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.
It occurs to me that the screw counter's main difficulty is in orientating the screws.
The machine does solve that (as a product of all the shaking and jostling and doubtless unjamming), but judging by the length of the feeder tube it's not a very fun step. And the end goal isn't to have screws that are each oriented in exactly the same way, but instead to have a specific quantity of screws placed in each of a series of containers.
All of that effort to orient them so precisely does make them easy to count using the nut dispenser mechanism, but that effort is otherwise ultimately discarded.
I'm lead to wonder if the process of dispensing 6 screws could be accomplished more simply (ie, with less fiddling and shaking) by reducing the amount of orientation necessary.
Perhaps by using a sorter that puts the screws in a line, axially, without a preference for heads-first or threads-first orientation?
> Perhaps by using a sorter that puts the screws in a line, axially, without a preference for heads-first or threads-first orientation?
Here's a vibratory bowl feeder doing exactly that.[1] This is the industry standard way to solve this problem. Look what happens once the screws are lined up without a preference for heads-first or threads-first. A very simple slotted rack gets them all from horizontal to heads-up. As is usual with such feeders, if something doesn't land where it's supposed to, it falls back into the bowl for another try. That's the anti-jam mechanism.
3D printing vibratory bowl feeders works.[2] Useful for when you need to handle thousands, but not millions.
This is more scale than the clockmaker needs, though. Unless his business scales up.
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.
> 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.
For 2D, sometimes I find it nearly easier to write the gcode by hand (or make a python program that writes the gcode for you). It really isn't as complicated as it sounds, especially if you can tolerate doing 3D in openSCAD.
I'm very into code-based CAD, I actually gave a small talk on it a year or two ago. A longer term goal of mine is to make some sort of hybrid CAD modeling tool which is mostly code-based, but has a GUI for certain things like defining sketches/constraints, and selecting particular geometric features that are hard to describe in code.
Here's a link to the talk if anyone can bear to listen to me for an hour:
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.
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.
If you count by threes, the ones place is unique until you pass each multiple of 30 - 3, 6, 9, _8, _1, _4, where was I? I hadn’t made it to 30, ends with a 4, must be 24. 27, 30, 33, _6…
Serves as a sort of checksum, as long as you know roughly how many you have and just the last digit.
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.
It is a clever mechanism to separate a wide range of parts. Like a vi Rating feeder, but without adjusting the device for different object sizes. It is a rotating tube, slightly angled.
How would that help?
Say - you have just one tube to seperate parts. You drop your first box of washers and route them into a sequential storage. They you do the next box with bolts on the same device and drop them into another sequential storage.
They dispensing remainds still manual as mitxela showed.
> 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.
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.
Readit News
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.
[1] https://www.youtube.com/watch?v=ZlyuHIxSC-A
It occurs to me that the screw counter's main difficulty is in orientating the screws.
The machine does solve that (as a product of all the shaking and jostling and doubtless unjamming), but judging by the length of the feeder tube it's not a very fun step. And the end goal isn't to have screws that are each oriented in exactly the same way, but instead to have a specific quantity of screws placed in each of a series of containers.
All of that effort to orient them so precisely does make them easy to count using the nut dispenser mechanism, but that effort is otherwise ultimately discarded.
I'm lead to wonder if the process of dispensing 6 screws could be accomplished more simply (ie, with less fiddling and shaking) by reducing the amount of orientation necessary.
Perhaps by using a sorter that puts the screws in a line, axially, without a preference for heads-first or threads-first orientation?
Here's a vibratory bowl feeder doing exactly that.[1] This is the industry standard way to solve this problem. Look what happens once the screws are lined up without a preference for heads-first or threads-first. A very simple slotted rack gets them all from horizontal to heads-up. As is usual with such feeders, if something doesn't land where it's supposed to, it falls back into the bowl for another try. That's the anti-jam mechanism.
3D printing vibratory bowl feeders works.[2] Useful for when you need to handle thousands, but not millions.
This is more scale than the clockmaker needs, though. Unless his business scales up.
[1] https://www.youtube.com/watch?v=kIPNdrbSYM4
[2] https://www.youtube.com/watch?v=ECx6L7z0T4Y
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.
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.
Easier with heavy objects, and needs the variation on weight to be low for the number of items you're dispensing.
> 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.
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.
Here's a link to the talk if anyone can bear to listen to me for an hour:
https://youtu.be/0wn7vUmWQgg
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.
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.
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)
I interpreted it as: For every three pills I put into the channel, I add +1 to my internal count.
"The last number is unique for every 30" means... if you have a mental count of 13, you have 39 pills channeled? I didn't quite follow...
Serves as a sort of checksum, as long as you know roughly how many you have and just the last digit.
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.
It is a clever mechanism to separate a wide range of parts. Like a vi Rating feeder, but without adjusting the device for different object sizes. It is a rotating tube, slightly angled.
How would that help? Say - you have just one tube to seperate parts. You drop your first box of washers and route them into a sequential storage. They you do the next box with bolts on the same device and drop them into another sequential storage.
They dispensing remainds still manual as mitxela showed.
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.