Here is the latest AR4 robot: • AR4 MK2 6 DOF ROBOT AR... www.anninrobot... Main Project Video: • AR3 6 Axis Robot All gear drive motors and drivers available at www.omc-steppe...
Chris, very nice work and awesome follow-thru on all the documentation for a complete project. I've been in the machining world for the past 40 years and have many CNC Mills and lathes, and this is a direction that I've been looking going to on some projects for lights-out operations. With the incredible amount of effort you have put into this project, it would be a shame for me not to build an army of these bad boys in the shop. :-)
Out of curiousity. Have you considered adding encoders not to motors, but to joints, so if there are any mechanical slop in belts/gears - it culd be compdnsated right away. Asking this because I am building CNC and I see people in hobby and industry using closed loop motors, but almost noone uses glass scales(like DRO on lathes/mills) as encoder. Why is that?
It depends what you hope to gain by closing the loop with a glass scale. They can certainly be more accurate but that doesn't translate necessarily into a better finish part. External encoders work best an extremely rigid systems such as linear Motors where the load is directly coupled.
@@adisharr I just have a few scales from DRO of decomisioned lathe and I found a controller board that supports linear scales. But still - I have a feeling that I am missing something...
@@scottjackson2812 so basically - noone uses them because it is overkill. If you have backlash problems on new machine that need a linear encoder to solve - you have other problems to worry about. :D Altough - linear encoders should be a good idea when retrofitting old mills to cnc.
Hello Chris Annin, Firstly, I’d like to congratulate you for the project. It is a really interesting and high-level electronic device. Cheer you up! I work as a technical maintainer in the automotive field, particularly in the painting line with arm ABB and as a hobby, in my free time, I’m trying to create my own arm robot. I’ve been observing Python’s and Arduino’s software but, because of the lack of comments, I don’t really understand the algorithm. Then, I’d like to ask you please if you could show me any flow diagram of the software to see how you carry out the L and C moves, the accelerations, and tell me the reason why you don’t use control PID. Thank you very much and I encourage you to write a digital book where we can easily buy!!! Best wishes!
Have you considered resonance compensation? Like how klipper does it but on a robot. Perhaps you could reach some really unprecedented speeds if you compensate for the jerk/bounce.
You can kill the servo squeel sound by using more look up table entries for the sinewave generator, or, if you have the computational horsepower free, just do the sine/cosine calculation in floating point.
That thing could be fixed onto a sliding track, as it takes clippings of plants to replace the delicious harvested ones, placed in a grow room of course, that you put your weed in. Yeah man yeah! Suppose it could have other uses though. Thanx for sharing, and inspiring.
Could I extend the lenght of the arm while using those same motors without overloading them? by how much you think? I don't really need the arm to be fast, although it would move slowly for a long period of time
Hi Chris, Great work! Do you think the aluminum can be replaced with stainless and he steppers upgraded to increase speed and accuracy/precision? I think this is a great project and I would love to apply it to low-volume pharmaceutical assembly and filling (where it's usually cost prohibitive to buy a specialized machine). I think it would be a great study to figure out cost tradeoffs and practicality (capping, filling, cartoning, labeling/inking, pick-and-place, inspection, etc.). Low-volume contract pharma is an interesting design problem. I would love to hear your thoughts! -Dean
Hello, yes It could certainly be made from stainless. I think to upgrade the motors it would add alot of cost but it could be done, I dont think it would be an easy change and might take some redesign. Please feel free to contact me with more ideas and questions on the website or forum. I try to reply to everyone on youtube but sometimes its easier to keep track of ideas and questions through the site. thank you. Chris
I have followed all your videos since 2017 , I wish to do the same robotic arm ,one day, with some improvements of course, as a hobbie, but my problem is to find the materials and components , here in Peru , I made a similar program AR3 in pure basic program , python creates a too much heavier .exe , regards
wat is the cost of aluminium cnc ? quoted a whopping RM19k (Malaysia ringgit RM4.56=1USD) coz here they cnc each part from one block of aluminium which i think is very unwise.
Not necessary ....if you are going to use it on high speed with a good amount of load then consider using one for feedback. If speed isnt that high and load is very less compared to stepper holding torque then u dont need one. Encoders also help in 3d reference inputs but you can also use other sensor instead of it like mpu or pot.
I’d like to know if you think an AR3 could produce four “no-roll pie crusts” within 30 minutes. Before discussing the tentative functional requirements, here’s a whimsical introduction. It’s been over 50 years since we put a man on the moon. Now some billionaires are actively working on sending people to Mars, yet here on Earth, billions of people are still toiling away each day to accomplish simple yet critical chores such as food preparation. Eating delicious, fresh, hot, homemade pies is a pleasure, yet making homemade pies is time-consuming. Elon Musk and Jeff Bezos undoubtedly can afford to have skilled pastry chefs bake delicious, fresh, hot, homemade pies for them as they conjure up images of earthlings venturing throughout our solar system. As for the rest of us, well, we could use a little robotic help. Or as I imagine Elon and Jeff might say, “Let them eat pie!” With that introduction out of the way, let’s get down to the business at hand. I would like to be able to measure out, say, 1600 grams of wheat flour, water, cooking oil, honey, salt, and yeast into a mixing bowl and then-say, within thirty minutes (that is, not very quickly)-have an AR3 produce four “no-roll pie crusts” in four adjacent glass pie pans. In case you are unfamiliar with “no-roll pie crusts” please watch the following video... *"Easy Shortcut No-Roll Pie Crust"* ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-bfHvG8PGIoY.html&t=425 for approximately 25 seconds, from the 7 minute and 5 second mark to the 7 minute and 30 second mark. The proceeding link, links to the 7 minute and 5 second mark of the video. Do you think your current iteration of the AR3 robot would be able to… 1. mix, say, 1600 grams of wheat flour, water, cooking oil, honey, salt, and yeast which I would have already placed into a mixing bowl into dough and then... 2. measure out a specified weight (for example, 400 grams) of that dough into each of the four adjacent glass pie pans and then... 3. transfer that 400 grams of dough into each of the four adjacent glass pie pans which I suppose might have previously been placed into a heavy cast iron frying pan to temporarily yet sufficiently secure it in place (that is, to prevent glass pie pan from slipping and sliding) and then… 4. evenly distribute the 400 grams of dough in each of the four adjacent glass pie pans thereby creating a “no-roll pie crust.” Do you think the above would currently be feasible with the current iteration of the AR3?
no problem inverting and mounting on the ceiling. I haven't tried on a wall. J1 wasn't designed for holding the weight side load so Im not sure that would be a good idea,
