- 5
- 21 080
curiosity-creates
United States
เข้าร่วมเมื่อ 29 ก.พ. 2024
Robotics, machining and endless possibilities.
3d printed 6-axis robot with automatic tool changer (end-effector)- Ft. Bay Area Maker Faire
I built a 3d-printed 6-axis robot that can automatically switch out its end-effector (or tool) depending on the item it is trying to pick up. For a box with a flat surface, it can use a suction cup. For an item without flat surfaces like a screw driver, it can drop off the suction cup and use a gripper. It uses ROS and MoveIt for control and inverse kinematics. I took this robot to the Bay Area Maker Faire to show it to other Robotics enthusiasts. This is the story of that journey, the successes, the failures and the learnings.
In my next video, I am planning to add some form of computer vision and object detection to this robot so that it knows what type of object it is picking up and then automatically chooses the right end-effector.
If you liked the video, please like and subscribe. Thank you for watching and see you all in the next video!
In my next video, I am planning to add some form of computer vision and object detection to this robot so that it knows what type of object it is picking up and then automatically chooses the right end-effector.
If you liked the video, please like and subscribe. Thank you for watching and see you all in the next video!
มุมมอง: 304
วีดีโอ
6-axis robot made with the only stepper motor based actuator you will ever need
มุมมอง 2.6K3 หลายเดือนก่อน
In my previous video, I built a stepper motor based belt driven wireless actuator. In this video, I build a 6-axis robot with the wireless belt driven actuators. By wireless, I mean that there are no wires that connect the joints to the central controller of the robot. Each joint is independent: it contains its own power supply and controller. Each joint communicates with the central controller...
The secret to DIY desktop CNC machine (or CNC router) feeds and speeds
มุมมอง 4.3K5 หลายเดือนก่อน
In this video, I talk about how I make my DIY desktop CNC machine mill metal faster. The online machining speeds and feeds calculators are geared towards industrial grade machines that are very rigid and can handle high loads. But they do not work for DIY or desktop CNC machines or routers. So, I will talk about how I found the right speeds and feeds to push my machine to its limits by using ad...
How I built a desktop DIY CNC machine (CNC router) that mills steel | Sturdy | Precise
มุมมอง 11K6 หลายเดือนก่อน
Although there are many options to buy a desktop CNC routers or milling machines, none of them seem to fit my requirements. I wanted it to be able to mill steel and be precise and cheap! So I decided to design and build my own desktop DIY CNC machine (or CNC router) and document the process along the way to share it with you all. I used epoxy granite to dampen vibrations and make the machine st...
Nema 17 wireless robot actuator | Massive torque | Position feedback | No backlash
มุมมอง 3.4K8 หลายเดือนก่อน
Wire-management sucks for any robotics project. So in this video, I build a wireless robot actuator (robot joint) based on a Nema 17 and Nema 23 stepper motor. The robot actuator has massive torque, no backlash and it provides full position feedback. I can control the actuator via a webpage on a computer or a phone. I put the actuator to test to see how much backlash and torque it has. The list...
Dwo sdedder modas - translation? Two stepper motors.
Gut Verständliche Erklärung 👍 Danke dir dafür Deine CNC Maschine ist erstaunlich geworden und ich wünschte ich hätte auch eine solche, aber ich selbst habe "nur" eine MPCNC PRIMO gebaut und in Benutzung. Dank deiner Videos ist die Flamme in mir wieder entzündet meine lange im kopf geplante cnc zu bauen😊 Herzlichen Dank dafür❤
I am glad you like the explanation. Thank you for your comment. Keep building!
So many issues with this machine desgine. Idk where to stsrt brotherb😮
Which CAM Program are you using
CAM software within FreeCAD called Workbench.
Why did you add an idler between the timing belts?
I had 2 reasons for that: 1) to reduce the distance between the two active pulleys, 2) to increase the engagement of the belt with the teeth of the smaller pully by making the belt wrap around the pulley a bit more. But in the next iteration of the actuator that I use in my video about the 6-axis robot, I have gotten rid of the idler and used a smaller timing belt. The actuator works great with this change too.
in the first 2 minutes of your video, you set goals for your machine, say, you cant find any suitable off the shelf solutions and continue to fail on all your goals. Very nicely done!
The design of the machine has received some decent constructive criticism, but your comment takes the award for being the most non-constructive comment. So, very nicely done to you too!
@@curiosity-creates my comment was purely sarcastic. Desktop scale? get a 6040 and machine steel with it.. it will do it if speed&feeds are right. will it look mirrorlike as with an kern, certainly not. also +- 0.1 is not precise, so not a real goal here. You target a cheap solution but go for Linear rails? Supported rails or even drawer slides are way cheaper and can be used also if you need to go really cheap i could give a few helpful tips if you like tho. as others mentioned, you need to put your guides as far apart as possible also the same with the carriages. you stack plates and use epoxy granite to fill these, which is great! Maximize the distance between the Y- Rails with using them 180 degrees rotated from each other, then you also can mount the plates on that directly. speaking of gantry plates: it might work like you did, but a more solid aproach with less of a 'z-bend' shape, also filled with epoxy granite should be way more stable. If you want to go for maximum stiffness, a moving gantry is not the way to go. You lose on y axis space tho, as you need more space to have the same travel. A fixed gantry could be made much more rigid due to not having to be moved.
😂 gayt damn
Unfortunately, there are some design faults, particularly on the gantry and the Z-axis. The machine is not as rigid as it could be. The distance from the milling cutter to the linear cart in Z is too large. It is also better to use two cart per rail for a greater support. Ordinary 24k Chinese spindles have a low torque at low speeds <12000. Better to use steel with 16000RPM and an HPC milling cutter (e.g. Holex Pro Steel HPC - 4 or 5mm). I mill steel with 1000 to 2200mm/min feed rate. You can find cutting values in my channel. I think you should work on your Z-axis again. Guide rails are never completely straight and if I have seen this correctly, you are simply setting up the second rail opposite the first. That's not a good idea. If you want it to be accurate, you need an exactly straight beam (e.g. granite) Greetings Anton
Thank you for your comment and suggestions.
Biggest is he's using one berring block for the Y 😂😂 Jesus.
You havent shown the controller..! Please put videos on cnc controllers and configuration
I will try to put something on CNC controllers. Thanks!
Wow, this is very impressive. I'm having a tough time milling aluminum on my desktop machine, and I'm in awe of the healthy chips and material clearing speeds you are getting.
Brooo can I get your phone no ? so that, I can meet well in India where do you leave ?
IMHO the design has too many weak areas with great moments. Factoring the type of materials used and cost, that machine is severely underperforming. A 4030 (on price point of view) out of the box (as reviewed by aurora tech) mills way faster and cleaner than that. For that size requirement and type of materials used, a mill type design would have been more appropriate.
Can the design be better? For sure! I am with you on that. I am actively working on improving and upgrading the design. It would be impossible to create something that is perfect and put the entire journey in one youtube video. So the approach is to record the journey. It might take a couple of videos to the ideal state of the machine. But, I can't agree with your statement about an out of box 4030 being better than this machine. An out of box 4030 comes with a DC spindle. That in itself makes it incapable of machining aluminum, let alone steel. If you decide to upgrade the machine to a decent spindle (with a VFD), it will cost you around $200 or more which may or may not make the 4030 capable of machining steel. There are several videos out there that talk about "upgrading" the 4030 to make it capable of machining metals. In short, an upgrade is essential to make a 4030 as capable as the machine in this video. In any case, thank you for your comment and thanks for watching. I will keep improving the design and I hope you will keep following the progress.
Design of this cnc is very Bad ! Sound of Mill ist Bad!maschine is instabil!
Loved it brooo even I want make it for my own custom parts. If you can help me out that would be appreciated. Pls
Glad you liked it. I can try to help! Please send me an email at pickyrobotics@gmail.com.
cool video man! keep us updated! there are a few things you could improve but for the price its a great machine. You learn operating a mill with time :)
You are right. It takes some time to get to know the machine. It has been three months since I built this machine. In that time, I have been able to develop a good set of feeds and speeds (I posted a video about that). I have also been able to make some decent parts with it now. I will try to post a video about that soon. Thanks for watching!
Hey I am working on a arm myself and any possibility we can be in contact? I have a few questions about making a motor controller like you did
Sure! Please send me an email at pickyrobotics@gmail.com.
Dobra robota fajna maszyna BRAWO :)
Thank you. You are doing some amazing work on your channel too!
@@curiosity-creates thanks ;)
so coooll, makes me feel very siilly and inexperienced lol, great stuff man
Glad you liked it. Silly and inexperienced is my middle name! But we have to start somewhere. Lets keep exploring and learning.
What was the total cost of making?
This is actually a slightly tricky question mainly because I had some stuff with me that I could repurpose for this project. That is also why I did not include a cost breakdown in the video. But I ended up spending around $800 to $1000 on this. Majority of the cost I incurred was spent on sheet metal stock for the metal structure, the spindle and VFD, and the table (which was surprisingly costly) to put this on.
Here are a few things I forgot to mention or didn't highlight as much in the video: 1. The robot needs to be tested further for its repeatability and weight carrying capacity. 2. I still need to work on implementing PID control for each joint. 3. If you look closely, the robot wobbles a bit as the joints are accelerated. This wobble is coming from the base of the robot not being bolted properly and the link between joint 1 and 2 needing design improvements. Any other suggestions or improvements, please let me know in the comments!
I see, you watch blondie hacks haha
That channel has some really good content and I watch it every now and then. But I am curious to know, what made you feel I watch the channel?
@@curiosity-creates I gave a comment on your cnc build video. Check that out.
Sorry, i don't see a comment from you on the CNC build video, @Manu2126.
Ah, here it is! Really awesome!
Thank you!
Very interesting project, I would recommend to use the strongest stepper motor for the x axis as it moves more mass and has to vercome more forces than the z axis motor.
Thank you! X axis does moves more mass but the Z axis is working against gravity. During my tests, I tried to see how fast I can move each of the axes. Despite having a bigger motor, Z-axis starts skipping steps at 1800 mm/min. The other axes (X and Y) can handle these speeds just fine.
Very impressive build and an interesting video! I immediately subscribed and I'm looking out to see what's next!!
Thank you! Next video on this is coming up very soon.
Very informative video. THANKS. Do you have the model numbers of the Nema motors you used?
Glad you found it informative. This is the one that I used in this video: 17hs24-2104s
A new video about machining feeds and speeds for DIY CNC machines (or routers) has been posted on the channel. If you want to see the machine in action with improved setup and parameters, please watch the video linked below: th-cam.com/video/i9bFZir_LxM/w-d-xo.html
Your videos are great! Have you gotten to replace the 3d printed bracket for the spindle?
Thank you! Yes, I have replaced the 3d printed bracket. All the results you are seeing in this video are with a metal bracket.
Very nice work, enjoyed watching, BUT pls don't clamp work like that, that's the root cause of this chattering and your piece is actually moving being pushed by mill, its not clamped strong enough. So either use some good vise, or clamp by the edges on sides so it can't move, not on top of it like that.
Thanks for watching! Yes, you are right. The clamping was makeshift in this video. I have since transitioned to a vise or better clamps depending on the need of the job. The improvements are very apparent in my other video about process parameters.
Got money, electrical and mechanical engineering knowledge and 0 knowledge of actual machining. Typical. Cool build but always people trivializing a vital part to making...parts.
There is only so much you can put in one TH-cam video. There is another video on the channel that does talk about machining parameters. But what you are suggesting is a great idea for a future video in which I make a part from start to finish. CAD->Tool path planning->Machine setup->Actual machining->Part inspection. I am already seriously considering it. Thanks!
Could you summarize if the use of belts is still popular with robotic arms in factories, and what are the advantages compared to cycloidal or harmonic drives?
Belts are still a popular choice for industrial robotic arms. The choice between belt drives or cycloidal and harmonic drives is really a matter of design preference and application. Belt drives provide a couple of advantages. For example, they are relatively easier to manufacture and can be a bit more forgiving to imperfections in manufacturing, they are lightweight and can absorb shocks better. Cycloidal or harmonic drives on the other hand need to be very precisely manufactured. So they can be expensive. But they can be relatively more compact and can provide higher reduction ratios. These are just a few pros and cons.
Very good explanation !
Glad you liked it!
Very helpful, thank you for posting this video!
Glad it was helpful!
Great breakdown. I like that you take the time to actually talk about the theory and reasons behind the numbers. I think it's important to consider the mechanics of what's going on and make a decision based on that (along with the material, setup, tolerances, etc) than just relying exclusively on rote adherence to a chart.
You are absolutely right. Each machine and situation is different. The fundamentals behind the process are the key and that is what I am trying to get across in the video. Thank you!
Good Explanation. To the Point.
Thanks for watching!
Why only one cart per rail? Especially for the Y axis. Picture where your center of gravity is, consider how much mass is there, and think of the forces on those linear bearings just as it sits. Now imagine it milling and the forces that produces. Yes, you lose a little bit of travel but the carts are cheap and you should be running two per rail. EGW rails have shortened carts too btw.
Very valid point. I might not be able to make this modification in my next set of upgrades but I will incorporate this somewhere down the line.
I can't believe it works at all!?!?! so top heavy and only 2 bearings! crazy! I guess the saving grace is the fact that force is at the same level as bearing so not high leverage ???????
What were the linear stages that you used for the x and z axis?
Here is the link to the ones I got. I have also added this link to the video description: www.amazon.com/dp/B07T89VFNP/ref=twister_B07YGLNCVS?_encoding=UTF8&th=1
Nice work mate, would love to see how it goes once the metal bracket is on. Would definitely recommend you 3d print or waterjet cut and bend some way covers for those rails though!
Thank you! I will surely post a video with updates. And yes, a cover for the rails and the ball screw is definitely on the list of upgrades.
hello, isn't better to use servomotors?
Not really. Servomotors, even the best ones, have backlash. Something like this is useful in applications where backlash needs to be kept at a minimum.
Perfect explanation. Thank you!
I am glad you liked it. Thanks for watching.
wooohhh ossum brother wating for your next video
Thank you! Next one is coming up soon!
Thank u for sharing, style of explaining the topic is perfect. i understand to N.m at now
Thanks!
That’s a really interesting design and I enjoyed your thought processes on the journey to this solution. I guess if you ended up routing power to these actuators, it would still be a much simpler task than the multiple connections it normally takes. Thanks for sharing.
Precisely. Worst case scenario is 2 power cables. Best case, no cables at all! Thank you for watching.
Great video! Just subscribed- I appreciate how you showed all of your design process and not just the "Right" answer!
Thank you!
very interesting, thanks for sharing, but the bigger problem is what are we going to do when AI driven robots make most of us useless?
Thanks! Haha. Then we just make better AI driven robots to make the older AI driven robots useless!