Humanoid Robot 3D Printable


#84

Here it is all assembled and wired. The OS and Code has been installed. When engaging with the robot I get an error. I checked all connections and it seems my Arbotix Pro Robocontroller may be bad. I just emailed Trossen Robotics asking for assistance with the troubleshooting of this in case it has to be repaired or replaced. So now it seems this may be on hold.

Here is an image of it fully assembled.

Just in case anyone else here may have experience with this controller here is the message I am receiving.

image


Interfacing a Rasperry Pi 3B to Arbotix-M Robocontroller
#85

That is an amazing amount of work!!! Very well done!!! :+1:t2::sunglasses:


#86

I received a reply from Trossen Robotics Support today. Here is a direct quote on some of the reply:

There is no direct replacement for the Arbotix Pro board, we stopped producing this in 2016 and carried inventory up until 2018 to accommodate those people that did purchase the OS1 platform. We still do have a forum for users of the OS1 http://forums.trossenrobotics.com/forumdisplay.php?140-HR-OS1-Development-and-Discussion and we are still hosting all of the documentation regarding the OS1 https://learn.trossenrobotics.com/38-interbotix-robots/164-hr-os1-humanoid-overview.html , we’re just not developing / supporting the platform anymore.

They have one spare Arbotix-Pro that they are sending me free of charge to help out. It s not new but they said it appears to work. It is supposed to ship today. Actually, as I was writing this I got the tracking number. It will arrive Thursday.


#87

So I am working on rev 2 a bit while waiting for my Arbotix Pro. Let me know what you think. The transparency is just for me to see inside while designing. I’ve added room for a 40 mm x 40 mm x 10 mm exhaust fan to help release heat from the RaspBerry Pi 3B and the Arbotix Pro. That is why it’s not fully enclosed at the top level, the heat needs to escape otherwise the RaspBerry Pi will throttle if it hits 70c.


#88

Here we hav the new RaspBerry Pi Mount printed. This is just for test purposes, to be sure all holes work as designed and all wiring fits. This will also allow testing the fitment of the fan.

Here is a time lapse video of it being printed.


#89

Here we have the Arobitx Pro Test mount. This is a time lapse video of it being printed.


#90

Here we have it partially assembled. The fan fits well. The Arbotix Pro fits, the RaspBerry Pi fits. I am also able to hide a good amount of the wiring for a cleaner look.


#91

Here are a couple new images.


#92

That is looking great!


#93

Thank you. The original design was just to get the concept of the 3D Printed parts to assemble and move as designed. That seems to be the case when I move everything manually. The second phase or REV 2 is to enclose it as much as possible while maintaining functionality. I am hoping the Arbotix Pro replacement (Due in today) will work as expected, just so I can finish this concept. Then I will start learning the methods mentioned about I2C.

I also left the Raspberry Pi connections open in the back so that I can connect to it if needed. But I will ask have SSH and WiFi enabled. It will also be upgraded to use the Raspberry Pi 3B + today. So it will have a faster CPU.

During all this my goal will be to minimize the visible wiring where I can as well as a method for restraining visible wires so they are not pinched in the joints. I am liking the results so far.

Any if anyone is curios, yes this is still 100% designed in Shapr3D. I did not use Fusion 360 on this design.

I hope to be able to do some test movements this evening or at least be able to stand under power.


#94

I keep forgetting to ask you, how long did the last two boxes take to print?

Also, is it better to have just the hole and tap the parts later or do you print the threads?

I’m really close to doodling a robot mostly made out of 1/4” thick wood. There’s a lot out there already.

I’d start off with a small scale model and work up…

Well, that’s the thought anyhooo… :crazy_face:

Oh and super work by the way!!! :+1:t2::sunglasses:


#95

6 hours for 1 and 4:45 hours for the other. This is printed with .2 micron resolution, 3 walls and 40% infill. I also print a raft, bridging and supports under bridging where needed. This allows the parts to be pretty strong, but still keeps them fairly light.

No threads at all were created. All holes are either through hole with a nut or were modeled .3 mm smaller hole. This allows the threads to self tap as I screw them in. So far that works well. However it probably wouldn’t hold up to being taken apart and put back together multiple times.

It would be cool to see what you come up with.

When I create my robots I figure out what hardware and electronics I will be using first. Everything is built around the servo motors / actuators. Without that info first there is no way to judge size and where joints are needed. This is important if I plan to actually make it.


#96

Thank You for the info!!! Greatly appreciated from someone on the outside looking in…

I like your approach with threads and hardware…

Another thing I’m playing with and laying out is rather than servo motors, I’m playing with some air mechanisms that have some neat potential for mechanical movements…

Thank You again!!! :+1:t2::sunglasses:


#97

Pneumatic is a great way to build robots as long as you have a constant air pressure to hold joints in place with some force. The other concern is backlash. But the robots would be fairly large I think.

Electric motors and ball screws are another method, it solves the holding power and lowers the backlash concerns too.

Then you have worm gears which can have holding power. And sizing along with gearing can add torque.


#98

Are you looking at pneumatic with piston/cylinder or air muscles?

I’ve always wanted to have a play with air muscles.

The motor/ball screw approach is really interesting. James Bruton did some interesting stuff on his robot dog project using that approach, worth a look on YouTube.


#99

The Raspberry Pi 3 B+ just arrived. This will be going in this robot.


#100

Hello Carl!!!

I just started checking out James Bruton on YouTube … That guy is absolutely great!!! I haven’t found his segment on his motor/ball screw approach yet…

As far as piston/cylinder or air muscles, I’d say both but mostly air muscle…

Here’s a link to a gentlemen that has utilized air power quite extraordinarily in my opinion … I very much like his approach fundamentally and mechanically…


#101

As promised Trossen Robotics sent me a used Arbotix Pro. The good news is the Raspberry Pi recognizes it and I can launch the software needed. The bad news is that it doesn’t recognize any of the servo actuators. So it seems I have 2 dead Arbotix Pro controllers. The other bad news for me is they don’t make these any more and no one has them for sale anywhere.

I will play around with it some more tomorrow. But it looks as though I have to start searching for other methods of powering this robot to move. It may be I2C as suggested. I was in high hopes that I could just use this and complete this project. But it seems that is not going to happen.


#102

That is truly amazing!!! I am speechless.


#103

That is a real shame. If you decide to go I2C, I am happy to help wherever I can.