Outdoor footage of MorpHex and a servo failure

Two days ago I did a second attempt to make a new video of MorpHex outdoor. This time I wanted to see MorpHex rolling down a short slope. I think this is the first time I've ever seen a hexapod robot rolling downhill. The outer sphere sections protects the inner parts of the robot very well.

To be honest, not everything went ok during the filming. 5 seconds after 3:31 on the video the inner body servo broke down!

I've had a bad feeling for this servo for a long time now. I did think of replacing it when I made the MKII version, but for saving time I didn't.. sigh..

This was a major set-back since my plan was to demonstrate some other features as well. Currently I'm home alone, because my wife and our 3 kids has already left for Easter holidays. I'm going to join them tomorrow though. This meant I got much free time for playing around with the new Teensy3.1 on a breakout board designed by Kurt Eckhardt. I'm thinking of using this board on several of my robots in the future. The Teensy3.1 feature the latest ARM Cortex M4 processor. Also having connectors for the Robotis servos, a holder for XBee RF communication and a speaker make this board just perfect. Thanks Kurt! You can read more about Kurt's Teensy 3.1 breakout board on the Trossen Robotics Forum.


But instead I had to spend about 10 hours in my workshop for resurrecting MorpHex. sad

On this picture (from the bottom side of the body) I've removed the LiPo battery holder and marked the white Nylon gears just to be sure..:


Removed the "body-arms". As you can see, the screws that hold the Hitec HS-5645MG servo is a real pain to reach:


After removing the servo I opened the servo case just to confirm that one of the brass gears was broken:


Luckily I have several spare Hitec HSR-5990TG servos. They are much stronger and feature gears of titanium. FIY, the Hitec HSR-5990TG servo are discontinued. Some years ago I bought about 40 of them, used but in good shape and for a very good price. 

The new servo in place!:


So far so good.? But no. At this point I found another problem. You see, the main reason for why the servo broke down was that one of the screws that hold the pairs of "body-arms" loosened after a while. To improve this I had to remove all the arms again and also remove the upper sphere section and the ARC-32 mcu board. Believe me, this made me pretty frustrated. What a mess!


As you can see from the picture below, the screw marked with a red circle became loose and came in direct conflict with a fixed part of the inner body. On the head of the screw you can see some scratches, actually this was more visible on the fixed body part of aluminium (no pictures of that). The solution was to use inset screw (the red arrow) and using locktite this time:


I don't think I've explained this before. The upper section are mounted to the inner body part using only three 4M screws (marked with arrows and a circle on the next picture). All in all, its a pretty simple task to mount.



After assembling everything together and recalibrating the inner body servo, I only had to do some minor changes to the code and MorpHex was working fine again! smiley

My next upgrade might be the new Teensy3.1 board. But that require a lot of work on the code, not sure if that is my priority #1. I've some plans for a new robot too..


I wish you all a great Easter holiday!





  • Morgan Gillespie says:

    I'm a mechanical engineering senior at Brigham Young University in Provo Utah and I just have to say that your work is incredible impressive.  I enjoyed it so much that I am presenting your MorpHex MKII robot tomorrow in a graduate robotics course. 


    I hope to see more updates on the MorpHex bot in the future and wish you the best of luck with its development. 

  • Micha Lucas says:

    I was amazed by your ant inspired robot. The fluidity with wich it moves is just amazing.

    Im currently doing a course for sportmasseur, and i get alot of anatomy lessons. I couldnt help but wonder how much more you could do with your robots if you would incorporate more anatomy-inspired parts like muscles, tendons and such. 

    I am as of now a big fan of your work and could only wish i had more knowledge of robotics so i could build things like this myself. 


    Kind regards,

    Micha Lucas,  the netherlands

  • Really cool!

    Out of curiosity, how does the spiral-looking thing work? Does the middle servo rotating it causes it to "expand?"

  • Jon says:

    Looks so lifelike for a robot and like something out of some sci-fi movie. Nice work!

  • Alex Ivanovs says:

    This is really exciting, and I had a good laugh from how it was able to accomplish all of those tasks; really great job and wishing you good luck in the future :)

  • Bernhardt says:

    A new way to discover the Mars!?

  • Faycal says:

    check this out


  • Ramon Valk says:

    I was so happy to see a new post, love your work!

  • Andrew Love says:

    You're on Reddit!


  • Jean Larrieux says:

    Well, I love robotics and I have a background as a mechanical engineer with strong emphasis on programing. I am wondering if you couldn't use object oriented programming techniques so that your programs would be processor agnostics. Again, I am not sure if you use the same language for all the different processor as I don't have enough of a robotics background beyond some parallax robotics kit.

  • Mike says:

    Absolutely fantastic.  I'm an engineer with 25 years experience and design quadcopters for fun.  Your robot is the best that I have seen so far.  Absolutely fantastic.  Great work.  How creative.


  • Juri says:

    Your robot is amazing! Imagine if you would add 6 hexapod engines to the other side, the ball could also fly like a hexapod! Much work but … someone must be the first one :D .

Leave a Reply

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>