Stephen, you can drop about $100 on a good HVLP sprayer, and some spray-able adhesive. that'll reduce the number of cans and time to coat for the foam. also invest in a good breathing mask and some coveralls

I would *love* to have the files for that belt winder, if you don't mind sharing them somewhere!!

You can add one more roller after 350$ cutter which can also wind it. If it is time consuming to add it electronically, just use mechanically, two GT2 pulley in a ratio to synchronise winder with cutter roller. Winder roller can be should removable(may be gear driven so that winder pulley stays in place). Another youtuber who makes this kind of cutting & winding machines. kzfaq.info/get/bejne/oa2pqNaUsJ-rkXU.html kzfaq.info/get/bejne/jL1gaqmSrtzQj5c.html kzfaq.info/get/bejne/gNJmjaSFv9_UcYU.html kzfaq.info/get/bejne/kNSVdct9u9nTlHU.html

These seem great. Thanks! I wonder if the Lumen or something like it could use those to automate screw placement and fastening?

@@nidavis there are definitely robots that are built to assemble things and use screw presenters. The mechanisms that would be required to activate everything is in OpenPnp, it's just a matter of writing g code for it. Currently centroid files are converted on the fly for pcb's.

@@nidavis further, if something was designed in pcb software, screw locations could be loaded into OpenPnp with the "feeder" being the location of the screw presenter, and an actuator for the Z axis pick and place commands. The screwdriver would just be on the head instead of a SMD nozzle. That being said, electric screwdrivers aren't light, and you'd have to accommodate for that. Nema 17 motors would have a hard time moving a screwdriver with any torque.

Yes, I was thinking jig1 would have locating PINS that would jsut line upwith jig2, then put it on a hinge frame that just picks it up and over and down into the jigs lining up perfect.. no more needing to put those smaller items in the foam to line up. could even make it where the glue shields are on hige mechinisms too that fold on to glue and fold out when done and fold the foams together... could make doing that job down to about 10-20sec per foam set..

As someone currently in this stage oh man yes.

Perfect! Know your strengths and know your weaknesses. And don't let your weaknesses become a chain holding you back.

*This may just be me*, but wasn't a lot of the Philosophy behind the Open Source Hardware / Maker / Mid-Scale Manufacturing movement to keep things local / in-house? Thus having resilient microfactories/makerspaces, not Ballet like hyper-globalized supply chains that can collapse like a house of cards?

@@ericlotze7724 The only thing they produce truly in-house is software, and the design documents/files. The circuit boards, the semiconductors, the metal extrusions and plate, the stepper motors, the timing belts, the wires and tubes, the cardboard boxes and foam, the nuts and screws, etc. all originate from external suppliers. Opulo sells a product that enables makers, people who usually produce small one-off batches on a tiny scale. The tools used by makers are superb for making prototypes and rapidly iterating for R&D and product design. They are also great at doing small-scale production for the first months or year of a startup, unfortunately, they scale terribly when you start getting hundreds of orders per day. There is no reason for them to restrict themselves with such limited tools if the goal is to have a LumenPnP in every makerspace around the world, similar to what Prusa has done with the 3D printer.

hah yes. this is one of those things where you get stuck in a solution and don't stop to think of an alternative idea.

How would you apply glue with roller from bottom, the foam is quite soft

how about a belt cutter and winder combo?

Was thinking the same thing

depending on "protective" sheet you put on, a lasercutter was standing with there. pop off old gummed up sheet, put on new fully covering foil, run throught laser (indexing pegs are already in the jig, maybe add some indexing to the cutter?) run job, continue glueing.

or apply glue with a roller...

... semipermanent mold release on the stencil. It's designed for composite molding & the thick adhesive layer will just peel off when you get tired of the gunk. Think commercial-grade RainX. You might even get away with a ceramic-based car wax.

The jig is made by wood which is already very cheap, throw the used one and laser cut new.

paint roller. or a CNC glue applier

I think a full breakdown of the kitting process and how much time every part of it takes could be an interesting insight into the mid-scale manufacturing life!

Add a cutter mechanism so it winds and cuts. Continuous unattended operation might be tricky without a way to keep the material coiled. I'm envisioning a magazine/cassette mechanism with about a dozen slotted holes for the coils to drop into. The holes could be arranged linearly, or radially like a DVD jukebox. Once the coils are in their respective slots, ziplock bags could be lowered (by human) onto each coil. Skipping my magazine/cassette binning process would be better if you could figure out a way to bag the coiled material directly but I am not smart enough to automate working with ziplock bags.

Lots of good ideas here. With the gluing process, I think we can all agree that excessive glue is wasted material and cost. Excess ultimately becomes scrap that requires additional cleaning processes so it's best to minimize the wasted glue, if not eliminate it entirely. I like the idea of eliminating the foam. It's surprisingly expensive and doesn't really add much value to the customer. One alternative is vacuum formed plastic. It is surprisingly easy, fast,, and inexpensive

@@brad3378 I hadn't thought of vacuum-forming plastic. That would work well.

they're using a spray contact adhesive. it needs both sides.

Use of roller both sides still works and is much simpler it would seem.

@@adamjackson9988 ah- except a roller would coat everything, and they can't do that because they need it to make a shelf for a part, not just a thicker piece of foam.

@@SignalJones but they are laminating 2 pieces? They make small rollers

@@adamjackson9988 yes, but that requires more skill than foam, stencil, spray, stencil off, place.

Stepper motor and arduino for precise, repeatable length cuts

Interesting - any suggestions of a brand / search terms or a link …? I have this problem looming in time too !

@@jonnyhifi I don’t have any brand suggestions, they are kinda made to order, can range in cost from £1k to £30k depending on need. - There are a few DIY models on KZfaq though, search ‘part sorting machine’

@@jonnyhifi they all work in pretty much the same way, vibration or spinning parts to separate them, then mass or sensor counting them, mass ones are less accurate because they rely on a regularised mass of the part, and sensor ones tend to need more customising for part shape.

he could even print them

@@l3d-3dmaker58 "get" wasn't intended as purchase, easily fabricated.

I’m holding out for that Yeti Mug 🤤

that's a very very interesting idea! with enough of them, you could source your PCB from a local machine and not wait for pcbway assembly and shipping!

Philipp are you talking about PCB manufacturing or assembly? Consumer-grade printed circuit boards take more than one machine to make, and there are very few board houses in the US. It’s not like 3D printing where there are a huge number of hobbyists with them and we just need a web platform to decentralize existing manufacturing capabilities. Assembly is another issue entirely, as you need the PCBs to be in a compatible PnP machine that is loaded with and can accommodate every component in your BoM. Traditionally, small-scale PCB assembly works by sending the board house your BoM and they’ll then source and invoice you for reels of each component, then send you the surplus components with your order. It doesn’t make sense for them to hold onto specific components you need for your board in their inventory in hopes that someone else will need the same component. JLC is able to simplify the process by having reels of the 1000 most commonly used components pre-loaded into PnP machines. They can be confident that they’ll be able to use the surplus, so they don’t need to source and invoice the components individually for each customer. If you have a component that’s not on that list of 1000, they can’t assemble your board because it doesn’t make sense for them to set up a PnP machine for one small custom order when their fleet can handle more common orders much more efficiently. Until there is a fundamental shift reducing SMD components to their building blocks, allowing a wide variety of components to be assembled from a small number of unique blocks, and we have machines accurate enough to PnP them, PCB assembly will be limited by these inventory and supply chain issues. And I think that’s an idea worth sharing, thanks for coming to my TED talk lol.

@@griffinvanhorne8695 I also know that JLCPCB does this. Problem is just we currently have a war and I am not very sure what side China will stand on. I am not willing to rely on China for this.

Philipp it would definitely be great but I don’t know how realistic it is quite yet. The cost advantage of PCBs and SMD components made in China makes it hard for international competitors to survive. Opulo is making strides toward allowing bare PCBs to be populated affordably in small workshops, and this could allow for a JLC-style PnP farm with a massive inventory of components elsewhere in the world, but decentralized assembly is still dependent on one single node of the network having a large inventory of components. In the meantime, this machine will be perfect for setting up for a few specific boards with a limited BoM and cranking out hundreds to thousands of them. Maybe the political climate, freight cost increases, retaliatory tariffs, etc. will make domestic manufacturing relatively more affordable, more domestic board houses open up, components can be sourced relatively more affordably domestically, and there’s greater demand for domestic quick-turn PCB assembly. The other option is to join or start a makerspace where you can split the cost of digital fabrication equipment like this machine. That’s probably the next realistic stepping stone.

Apparently it's for "Function" aha

Super true, we're also looking at getting folded cardboard packaging to replace the foam! The foam die will help us in the short term, but we're definitely switching to a more sustainable choice! The way Prusa kits arrive with mostly folded cardboard and some recycled foam inserts is the approach we're considering.

@@stephen_hawes I don't work in packaging, but mycelium (mushroom-based) is a recent option I've heard about that might be worth at least a cursory look.

@@eliotsandberg8979 they looked into mycelium. One of the main issues for them is that 1) it takes like a month to grow, and 2) you need a lot of molds, which adds up in price. They wanted to hold off on it until they have their packaging decided pretty solidly. Then they might consider it.

​@@natalie5947 First result in google reckons it grows in 7 days, but that's still way longer than bending some cardboard and whatnot. Thanks for the info, and fair point on needing a lot of molds :-)

​@@stephen_hawes That's good to hear, I was thinking of the Prusa packaging too. great to see you make progress with your company

congrats (?)

Well done, feels great doesn't it?