In terms of feeder behaviour there is quite a lot of variation of tape that needs testing with - paper, clear & black plastics that can have different stiffness, as well as different thicknesses of plastic - parts are more prone to jumping out with thin tapes. Part pocket size can also affect how the tape bends. Pocket depth can vary - 1210 caps and PLCC-4 LEDs are probably the deepest on 8mm tape. Wider tape has a much bigger range, e.g. connectors, inductors and electrolytic caps. Cover tape peel force also varies. Paper tape can be a bit abrasive, and also sheds dust which can cause issues, e.g. you don't want anything that needs oiling near it as the dust turns it to sludge. Feeders are hard. I suspect you'te only at the start of a long and frustrating journey...
2:55 TVS diodes aren't very useful at protecting from 24V , as the 24V will likely toast them - you need to combine with additional protection like a polyfuse. However he best thing I've found are Murata PRG18BB PTC thermistors - these trip to a high-resistance state fast enough to protect RS485 and HCMOS from shorts to +24V with no additional TVS or zener protection. I usually use the 100R ones, as this also provides useful edge damping, or 33R if I care more about edge rates - the latter still offer good protection.
Mike -- I agree with your assessment that the TVS diodes at 2:55 would get fried by the 24V hazard that Stephen described. They are more for ESD which is high voltage but relativeIy small total charge I think. Your PTC suggestion-- presumably those are in series with the input, but what sinks the current to get them heated up? Is this relying on the protection diodes in the RS485 chip (ie: to the rails)? Note to Stephen: From following Mike's channel -- he has done a great number of RS485 projects, and has a lot of field experience with it!
Full disclosure, I work at a filament company, so of course I’m going to be biased toward 3D printing, but I still wanted to share my thoughts. When you design a project, you tend to design parts in a way to be easily manufactured. If you want it to be machined, molded, or printed, you’ll take different approaches to the design. For example, something that is designed for molding might be hard to 3D print, and vice-versa. This then boils down to a big-picture question “how should this be made in the long term”. Injection molding absolutely makes sense if the end goal is for this to be an off-the-shelf product. BUT, leaving something 3D printed enables and encourages the design to be changed, improved, and customized by anyone who has their own printer. Some of these 3rd party design changes might even make it back into the official product. This ideology is very similar to Voron and Prusa 3D printers. For example, Prusa sells hundreds of thousands of 3D printers, but they still print as many parts as possible because it lets them make changes, and allows hobbyists to hack their own machines. Parts that are designed to be injection molded will tend to focus on different things, such as minimizing/hiding flashing, adding small chamfers to seemingly flat sides to make parts pop out of the mold easier, and avoiding any internal structures. It IS possible to design something that can be printed AND injection molded, but it’s certainly another layer of complexity.
@@MrWizard65 they are, perhaps if he reaches out to Dragonfly Engineering, he might be able to do a collaboration on having some parts injection moulded 🤔
I agree. Full commercial injection molding is likely not something that will be super viable for this company for a while though. It's expensive and thus impedes the quick and creative iterative design paradigm that is finding them success right now. I would say though, that it may be prudent to seek a middle ground and possibly manufacture casts of their own and do injection molding themselves. There is a youtuber that does this and shows how to do it. He even works out a lot of bugs in his videos. Right now he's building sets for a stop motion animation he's doing. Shouldn't be too hard to find as he basically is fully funded by Patreons. Despite the speed angle, which is very valid as 3D printers are crazy slow, injection molded parts are smoother and much stronger. Less chance of failure.
Awesome work! I've been a part of quite a few software startups as well as currently running my own manufacturing business... It's never easy, there's never enough time in the day, but it seems like you've got it under control and it's exciting to see how things are progressing! -Aaron
I've loved following your journey and you guys have done so well! It's great to see how you have evolved; embracing both lean manufacturing principles and continuous learning. I think those are two of the key aspects that help drive success in a startup like yours. Smart move to MVP smaller 8mm feeders first - smaller feeder also prints faster to reduce iteration times. Focus on getting it right and then out the door to offer maximum relevance and value to customers......use lessons learned to develop the other sizes. Congrats on the win - totally deserved!
It's wonderful to see the growth. I've been watching this channel grow since the beginning. Glad to see all the commercialization happening and contribution to the hardware dev community.
12:44 Very nice, only comment is perhaps more fold up trestle tables (you have a couple in frame) for final packaging assembly; packing on the floor isn't the nicest way to do it. Besides, fold up trestle tables are generally useful, cheap, fold away to nothing when not needed and easily on-sold if you ever want to get rid of them. Very proud of what you guys have done (an old [literally] Linux dude from way back).
Super hype for the new feeders! They’ve already changed a lot since I saw them at MRRF this summer (they were CHONKERS then), and I’m excited to see the changes that continue to happen after launch too. Also, super cool seeing the new gold color actually being turned into parts now! (Sorry if this is a duplicate comment, I think my last one got removed because I posted a link to a pic)
Engineering students (as I am) can learn so much from watching your videos to understand what it takes to take an idea to production ! I love watching your progress and the way you tackle your problems and solve them ! Great inspiration for me !
On your points about efficiency. Looking at your shop. I would investigate the 5S and 8 wastes from the Toyota production system and how to implement lean effectively. (I say this cause I am seeing a lot of batching) Also, think through 3d printing vs polyurethane casting. Casting can be cheaper and faster (depending on the liquid plastic you get). I 3d print only prototypes to make molds. I am not yet at production level to need to either buy your machine or make it, but I am watching closely, over time it will make sense.
Just started researching PnP machines because I want to expand into producing my own SMD boards in-house and I have to say that I am very excited by this. The alternatives from the maker community don't seem to have the support and openness I'd want in order to feel confident about investing my time, and frankly the budget machines (by which I mean 2x, 3x, 4x the cost of yours) coming out of China seem to also have questionable support and some unexpected deal-breaking limitations. With expandable, affordable feeders this looks like a genuinely industry disruptive option. I can't wait to see how this progresses! I love your commitment and enthusiasm. Great stuff!
You got it right for sure on the decision to just stick with 18mm feeders first so you're not splitting up your attention, focus, time. Get that one totally figured out, any and all bugs squashed and by the end you'll have this really refined product as opposed to maybe 4-5 feeders each with their own weird problem that you've really gotta tackle together because you could back yourself into a corner. I've loved following all of this bud, and this growth couldn't be happening to a nicer person :) and I can't wait to see where and what the future holds for you guys 😃
Also, for the 3D printing shelf. I know that a number of filaments can soak in water vapor from the air and it can affect the quality of the prints. It may help with quality control to build a bubble or something around the printing shelf that has a dehumidifier in it. When spools start, things may be okay, but the longer it sits there, you may end up with issues with the prints. Just a thought.
Good suggestion, but I think at the production volume they're at the spools don't really last longer than a week maybe, so unless they're having really humid air I think they'll be just fine.
Another fab update video, Stephen - I really love following along on Opulo's journey and well done on the recent competition win! If you were willing I'd be really interested in hearing how you have found it moving from just you to 6 people and more of the business side of things.
To avoid the shorting issue, I would put a NO cherry switch on the feeder mount that only allows power to flow through the connection when the mounting trigger is released and closes it.
I'm finishing up a run of 50 christmas tree ornaments with 60 components. Nothing like placing 3000 components by hand to make we wish I had a LumenPnP!
Hey Stephen, I can share some insight with you based on a commercial P&P. I have a Essemtec Paraquda and their Hy-Q feeders are similar in function to your feeder (good job btw). The issues I run into with the commercial feeders are mostly with the thinner plastic carriers either causing the parts to jump (and sometimes flip) in the pocket after the cover tape is peeled or for the same thinner carriers getting jammed / bunched up before it can exist the machine causing the feeder to jam. The other issue we've run into is with cover tapes not peeling off nicely from some reels, and is hard to detect because even with the machine camera you can't see the clear tape easily.
Hi relatively new here and first comment but been really enjoying the development of this system. Great idea for the design change. From the electrical schematic it looks like +24v is the pin that would connect first, is good practice to have GND be the first to connect in this sort of hot sawp configuration.
Hey Stephen! Sorry if you've covered this in your other videos, but -- have you tried applying the OpenPNP computer vision to actually identifying the solder pads of the individual components, and accurately dispensing solder paste this way? Really interested in adopting your product. Thanks. -TC
Oh boy, I can't wait for an update. I'm from Germany, working on those machines and as an big electronic enthusiast, I will some pick&place on my own because we wanna to get things small (except thin I can't mention here ;)) and if you need some support on 3D printing or mechanic electronics, I will try to help you :) awesome project and I hope u will get rich :D
If your doing machine vision to test position delta on the feeder strip then why not incorporate the machine vision to negate that variation while in production?
Any P&P machine needs vision anyway as part position in tape will vary so that will take care of some of it, but it's always best to start from the best situation you can.
Cool! Sounds like it's on track for being ready for prime time by the time I'm in the market... currently looking at 2Q23-ish, depending how things go here. It'd be nice to have an open-source pick & place machine; the existing Chinese machines in my price range seem a little short on documentation and suchlike details. Supporting a domestic startup would be a big bonus.
Hello, I wanted to through this out there, Your package box that has the camera sight on it. Someone might mistake that for a bullseye target and think there is a gun in there. Just saying.
If you want to do a bit of research on your own, search for over voltage protection using diodes. Diodes prevent “reverse” flow. Basically, if you tie your sensitive input line with one diode in reverse going high (aka to 3V source) and the other to ground, then you protect your circuit. If the input voltage ever gets too high (eg 24V in this case) then the voltage will “dump” out into the 3V source. Now of course you don’t want to short circuit 24v to 3V, so you can use a resistor to limit how much current can actually flow. Digital/analog inputs don’t need much current since they work on voltage instead. That means you can use a 10k, or 50k resistor and not have any problems
oh :( you were at electronica? We will buy one LumenPnP in a few months and that would have been a great place to meet you :) Greetings from Germany :)
Seeing your open air print farm has me worried you may be breathing in microplastics. See this video : 'Is breathing plastic fumes good for you?' By 'Nathan builds robots'.
Your waaaay to giddy.. that sound in peoples voices when they talk to babies or when little kids get excited or adults are faking it for a children’s show.. yea, you got that problem.. like the way most people sounded on sesame street….
In terms of feeder behaviour there is quite a lot of variation of tape that needs testing with - paper, clear & black plastics that can have different stiffness, as well as different thicknesses of plastic - parts are more prone to jumping out with thin tapes. Part pocket size can also affect how the tape bends. Pocket depth can vary - 1210 caps and PLCC-4 LEDs are probably the deepest on 8mm tape. Wider tape has a much bigger range, e.g. connectors, inductors and electrolytic caps.
Cover tape peel force also varies. Paper tape can be a bit abrasive, and also sheds dust which can cause issues, e.g. you don't want anything that needs oiling near it as the dust turns it to sludge.
Feeders are hard. I suspect you'te only at the start of a long and frustrating journey...
2:55 TVS diodes aren't very useful at protecting from 24V , as the 24V will likely toast them - you need to combine with additional protection like a polyfuse.
However he best thing I've found are Murata PRG18BB PTC thermistors - these trip to a high-resistance state fast enough to protect RS485 and HCMOS from shorts to +24V with no additional TVS or zener protection. I usually use the 100R ones, as this also provides useful edge damping, or 33R if I care more about edge rates - the latter still offer good protection.
Mike -- I agree with your assessment that the TVS diodes at 2:55 would get fried by the 24V hazard that Stephen described. They are more for ESD which is high voltage but relativeIy small total charge I think. Your PTC suggestion-- presumably those are in series with the input, but what sinks the current to get them heated up? Is this relying on the protection diodes in the RS485 chip (ie: to the rails)? Note to Stephen: From following Mike's channel -- he has done a great number of RS485 projects, and has a lot of field experience with it!
Injection molding is not as expensive as you may think. It might make it easier to scale production instead of scaling out 3D printers.
Molds are incredibly expensive. Need a very high part count to spread mold cost over.
mold: hehe
Full disclosure, I work at a filament company, so of course I’m going to be biased toward 3D printing, but I still wanted to share my thoughts.
When you design a project, you tend to design parts in a way to be easily manufactured. If you want it to be machined, molded, or printed, you’ll take different approaches to the design. For example, something that is designed for molding might be hard to 3D print, and vice-versa.
This then boils down to a big-picture question “how should this be made in the long term”. Injection molding absolutely makes sense if the end goal is for this to be an off-the-shelf product. BUT, leaving something 3D printed enables and encourages the design to be changed, improved, and customized by anyone who has their own printer. Some of these 3rd party design changes might even make it back into the official product.
This ideology is very similar to Voron and Prusa 3D printers. For example, Prusa sells hundreds of thousands of 3D printers, but they still print as many parts as possible because it lets them make changes, and allows hobbyists to hack their own machines.
Parts that are designed to be injection molded will tend to focus on different things, such as minimizing/hiding flashing, adding small chamfers to seemingly flat sides to make parts pop out of the mold easier, and avoiding any internal structures. It IS possible to design something that can be printed AND injection molded, but it’s certainly another layer of complexity.
@@MrWizard65 they are, perhaps if he reaches out to Dragonfly Engineering, he might be able to do a collaboration on having some parts injection moulded 🤔
I agree. Full commercial injection molding is likely not something that will be super viable for this company for a while though. It's expensive and thus impedes the quick and creative iterative design paradigm that is finding them success right now.
I would say though, that it may be prudent to seek a middle ground and possibly manufacture casts of their own and do injection molding themselves. There is a youtuber that does this and shows how to do it. He even works out a lot of bugs in his videos. Right now he's building sets for a stop motion animation he's doing. Shouldn't be too hard to find as he basically is fully funded by Patreons.
Despite the speed angle, which is very valid as 3D printers are crazy slow, injection molded parts are smoother and much stronger. Less chance of failure.
Pleasure ourselves to have you and your team with the LumenPnP at the electronica fast forward booth. Stay Motivated and keep on doing great work!
Awesome work! I've been a part of quite a few software startups as well as currently running my own manufacturing business... It's never easy, there's never enough time in the day, but it seems like you've got it under control and it's exciting to see how things are progressing! -Aaron
I love your enthusiasm and upbeat attitude! I hope you are super successful with this!
His face in the thumbnail picture is the exact opposite :D
I've loved following your journey and you guys have done so well! It's great to see how you have evolved; embracing both lean manufacturing principles and continuous learning. I think those are two of the key aspects that help drive success in a startup like yours.
Smart move to MVP smaller 8mm feeders first - smaller feeder also prints faster to reduce iteration times. Focus on getting it right and then out the door to offer maximum relevance and value to customers......use lessons learned to develop the other sizes.
Congrats on the win - totally deserved!
It's wonderful to see the growth. I've been watching this channel grow since the beginning. Glad to see all the commercialization happening and contribution to the hardware dev community.
12:44 Very nice, only comment is perhaps more fold up trestle tables (you have a couple in frame) for final packaging assembly; packing on the floor isn't the nicest way to do it. Besides, fold up trestle tables are generally useful, cheap, fold away to nothing when not needed and easily on-sold if you ever want to get rid of them. Very proud of what you guys have done (an old [literally] Linux dude from way back).
Super hype for the new feeders! They’ve already changed a lot since I saw them at MRRF this summer (they were CHONKERS then), and I’m excited to see the changes that continue to happen after launch too.
Also, super cool seeing the new gold color actually being turned into parts now!
(Sorry if this is a duplicate comment, I think my last one got removed because I posted a link to a pic)
Congrats! Well deserved success. Thanks for sharing the insider views. Fascinating stuff
It is so satisfying to see good design and proper engineering process. You're nailing it.
Awesome how far you came in the time. Nice to see it and that you still have your great mood!
Go on, be happy.
Engineering students (as I am) can learn so much from watching your videos to understand what it takes to take an idea to production ! I love watching your progress and the way you tackle your problems and solve them ! Great inspiration for me !
I like your enthusiasm and finding out to grow from a hobby startup to a real product and share those experiences.
Best wishes for 2023.
Thanks for your tour! I found it exciting enough 👍
On your points about efficiency. Looking at your shop. I would investigate the 5S and 8 wastes from the Toyota production system and how to implement lean effectively. (I say this cause I am seeing a lot of batching)
Also, think through 3d printing vs polyurethane casting. Casting can be cheaper and faster (depending on the liquid plastic you get). I 3d print only prototypes to make molds. I am not yet at production level to need to either buy your machine or make it, but I am watching closely, over time it will make sense.
congrats on the growth! Love seeing your growth and enthusiasm for all things grow and expand!
Just started researching PnP machines because I want to expand into producing my own SMD boards in-house and I have to say that I am very excited by this. The alternatives from the maker community don't seem to have the support and openness I'd want in order to feel confident about investing my time, and frankly the budget machines (by which I mean 2x, 3x, 4x the cost of yours) coming out of China seem to also have questionable support and some unexpected deal-breaking limitations. With expandable, affordable feeders this looks like a genuinely industry disruptive option. I can't wait to see how this progresses! I love your commitment and enthusiasm. Great stuff!
You got it right for sure on the decision to just stick with 18mm feeders first so you're not splitting up your attention, focus, time.
Get that one totally figured out, any and all bugs squashed and by the end you'll have this really refined product as opposed to maybe 4-5 feeders each with their own weird problem that you've really gotta tackle together because you could back yourself into a corner.
I've loved following all of this bud, and this growth couldn't be happening to a nicer person :)
and I can't wait to see where and what the future holds for you guys 😃
What a stud… thank you for all of the great lessons you share with us.
Your a genius! Young but what a potential! So proud about our youth! Future is bright because of you!
Also, for the 3D printing shelf. I know that a number of filaments can soak in water vapor from the air and it can affect the quality of the prints. It may help with quality control to build a bubble or something around the printing shelf that has a dehumidifier in it. When spools start, things may be okay, but the longer it sits there, you may end up with issues with the prints. Just a thought.
Good suggestion, but I think at the production volume they're at the spools don't really last longer than a week maybe, so unless they're having really humid air I think they'll be just fine.
I've really enjoyed watching the development of this product that I hope to eventually use! My PNP is ready!
Thanks for sharing all these informations ! As a small hardware startup, it really helps me organizing my production ! 🤙
You've got the passion! That's all that's needed to make a viable project!
BS! The knowledge, and long hours of intelligent work are what is required, which Stephen has in abundance.
love your videos every time, cant wait to have my own luminPNP!!
Super cool - thanks for the tour: It's an amazing setup you've build. Keep it up :)
When you first moved into this location, did it look huge? How about now? Do you feel cramped?
Great update!
Another fab update video, Stephen - I really love following along on Opulo's journey and well done on the recent competition win! If you were willing I'd be really interested in hearing how you have found it moving from just you to 6 people and more of the business side of things.
Great to hear you went to electronica!
Will check out the booth next time.
Can't wait for this!
Prusa minis are a great pick for a farm. More print heads >> bigger build plates. Consider .6mm nozzles if you want to print even faster
To avoid the shorting issue, I would put a NO cherry switch on the feeder mount that only allows power to flow through the connection when the mounting trigger is released and closes it.
I'm finishing up a run of 50 christmas tree ornaments with 60 components. Nothing like placing 3000 components by hand to make we wish I had a LumenPnP!
Hey Stephen, I can share some insight with you based on a commercial P&P. I have a Essemtec Paraquda and their Hy-Q feeders are similar in function to your feeder (good job btw). The issues I run into with the commercial feeders are mostly with the thinner plastic carriers either causing the parts to jump (and sometimes flip) in the pocket after the cover tape is peeled or for the same thinner carriers getting jammed / bunched up before it can exist the machine causing the feeder to jam. The other issue we've run into is with cover tapes not peeling off nicely from some reels, and is hard to detect because even with the machine camera you can't see the clear tape easily.
I don’t have a need for a pnp but I love watching this startup journey
it was great meeting you! Very exited for the feeders, eventually Iḿ gonan build a lumen...with that little twist. and hopefully all the feeders...
Hi relatively new here and first comment but been really enjoying the development of this system. Great idea for the design change. From the electrical schematic it looks like +24v is the pin that would connect first, is good practice to have GND be the first to connect in this sort of hot sawp configuration.
Hey Stephen! Sorry if you've covered this in your other videos, but -- have you tried applying the OpenPNP computer vision to actually identifying the solder pads of the individual components, and accurately dispensing solder paste this way? Really interested in adopting your product. Thanks. -TC
Oh boy, I can't wait for an update. I'm from Germany, working on those machines and as an big electronic enthusiast, I will some pick&place on my own because we wanna to get things small (except thin I can't mention here ;)) and if you need some support on 3D printing or mechanic electronics, I will try to help you :) awesome project and I hope u will get rich :D
So badass. PNP with feeders = 100% worth it.
Please get a gimbal for handheld shots :) Love the update, so much progress everytime! Keep it going.
yeah those handheld shots had some wild motion blur with every step, feels like just recording on a phone would be way better
Приветствую, классная идея все вообще, удачи в творческих процессах , жму крепко руку настоящему инженеру своего дела.
Congratulations on your award!
That's nice bro 👍 looks pretty cool
Very cool place you have
Congrats on the award! Definitely well-deserved.
How's Lucias the Eternal?
If your doing machine vision to test position delta on the feeder strip then why not incorporate the machine vision to negate that variation while in production?
Any P&P machine needs vision anyway as part position in tape will vary so that will take care of some of it, but it's always best to start from the best situation you can.
YO!!!! You're based out of Pittsburgh! I love this. I know right where your office is. I'm going to keep an eye on this...
This is an excellent initiative - and if we had the room at BOKO - I'd totally host you in an adjacent workshop space!
Cool!
Sounds like it's on track for being ready for prime time by the time I'm in the market... currently looking at 2Q23-ish, depending how things go here.
It'd be nice to have an open-source pick & place machine; the existing Chinese machines in my price range seem a little short on documentation and suchlike details. Supporting a domestic startup would be a big bonus.
Amazing you have high ambition as you should
Ship everything fully I insured with signature if you have the product on the box. Nice graphics though. Good work
When will there be kits available I will buy like pnp kit.
Thanks for sharing, it's inspiring, and I hope that the new feeder design would be open and you could update the GitHub repo.
does this machine place 0603 parts? If so, NEED THIS in my life!!!
You should make a video about how customers use your device to show more in detail some products that your machine is making.
I have a question. Why do you not use Belt Printers for the PrintFarm?
what a cool product .. i need to save up and get me one :)
Hello, I wanted to through this out there, Your package box that has the camera sight on it. Someone might mistake that for a bullseye target and think there is a gun in there. Just saying.
Could you do a deep dive on adding protection to the circuit?
If you want to do a bit of research on your own, search for over voltage protection using diodes.
Diodes prevent “reverse” flow. Basically, if you tie your sensitive input line with one diode in reverse going high (aka to 3V source) and the other to ground, then you protect your circuit.
If the input voltage ever gets too high (eg 24V in this case) then the voltage will “dump” out into the 3V source.
Now of course you don’t want to short circuit 24v to 3V, so you can use a resistor to limit how much current can actually flow. Digital/analog inputs don’t need much current since they work on voltage instead. That means you can use a 10k, or 50k resistor and not have any problems
OMg you were at electronica and didn't say anything??? I was there too and would have been so toked to say just hi to you!
I will wait till laser soldering is a add-on ^^
Sound nice :-) Will there be a EU supplier?
Any update or new video ?
oh :( you were at electronica? We will buy one LumenPnP in a few months and that would have been a great place to meet you :)
Greetings from Germany :)
DVT....ak....that acronym has worn out its welcome in my household = blood clot. Tough year here.
Awesome progress though. Love the work!
any issues in operation at -5 °C (23 °F)?
wtf?
Good video bro
I see on your hand tattoo. You know that tattoo meaning ? I like it
Om
Would be excellent to get some interviews with actual users on the channel at some point
Seeing your open air print farm has me worried you may be breathing in microplastics. See this video : 'Is breathing plastic fumes good for you?' By 'Nathan builds robots'.
It's really nice to see how excited you are about your business. It makes it even more satisfying to see how it's expanding.
You're going to need a bigger space
dont skimp on qa, even if you target qx
dont swear/promise, because you are not owned by yourself
consider variable toolheads, not only one function, single coordinate translation frame, like laser engraver toolhead
see, your communication is lacking clarity, dont even know what you are trying to make, or why, pitch better
Your waaaay to giddy.. that sound in peoples voices when they talk to babies or when little kids get excited or adults are faking it for a children’s show.. yea, you got that problem.. like the way most people sounded on sesame street….
What drugs do you take ?