3D-printed automation - Screw packaging
ฝัง
- เผยแพร่เมื่อ 2 มิ.ย. 2024
- This is the beginning of a video series about a machine I built using 3d printing to count screws.
In each video we will go through the development.
The design files will be released at some point. I have not yet decided in what form, at the moment i am focusing on finishing the machine. But my goal is to develop things that people will use, so if you want one yourself - be patient, it will come soon.
Chapters:
00:00 Into
00:39 basic feeding path
02:44 solenoids
03:10 alternating dispenser
05:16 binary dispenser
06:10 sensor based dispenser
08:03 speed - วิทยาศาสตร์และเทคโนโลยี
6:44 “I am having an internal discussion with me” is the most engineer thing I’ve ever heard ❤😂👍🏻
incredibly clean and precise builds here. highly impressive
thank you!
I have a suggestion for an alternative metering method.
Place a small driven “sprocket” near the bottom of the magazine, so that it’s “teeth” poke into the side of the magazine slot and will catch the bolts as they fall.
Words are bad for describing mechanical systems, so here’s an example of functionality. If the sprocket had six “teeth” and I wanted to dispense 15 screws. I would rotate the “sprocket” 360/6 * 15 degrees + .
I see a few options for powering such a “sprocket.”
1. A stepper motor
2. A DC motor running through a continuous-to-discreet linkage such as a Geneva drive
3. A solenoid pushing a rotational mechanism that indexes one position every time the solenoid actuates (unfortunately I can’t find an example of this right now, but I know I’ve seen them.)
The biggest problems I see are getting the teeth of the “sprocket” to cleanly engage the screws without jamming, and the weight of a stack of screws constantly loading the drive system.
Thank you so much for this comment! I really appreciate it!
All very good suggestions.
My first thought on this is:
I'm not sure if the teeth of the sprocket can really reach far enough into the magazine to hold the screws 100% tight without the tooth that "swings into the magazine" next colliding with the screw next in line. Based on pure intuition, this would require a very large diameter (and many" teeth") of the sprocket. This would require a very large diameter of the gear. But I haven't thought this through yet.
I will test it in the next video.
@@christopherhelmke i actually have a few of these running and they work excellently if not a bit slow. just a sprocket directly on a nema stepper (or geared if the holding force of the stepper is too low).
my main issue with the ones i have built is that the sprocket is 3d-printed to fit the specific task and as thus, are plastic. they work perfectly when new but wear out too fast for me.
also, for the solenoid approach, please do use fail-safe design if available (the springs pushing the rod into the magazine and current retracting it). i know it's probably not as bad for youbut it can at least be annoying. i got flashbacks to a solenoid hopper feeding 400 bars straight into a lathe bar feeder when it lost power.
You could just make it such that the gear is directly at the output and a single screw falls into the spot between teeth and then just turn it around to dispense letting them fall out
@@lhadnusj there is a new video
@@christopherhelmke Look at two gears one on each side, then sprocket would be 1/2 the size.
One of my long term goals is to make a “BOM Vending Machine” where you basically walk up to it (at a Makerspace etc) insert your Bill of Materials (BOM) and it spits out all the parts you would need for that from a wide selection all counted/cut to length etc
You seem to have made a part of that system already then, well done! I especially love the modular aspect.
I’m excited to see how this project goes!
glad you like it!
My hardware store has something way better. I can buy srews, nuts, bolds, washer etc. by weight. They have different prices for the different materials, standard, galvanized and stainless. I put everything in a bag what I want, at the counter they put it on a scale an I pay it for the weight.
@@jonasstahl9826 Almost like those candy or icecream stores, that’s neat! Still leaves the hand picking + hunting for what you need issues which is what i hope to eliminate.
@@ericlotze7724 Year pretty much like that.
Picking not realy they are dirt cheap.
Galvanized is 10€/kg
A M5 Bold 30mm long weighs 5 gramm
Means a single Bold cost just 0,05€.
That is basicly the same price per piece as if I would buy a box of 1000.
I dont count, I just grab a hand full of what I need.
When you said controlling the speed I saw it clear. You could try to make a gear wheel that engages the column of screws through the lateral slid of the magazine. With a stepper motor you can make it turn and control the number of screw you releases at any given moment. This will substitute all the selenoid and even one sensor from top. You could solve the hysteresis problem, starting a counter which is counting turns of the wheel (the one I proposed) when the magazine is complete, then when you release the number os screws you want to leave space the counter would reach a number of turns then you send the signal again to start the picking screw up wheel ( the one you actually have). I hope I've explained myself, thank you very much for your time.
I feel like this is going to be one of my favourite channels on TH-cam. Great video, great designing, and as you said in another comment, that you would like it to be pleasant to watch your videos. I can tell you that I really felt that this video was pleasant to watch. So you are already on your way. Thanks and good luck. 👍🏽
Thank you! :)
you could have a toothed wheel at the end of the mag to dispense the screws, you just need a sensor to check if there is a screw to be fed and then having an encoder on the wheel, you could just command a certain number of screws to be dispensed (sorry if you said this, I tend to jump around videos) Hey you should have your store linked in the description or on your channel, I'm curious what you sell.
yes, I tried that in the new video :)
That was my thought too, this method is impressive, but seems overly complex for a screw dispenser, looking forward to the cog method when I watch the next video.
Yeah or even go for like a pez dispencer approach
Screw counters/feeders for automatic screw guns count or sense the screws as they are leaving the feeder, and shut the feeder off as soon as the total count is reached. Then the output ramp is vibrated separately. Automated pill counters and vibratory bowl feeders work in a similar manner (but often without an output ramp). I think it is worth asking if the magazine portion is essential at all if the output control and counting could be done earlier.
Very interesting project and designs so far. I’m interested to see where this goes.
Also gotta say, very well-made video. I was shocked to see how small this channel is.
I send the link from your video to my engineering department, very impressive of the clean engineering and buildup of the design.
I like your drafting board. Brings back memories. That's how I learned to design 35 years ago. Nice project btw.
Great job! My respect for showing the way of finding possible solution in such a clear and fast manner
I've been wondering: Why didn't you countersink the screws on the feeder wheel, so the push-off-arm could be static? I don't know the exact design of the wheel, but I feel like it should be possible, and it would remove a moving part, which is always good.
IMPRESSIVE！Do more uploads pls!
That’s an awesome project! :)
thank you!
This is like real life Satisfactory!
I love seeing challenges and solutions to problems like this. Of course, like all the other viewers, I have a brain full of alternative answers that I would like to see tried out.
Videos like yours make me realise that I probably chose the wrong path in life... too late now, I am enjoying my retirement, watching plenty of TH-cam, and frequently saying to myself "I wonder if..."
Thanks for the excellent videos; Liked & Subscribed.
Really cool project! You might only need the lower inductive sensor. When it no longer detects a screw, run the collector until it detects a screw again, and then for a certain additional length of time to build your hysteresis buffer.
Really liked the modularity of the system. Great work
"i am having a bit if an internal discussion with me". That is such a good line!
Hey! We had a dispenser, that only consisted of one solenoid, but had a slider. The slider was seperating the screws and kept them from falling. The nice thing about it was, that screws never rushed through it, if ther is no power or a malfunction.
Downside: For 5 screws, the solenoid needs to work 5 times.
a gear with long teeth at the mouth of the magazine or reaching into the side of the magazine near the mouth can be used to control the dispensing. the teeth, with pitch equal to the distance between each bolt in the magazine would allow the the gear to dispense one bolt per tooth. an optical or other sensor can be used on the gear to count the number of bolts dispensed per angle of rotation. a solenoid activated clutch can be used to seize or release the gear at any time, otherwise the gear is free to turn by gravity of falling bolts. this should be very fast in dispensing and also very fast in stopping.
to improve the reaction time, the dispensing gear can be used to drive a larger gear with various adjustable markers placed on its face. a sensor is used on these markers to trigger the clutch solenoid whenever a marker is detected. a user activated pulse switch is used to release the clutch until the next marker is detected which reactivates the clutch. the reactivation interval and thus the number of bolts dispensed is programmable based on the adjustable markers. or if it is easier, make multiple pre-programmed large gears that are swappable.
To slow them down I'd set the solenoid to oscillate to only let one screw through at a time. It would still be fast but, controlled.This also removes the need for an optical counter. However, maybe a better solution would be a motor and a bucket sort of system to grab individual screws and deposit, get it to spin only the required number of rotations to deposit the correct number of screws.
edit: another idea would be to use compressed air, like most industrial sorting systems. You can see them in action at very high speed combined with optical sorting in food factories, blasting away anything that doesn't pass QC. Or maybe just an actuated pusher. Or... again some sort of rotating collector at the bottom of the mag.
I like the pusher idea. Push/slide them out front or sideways one-by-one out of the magazine with a pusher like the bolt do in a semiautomatic/automatic gun without any sensors.
6:54 you can get those inductive sensors with a narrow sensing window too. The wide ones got the tall plastic sensing end like the one you got, and the ones with the narrow is just a thin disk on the end.
If I understand how an inductive sensor works, then I agree, because any type of light sensor would have a tough time differentiating 2 or more bolts/screws falling tangled or over-lapped with each other, which you can see they do in the slow motion video.
@@DarrenHughes-Hybrid I would have put it so it senses the screws from the side, just behind the head. That way the size of the head helps keeping the thread a bit apart, and then I would use a sensor with only 2-4mm sensing range with the narrow sensing window. Probably something like a IMB08-02BDSVU2K
@@mikeydk I think I get the picture you are describing and that sounds like a good idea, but I am not very familiar with the IMB08-02BDSVU2K sensor.
looking forward to the files.... very nice
This is so cool, great work!
Interesting end well presented!
I am glad i found this cool new channel.. Keep the content coming.
very good idea 💡
Beautiful design, inspired to build my design skills
Amazing work
Very cool! Subscribed!
Great build!! Something I always try to do is to pause for a sec and ask myself: “how would someone solve this problem 500 years ago? (With basic tools)”
You can use gears! Where the space between the teeth is big enough for your bolts, that way you can accurately count them, pause at any time and go as fast /slow as you wish.
I hope this helps.
I too am always 'having internal discussions with me'. Loving the channel!
Thank you!
Another complication with the sensor method is that acceleration is on a curve so the timing of when to close the gate is not consistent. Super rad project though.
Great video! Your prototypes look better than the iPhone 15. To battle the issue with falling speed you can add an insert wheel that has cutouts for screws and which rotates a required amount of degrees per screw to dispense.
Subbed right from the getgo. Your channel will be reallly interesting! :)
Side note: Optical counting could be tricky because the end of the screws could be touching so the heads, so you only can detect somewhat constatly at the neck of the screw. Also the timing will be tricky as well because of acceleration.
You could make a bottom drum attachment with the screw negatives in it. The screws fell from the magazine fell into the negative holes. Then you can rotate the drum with a stepper motor (or simply some divits on the drum, a switch and a motor) to turn the exact degree to feed one screw to the bottom.
Nice video! I'm going to watch more of them.
But I really don't understand, why the complexity of counting/separating screws from the magazin is needed, when they are already separated and easily countable after the feeder. Having two solenoids, the magazin could be split in an upper and a lower chamber, so the feeder could count the next batch into the upper, while the lower is available for dispensing an already counted amount of screws. If the the upper chamber has the right amount and the lower chamber is empty, screws can be quickly send downwards to the lower chamber, probably without stopping the feeder.
Yes, the feeder might need to stop more often. If this really is a problem, especially with low-count dispenses, additional solenoids may be used to split up the magazin into more chambers. By this technique, any later counting (Alternating/Binary/Sensor) isn't needed. Additionally, you don't need the induction sensors.
The total output rate is limited by the feeder speed anyway,
Wow! Cool stuff, just came across this channel.
Very impressive video! Really enjoyed your breakdown and approach to the problem. Have you considered using a load cell (I would imagine the screws to be consistently weighted)? By locating it at the bottom of the magazine, you can set up your weight limits in your program based on your number of screws required and use it in line the two solenoids. Thank you for the video, I am a fan of your work.
I have not yet thought about a load cell... currently i'm testing another approach (more about that in the next video) - if that doesn't work this could be an alternative. Thank you for your input! Also the feedback is extremely helpful for me!
The speed of the screws also vary by how many is unloaded at the same time because of gravity accelerate objects, so it is a non-linear problem. In addition to this fact I also think the binary solution is most promising because the lock pin is never interrupting/intercepting between screws in motion and/or makes potential jams. I think the sensor solution is possible(after ALOT of work and problems). But the question is if the added complexity is worth the faster dispensing time of a sensor? If in need for faster dispensing you could always add another solenoid for some additional addition options, or drop the whole magasine with the sequential or binary system and filling it up again activated by the empty magazine proximity sensor. Either way you seem very reflected about the pros and cons of added complexity, so I am sure you make it work to your benefit either way.😊 Just found the channel, keep up the good work!
You could put a carousel with slots under the chute. A screw would drop into a slot, and to dispense a screw you would rotate the carousel until the slot was over a hole, letting the screw would fall into the collection box below.
Well done.
You‘re lucky the screws are heavy, the machines I learnt on used vibrating pans with elaborate mazes to sort very light but far larger parts.
Pick and place arms have come a long way and depending on the properties of the material you‘re sorting or dispensing you can use them to your advantage (like magnets and precisely setting the force they have on what you want to sort).
I‘d also have a look at modifying the angles of the curve to space the objects out and then using a simple latch and two sensors (one counts, one for item present before latch).
You don‘t really need a magazine IMHO if you use a latch and two sensors next to (de)acceleration of the items by different angles in the curve.
We used to sort and place caps on vials at a machine speed of >600 bottles/min with a simple latch and two sensors as I described above. Two prongs would hold the cap in place at an angle and the stoppered vial would pull it out and move on to the next position for crimping.
Not strictly a continuous process but also not a sequential one either.
Solenoids can be controlled very precisely if you need to.
Similar principle like microstepping, one controls the buildup and draining of the magnetic field in the coil.
One could fill 10 vials/s (1ccm with
great vid
Super video, I directly had to abo and like.
wow this is so cool
You can put the magazine under an angle so the screws don’t go as fast
There's a video of how the part of this wheel box that holds the screws looks like, how the box is assembled inside the other box, very cool, really cool this project is sessional
Yes, there will be videos about that as well. My plan is to work through the whole system in the videos.
However, there is really a lot going on in the seperator, and it is quite complicated to explain - so there will be several videos about it. Also due to the fact that the separator works a bit different for nuts/bolts/washers.
That's why I decided to start with the elements that are easier to explain, my goal is to explain it well and to make videos that are pleasant to watch.
Hopefully it won't take too long until we get there.
Subscribed.
Cool project! I think slowing them down using an electro magnet might be worth a try. If the screws were non ferrous just a bank of permanent magnets would nicely slow them down because of the Eddy currents generated in the screws. But with ferrous screws you might need to pulse an electro magnet.
interesting idea, I will think about it, thank you!
keep it up!
Oh and a Pentagon Wars recommendation. Double plus good
Found your second video first, then came back for this one. I hope a future one will be about how you did the orientation! Anyway, super interesting; subscribed… though I will admit to being disappointed with your suggestion in the Prusa video to stop open sourcing… that’s sacrilege, in my book. It’s about freedom, after all… (have you read the FSF’s post about the philosophy of free software? If not, I hope you’ll consider doing so!)
Guess who won a new subscriber
I have seen your later video with a stepper motor drive so my comment is purely hypothetical.
The speed of the first screw falling down is different than the last one in the sequence and the speed of the last one depends on the height of the stack. So the demand on the system to count the and close the chute is a bit higher than average you calculated from dropping 60 screws.
In a stepper based system you control the “rate of descent” rather than gravity.
Another idea you can consider is to use a horizontal conveyor moving at constant speed, carrying one item per “tray” then using a solenoid or pneumatic cylinder to push the part off the conveyor.
Complicate this with computer vision and you could package different parts into a single container, as long as there is only one part per section of the conveyor.
Istead of "locking" the 2 solonoids of the original system together with a fixed printed part, why not design the holding mechanism with adjustability in mind?
The bottom one could be fixed to the magazine and have for example a toothed rack extending up. And the upper one would have a pinion gear which can be tighened down once in the right position, so it stays in place for having the holder tightened as well. Alternatively there could be just made space for a literal spacer "stick" of the required length, that keeps the two apart.
crazy ^_^ I would live to build one and maybe tinker to modify it. Please release it even under some sort of payment
I know you've solved this in the next video, but an idea for this would be to use a alternating channel with something like a pinball paddle pointing down that simply flips back and forth to release a single screw, loading while the one before it is falling, you become limited mostly by how fast the paddle can swing back and forth. The paddle can be driven by a really dumb DC motor.
Instead of a parametric designed model for the solenoid bracket. I suggest the bottom solenoid be fixed and the upper one be mounted with some sort of quick clamp. Then you just slide the upper one to the desired limit
I'm here before he blew up
Sorry if I'm getting ahead of things, but your sorting unit, does it use magnets to pick up individual pieces? If so, how do you manage non-ferrous fasteners, say, stainless or nylon?
Thanks for posting your journey!
You mentioned that you still stuff online.
What do you sell?
Post a link to your web store!
This is awesome. I know it's late, but why not have a gear with teeth that interlock with the screws and turn it a certain amount for the number of screws to dispense?
very nice design, i was thinking of the screw counting and wat came to my mind was multiplication, if you have two versions interchangeable with odd and even number of screws between the solenoids or even static mounted, you can maybe use just simple multiplication, so if you need 20 screws you just set it to 20 and it will figure out how many times the solenoids need to open and close.
And why isnt it possible to make the upper one variable in hight?
Any plans to share files?
Did you have an overview video of why you are building screw feeders? What is the back story here?
Rather than a solenoid and sensor setup, I'm thinking a stepper motor with gear dispenser may be easier here.
Allows you to dispense a precise number of screws, though it will be slower than pure gravity for larger number of screws.
Could you not have a little wheel with notches in (resembling a sprocket) that rotates with a motor, stepper or with an encoder, and it just rotates the correct number of notches to count the screws out? Would need a different sized wheel for each screw though
Would it make sense to make the upper solenoid’s position adjustable? In general you’re going to package the same number of screws. Fix the lower solenoid, and make the upper solenoid position adjustable on the magazine. The manual adjustment is relatively infrequent for a given screw type.
thanks for your comment! yes i think that would work, but i like the flexibility the sprocket design brings (in the next video)
what about slowing screws 4 times slower to keep electronics easier? Maybe if could be some rubber wheel dispensing screws or magnetic field could slow them down? I am not an engineer but I am interested in your thoughts about this one
The euro boxes you use - where do you get them?
IKEA needs one of these
Why dont you make both solonoids movable with the same screw clamp connection you made for the 3d printed mount?
This way you can set the distance manually and tune it by hand
Just came here after watching part #002. Excellent design overall, and the servo motor is pretty fucking wicked, but I'm wondering why not just program the rotation of the feeder, and have a sensor at the uptake (the top of your mag). You could just add an amount of rotation if no screw was detected being picked up. I see you thought of something similar and of using a solenoid, but a solenoid or an additional motor are not necessary if you were to implement it this way.
I am looking for solenoids on that style, where did you get those?
Can you build a bolt sorting robot from scratch?... congratulations on the contribution, and complement...
Could you use a motor with a sprocket shape on it to dispense screws depending on how much you rotate the sprocket? Depending on how many teeth the sprocket has determines how many screws are dispensed per rotation of the motor or you can do fractions of a full rotation to dispense an infinitely custom amount of screws.
I am just realising that the next video is based off someone else’s comment with the same idea ahahaha.
Hi Christopher, this design is brilliant! After watching your 2 videos, I was hoping to see a TON of other videos from you on how you made the sorting machine, and was sad there weren't any videos on this. Do you have another channel or more videos on how you built the sorting / pickup mechanisms? Thanks!
Yes, there will be videos about that as well. My plan is to work through the whole system in the videos.
However, there is really a lot going on in the seperator, and it is quite complicated to explain - so there will be several videos about it. Also due to the fact that the separator works a bit different for nuts/bolts/washers.
Another reason why I started with the counter module is - that it is simple. I want to make videos that are pleasant to watch. This way I can get some experience with video making before we move on to the more complex ones.
@@christopherhelmke Fantastic news, thanks!! I cant' wait to see those videos! Also, I didn't see info in any of your videos on why you're counting these and/or links to the products or items you're selling/offering. I'd love to see that as well.
Looking forward to more videos!
How does it know when it’s a screw or a nut to open or close that?
What about a sprocket-wheel on a stepper motor that feed one screw for each sprocket? effectivly using the sprockets to feed and stop the screws instead of a pin of a solonoid?
I just klicked the next video and realized I was far from first with this idea, haha! ^^
Hi! This design is incredible!!! Will you be posting some of the parametric designs online? I would love to 3d print and use the screw sorter! (washer, and bolts)
I have not decided on that yet. For me it depends on how this channel develops. At this stage I'm focusing on getting it completed and functional. After that we will see. Since i want to design stuff that will is used, in some form at some point probably yes :)
@@christopherhelmke Great to hear! This content is greatly accepted by the community!
What about a gear style dispenser?
i saw that solenoid in a golf ball return machine..
Why not use weight approach? Those screws have similar weight more or less?
Might be a silly question but what means have you been using to separate nuts and bolts at the first place? I understand that the disc contains equally separated magnets but what happens in case a bolt is set horizontally as it can been seen in an instance. Great great video thanks !
Also as far as the second video is concerned maybe you could also use a belt system for the 2 1 reduction just an idea
I think I was a little unclear, I call it a seperator because it separates the individual screws from each other. You can only insert either bolts or nuts. Not both at the same time. Thanks for the comment, I was not aware that could be misunderstood!
@@christopherhelmke dude I have literally lost my mind it's not your fault. The nuts I had been watching were the ones fixed on the rotating disc...I just hope you had a good laugh
Hello beautiful design, can this also be purchased somewhere? Or do you know how or where I could have this built?
Hi, thanks for your comment! I will write you an email later
Will this system work with non-magnetic fasteners?
How about gear like arms on a stepper and let it rotate the Needed amount?
Are the designs file going to be released ?
I have not decided on that yet. :) For me it depends on how this channel develops and if people even want the files at all. At this stage I'm focusing on getting it completed and functional. After that we will see :)
👍👍👍👍👍👍👍👍👍👍
Checkout Wintergreen where Martin with the community are building a marble machine instrument. I think you will find the marble droppers fascinating.
Yes, i absolutely love its channel! It is indeed fascinating.
How is it that only one number of nuts or screws are attached to the magnet?
I would go for the binary! I would go with a 1 plus 9…
Wow, impressive - your content and construction are high quality
PS: Dich hätte ich gerne in der Berufschule als Lehrer gehabt 😉 👌
Vielen Dank! :)
Hey, very cool project! I'm curious what it is, that you're selling online! By any chance 3D-printed screw counting machines hehe? :-)
These kinds of nut/stud feeders go for about $25k-35k each for a reliable system in the automotive manufacturing industry.
To what, calculate the speed the bolt will be falling? That seems like the most complicated option to use a sensor
Industrially the do it by weight and divide by the weight of a single bolt
If the point was to make something cool, you did it
If you use a magnetic sensor it would rule out stainless screws.
Binary approach is very... programmatic :) not the best in any way, but fun!
Seems you would just use two, with the upper one on a adjustable slide.