hckrnws
That's very nice. The nut dispenser is very effective. Small, and feeds well. The screw dispenser is starting to run into jamming problems. It will probably start to jam more as the acrylic gets scratched and friction becomes worse. But it's manual and low volume, so jamming isn't a big issue.
He's discovered that dispensing is easy, but order from chaos is harder.
There's a whole theory of feeder design.[1] There are clever tricks to orient strangely shaped parts using feeders made from passive components. A basic trick is to get parts aligned in one axis, then arrange it so that the ones that are backwards or upside down hit some obstacle or are not supported, so they fall back down for another try.
When we count pills, we use a simple device that has a flat part and a channel to count into. A hockey-stick-like spatula moves pills into the channel. The excess is poured back into the bottle the opposite way and the channel is then emptied into the bottle.
I count in multiples of 3 so I don't lose my place. The last number is unique for every 30.
https://share.google/VshUpiSioUh6rLg4q (Image link)
At scale, use weight and supply 1 or 2 extra.
This is how pretty much every IKEA, LEGO, etc works with very small, cheap parts.
End users benefit because it's easy to drop/lose/break one.
So that explains why the smallest parts often have spares in ikea and lego builds. Is this done because of the error in weighing the smallest parts, so they have a margin for error by allowing for an extra 1 or 2?
> Is this done because of the error in weighing the smallest parts, so they have a margin for error by allowing for an extra 1 or 2?
This is a secondary benefit, the primary benefit is if the end user loses/breaks one. That part very well could be show stopper (Ikea 110630 anyone?). Now the end user is stuck - has to call, you have to ship, do you charge? do you give for free? they have to wait. they're annoyed, you're annoyed.
No one is happy.
The supply chain headaches for giving exact number of tiny parts is terribly expensive, relatively speaking. So you give spares because in the long run it's way cheaper.
Being aware of this, I am waiting for a solution to what to do with the leftovers besides chuck them into a landfill. The problem, of course, is scale. No one is mailing 3 screws and an Allen wrench anywhere. Maybe once you hit 5 pounds of spare Lego . . .
If you have an IKEA store they do have a place for spares, and you can return them there. Assuming you go back from time to time.
For stuff bought online, e.g. Amazon, not much you can do.
Just tacking on to mention the smallest parts are most likely to be lost, they’re the ones that - if dropped - seem to bounce and roll under a refrigerator or into the ether. They don’t give extras on the larger parts because they’re not likely to be lost. Frequently enough all it takes is a violent/careless bag opening to send the small pieces flying.
How does this work without dispensing onto the scale one by one? Just shaking them out of a hopper?
You're weighing the bag. Dispense a load in and divide the total weight by the unit weight and you know how many you've put in.
Easier with heavy objects, and needs the variation on weight to be low for the number of items you're dispensing.
Sure, but how do the parts get into the bag?
Or a vibrating seperator which can give perfect counts if needed.
> You can probably guess my opinions on it though, the software is very good but the cloud-based vendor lock-in is grating, and the free tier is hobbled beyond the point of usefulness. On the plus side, being browser-based, it works perfectly on Linux.
> The 1.1 release of FreeCAD should be soon. I really want FreeCAD to succeed, but blimey they have a big hill to climb. My fingers are crossed.
Since these parts mostly seem to be laser cut acrylic (so mostly 2D), it seems like solvespace would do a good job at cranking out the designs. I haven't used it for a larger project like this though, maybe it was already considered.
For what it's worth I've been using FreeCad 1.1RC2 lately, and for me it's the first FreeCad version worth bothering with. It's now a tool I actively reach for over OpenSCAD and Blender for various projects. Previously I couldn't make the simplest part with it.
I can't wait for the release proper, but I can heartily recommend everyone try the release candidates as well. I've got a feeling this is the tipping point for FreeCad like 2.5 was for blender.
I'll give it a shot, I remember absolutely hating FreeCAD's UI last time I tried it.
I mean I'll be honest, it's still a car crash of a program, but at least it's now a usable car crash. I've found the following workflow to be pretty good, using the part design workbench:
- create a part - create a body - create a sketch - pad/pocket/revolve/etc - repeat with additional sketches to taste
I've also been using the proxy object thing, I forget the name, the button is a green blob, to "import" geometry from a master sketch into more specific sketches.
> I mean I'll be honest, it's still a car crash of a program
I'm glad you said that so I feel a little less mean...
I gave it another try but it still feels pretty dire. FPS is bad on a macbook pro with a 120Hz screen on simple models and sketches. I explicitly selected "touchpad" as the navigation scheme, but I still can't figure out how to rotate, and even figuring out panning took me longer than almost every other 3D program out there (blender, PrusaSlicer, macos quick look STL viewer, solvespace).
It still has a splash screen and takes quite a long time to load, like an application from the 90s.
Buttons and actions that are completely irrelevant to me are shown, but disabled, which gives a really cluttered feel.
There's still "part design" and "part" benches. No idea what distinction is being drawn there.
Obviously part of this is from me being inexperienced with the tool, but as a new user all these issues add up to something that doesn't feel approachable or enjoyable to use.
Solvespace has its own issues, but at least it opens instantly and is generally a joy to use.
I'll watch some others slog through FreeCAD 1.1 though so I don't have to, and maybe I'll learn something.
You definitely need to use a mouse with a middle button/scroll wheel, IME. With this, there are presets which work just like other CAD programs e.g. Solidworks. Without it, it’s very difficult to use, but that’s not unique to FreeCAD.
The base UI is quite bad but there are ways to improve it - either through settings and better organisation [0] or via plugins.
I’d suggest to watch a couple of tutorials specifically on 1.1 ([1] was my entry point) as every CAD program had quirks and frustrations at first. I’d say that for hobby-level creations, 1.1 now has ~80% of the usability of Solidworks, once you figure out how.
I’m not sure what’s going on with the performance on your system; I’ve used various 1.1 versions on a Windows laptop and a MacBook Pro and they’re both sufficiently performant. (I usually use a development or RC build from GitHub [2])
[0] https://youtu.be/LKq7hgbu7ks
> It still has a splash screen and takes quite a long time to load, like an application from the 90s.
The splash screen can be disabled and it takes 3 seconds to start on my mac. Fusion however has two splash screens (first a regular one, then one that covers the whole app window) and it takes 32 seconds to load! (to be fair, once loaded it's much better than FreeCAD).
Oh yeah, you won't find me defending Fusion. I can understand when you're loading a heavy scene or recomputing _everything_ in a complicated model, but I can't understand multi-second loading times and splash screens for loading...the start screen.
> It still has a splash screen and takes quite a long time to load, like an application from the 90s.
Lots of it is single-threaded, which is an endless frustration on a machine with umpteen cores. Especially frustrating given that it means compute happens on the UI thread.
Solvespace can also be used for 3D.
It is sad that FreeCAD gets all the attention. If Solvespace had some of it, and the development time following from it, it could get improvements and some of the cool stuff in their pipeline. That would IMO make it a much better CAD program than FreeCAD could ever become.
I know, it's just that in this particular blog post, the designs mostly seem to be extruded 2D sketches which solvespace is particularly good at with its sketch interface.
Solvespace can also do a lot of useful 3D stuff, but it's also missing a lot so I can't in good faith recommend it for any arbitrary CAD work.
A beautiful balance of effort and benefit.
I don't know any better, but the screw counting mechanism seems awkward. Imagine the set has 10 components..
I'm surprised there is no standard solution to this - like a tape and reel solution? A counting and dispensing gun that works for different sizes? But how much more would anyone pay for M3 bolts on a tape?
Helmke had a tube feeding his dispensers in one of the videos, with bolts lengthwise. That tube idea could be used for a manual dispenser - imagine a drink dispenser, but giving 3 bolts. Maybe easier to store away, but just as awkward to load.
> the (OnShape) free tier is hobbled beyond the point of usefulness.
The free tier is identical to the standard tier except you can not create private documents and it has a no commercial use clause. This has been the case for many years, so I'm not sure where "hobbled beyond the point of usefulness" is coming from.
This is a commercial use, isn't it? Might be clumsily worded, but it's out of the running for that reason alone.
I love this so much. Such simple machines, for human-scale problems. I often get pulled down rabbit holes of machines and automation - this is a nice reminder that you can solve a lot of problems without reaching for an arduino or a servo.
In a similar vein: https://www.youtube.com/@christopherhelmke
Awesome solve!!! Lasers and 3d printing is my side hobby business and is what keeps my sanity intact. I love seeing the practical creations that are realized by them! One of my core tenets is being self-sufficient and achieving efficiencies. This post is exactly that. Well done.
a small scale problem = assisting device
a medium scale problem = a better assisting device
a large scale problem = hire people
Oh come on, now I want that clock.
Push button, nut.
Wow, that's so cool!
Insane!
>> I have wasted a significant chunk of my life counting out small numbers of parts into bags and posting them to people.
He's counting out like 6 at a time. He needs a fast way to pick small quantities precisely, not a fast way to check large quantities. Once they're picked they're easily verified by eye.
Yes it's by weight when you need exactly 20k tiny screws in a box. But when you need six that won't save you any time.
Then just use a small cup to scoop out some screws.
In volume, small parts are dispensed by carefully designed machines, and then the result is counted by weight. You still need control of the dispensing, and as he's putting in small numbers of items the counting is the easy bit.
Until counting machines got ubiquitous, banks in India would count notes/bills by weight as well.
It wasn't very precise but you could move a lot of money in ball park with this method. Atleast internally across branches.
Up to roughly 100 bills it's pretty much bang on - even with a cheap $10 scale (American Weigh Scales Digital Pocket Scale has a bunch of different options). Each bill weights roughly 1 gram. So - accurate to within 1% - and presumably the banks have better scales.
I suspect at scale (moving either a lot of batches or large batches), you also need to take variance into account more. Some bills might be dirty or have stuff stuck to them, some bills might be damaged and have bits missing? And other things that occur in practice that I can't think of from the comfort of my armchair in 30s.
Crafted by Rajat
Source Code