plastics and prototypes, A mill like that for small parts, and you send out to services for anything you can’t build at home.

What excites me about all this is the effect it could have on innovation. The rate of innovation is largely a function of how many different and diverse people are exploring the ‘adjacent possible’. Giving potentially millions of people the power to design and build anything they want without major resources or management approvals will accelerate innovation in hardware just as cheap ooen source software has accelerated software and enabled an entire industry of work-at-home developers and distributed teams of amateur developers working on open source projects.

Add to that the new generation of microcontrollers, sensors, and support hardware that are available dirt cheap and easy to work with, and the availability of professional CAD/CAM software at hobbyist prices, and individuals are being empowered to create what they want with very little money (comparatively) and little space.

Yeah, the properties of filament are becoming well understood. PLA is biodegradeanle and food safe. ABS is like any other ABS. PETG is also food safe. The filaments thst give off hazardous fumes when prjnting are well documented. There are plenty of engineering studies looking at how to validate 3D prjnted parts for strength and consistency, and there are filaments with embedded carbon fiber for great strength.

NASA has been testing 3D printed rocket nozzles and combustion chambers:

Not to be outdone, Relativity Space is 3D printing an entire orbital rocket:

And SpaceX’s SuperDraco engines use 3D printed combustion chambers printed in inconel.

Keep us informed! I have yet to use my 3018. I bought an upgraded spindle for it, but haven’t installed it yet. Once I finish the current ohase of prototyping, I’m going to try physically etching a PC board. It sounds finicky, and you need to do a mesh map of the surface height, somI bought a cheap probe for that,

It would be nice to share notes with someone else playing around with a 3018 machine,

Lots of videos on youtube about leather work and the 3018, as I’m sure you already know,

Oh, it will never be lots of people… It’s technical, finicky work. But it just has to be accessable to hobbyists - Hams, R/C hobbyists, train hobbyists, people who like tinkering with hardware, craftsmen, etc. I have a friend who isn’t technically trained but he’s a good backyard mechanic, and he’s now got a 3D printer and is looking at a 3018 and is absolutely stoked for the possibilities it opens up for him.

The kind of people who design decorations or small useful home items for sale on places like Etsy are also going full steam into this technology.

Yep. Thst’s already happening. But it’s more than that. People have uploaded thousands of designs for replacement parts for various appliances that ate appropriate for 3D printing. Break a knob on your stove or a plastic clip for your dishwasher valence panel or something? Look up the part, send it to yoyr printer and you have a new one in a few hours.

Quality is totally up to the person. Where 3D printing falls down is mostly in volume - it’s a very slow process. The lamp I am building takes a total of about 30 hours to print one. I have seen some people who have set up dozens of printers to make products for sale, but that strikes me as the wrong use for the technology. If you need to make many parts, do the prototype on a 3D printer, then once you are happy with it get a service to manufacture them in whatever bulk you need. That too is now surprisingly cheap and easy to do.

Not just verifying the design, but the actual product. 3D printing on cheap printers still has issues with repeatability. You need to be able to validate that each part you make fits within the specs of the design both dimensionally and in whatever strength parameters you need. so at this point I would not try building parts for safety-critical applications, or where part failure could cause larger failures.

Current markets for home 3D printed producrs include things like parts for hobby drones and airplanes, small figurines for wargaming, Cosplay accessories, replacing small broken parts that are hard to get, etc. I believe Jay Leno has a metal 3D printer and a waterjet and CNC mill, and uses them to replace parts that are absolutely impossible to get and would cost a fortune to have made traditionally, like replacement pump parts for his Doble steam car.

I did not, actually, thanks. The leatherworkers I follow haven’t done anything yet. I took a look now, and see people actually cutting out parts to sew with the router bit, including the stitch holes. I hadn’t considered that before, I probably still wouldn’t go that far, round holes aren’t the same as diamond awl holes.

Most of the videos are about using the laser, hut there are a few that show using the carving bits on leather as well.

If this is say an electronics case, would it pass a flame test? At work, I had prototypes 3D printed, and then limited to only use in a safety lab after watching the flame test. Flame test is effectively, a butane flame on the device for 10 seconds. Then if it keeps burning and the flame gets bigger, it fails the test.

Parametric modeling really is amazing, incidentally. I got my start with SolidWorks and switched to Fusion 360 since it was free. But regardless, grokking parametric modeling turned it from an immensely fiddly process into something that more resembled software development. Change a constant like wall thickness, and it takes zero effort if you’ve done things correctly. All the constraints like centerlines and tangents and so on just automatically do the right thing. Complete game-changer for me.