Orion Robotics FireAnt:

The modified FireAnt abdomen 2DOF joint:

Compared to the original A-pod:

—

Initially, the 2S 2200 mAh LiPo battery was ment to be placed under the mcu boards inside the main body. When it comes to LiPo batteries I like to charge them in a safe area. This mean that it should be easy to replace the battery. By placing the battery inside the abdomen, replacing the battery isn't a problem anymore and you get more free space under the mcu boards for servo wires etc. I've also used some shorter standoffs for lowering the boards. The new digital HV-220 servos are fitted with a very long servo wire and thats a relief when it comes to the tibia servos. That also makes it easy to guide the wires without the need for extending the wires.

I'll take some better pictures of it when I've uploaded and tested the software. I'll post more information about the mcu boards and the servos later.

As you can see I'm using two power switches; the rear switch is the main switch for both electronics and servos, the front switch is for the servos. Personally I prefer having a separate powerswitch for the servos.

The mcu boards are programmed through an USB cable, unfortunately the usb port on the DaVinci main board is a bit hard to reach every time you need to update the software. I solved this by using a mini USB roll-up cable that is permanently connected to the board. I placed the roll-up holder inside the abdomen as well:

Like I said, I'll give you some updates when I've tested the software for the FireAnt.

A few days ago I got the great news from the Boca Bearings company that my contest entry won their grand prize! More information about the contest results can be found here. I do feel very honored for getting a prize like this. I'm also very humble thinking of all the other great projects that was submitted to contest. One of my favorite projects was the human powered helicopter, a very impressive achievement indeed.

As you may have noticed I've not had much time for updating my blog lately, its mostly caused by little free time. When it comes to the MorpHex project I've done some progress on the next version of MorpHex, called MorpHex MK2. Or maybe I should call it Mark 3.. You see, I've had a few setbacks lately. I earlier mentioned that I would try to go for a fully symmetric version and using a completely different leg design. I did some progress on that attempt, but after a while I realized that it wouldn't work under one critical condition; transforming from ball to hex. Here is a rather low quality picture of a very different coxa with dual femurs I made under this attempt:

After a lot of thinking I came to a decision of going for an asymmetric design again. To be honest I don't think a symmetric version of MorpHex would be very appealing. You see, a symmetric design mean that the upper and lower leg sections share the same coxa servo. So when walking the upper sections always need to follow the lower legs. An asymmetric design where the upper section can keep the half sphere shape is something I really liked with the first version. So this time the I'm going for an upper section with 2 DOF. I do hope this will work better.

So far I'm still working on assembling the lower leg sections. I'm using a slightly different coxa design (don't have any pictures ready for that yet) and a tibia bracket that are going to make the sphere sections much stronger. A picture of some the new brackets here:

There are still a lot of work left when it comes to hardware assembling, calibration and programming.

Again, I want to thank Boca Bearing for the very generous prize! Hopefully many more robots come in the future.

-Zenta, "Having fun with robots".

But you might find it interesting since it demonstrate different gaits and high speed walking.

To be honest, my first impression of how it performed using the standard code was not good. The overall performance couldn’t match the Phoenix hexapod. Especially when it comes to smooth movement.

Luckily, my friend Kurt Eckhardt had just bought a PhantomX. Since we both have a lot of experience of working with Jeroen Janssen’s (aka Xan) Phoenix code written for Basic Atom Pro microcontroller, it became more natural to continue using this code. When it comes to programming, my skills are very far from either Kurt or Jeroen’s skills, but I love to play along! So it was a great relief that Kurt decided to port the original Phoenix code over to C for making it work on the ArbotiX board. You can read about Kurt’s work in this thread. I’ve had a lot of fun playing along!

Oh, one thing. I can’t say this enough. If there is one person who really deserve lots of credit, my friend Jeroen Janssen is the man! For me this hobby got a lot more interesting from the day he released the Phoenix code.

In Kurt’s thread I posted about some hardware modifications like giving the tibia more freedom and making it easier to replace the LiPo battery. Being able the to replace the battery is very convenient since the AX-18′s drain the battery pretty fast.

One of the first thing that struck my mind when I saw the PhantomX hexapod is that it is 100% symmetric. That combined with the servos 300 deg of travel range made it a perfect for upside down walking. By adding a sensor (accelerometer) it became a relative easy task.

A few days ago I managed to fix the balance mode. While doing this I was visiting my parents who live in the wilderness (more or less.. LOL). You see, the balance mode combined with tripple 16 gait (a modified tripod gait) make the robot very capable for crawling outdoor without the need for any sensors on the legs. So I decide to take the hex into the forest for some pics:

PhantomX together with the XBee based DIY remote:

After playing in the forest for a little while I decided to shoot a new video. (Official youtube video).

Enjoy!

I’ve found a way around the wiring issue but I might want to go for a totally different leg design too. At this moment I’m not sure.

Anyway, here is the new video:

I hope you enjoyed it! I’m also using this video to promote my April contest entry at the Boca Bearings 2012 Innovation Competition. The video with the most votes will be announced winner at the beginning of the following month. Out of all monthly winners the jury of Boca Bearings will select a grant prize winner who will receive $10000 in cash. The two follow up winners will receive their own Makerbot 3D printer.

To be the monthly winner I need your vote.

If you want to see more advanced robots coming out of my hands leave your vote here. Thanks a lot, its highly appreciated!

After my last post I found it necessary to some additional improvements, like strengthen the ribs on the lower sphere sections:

The solution was pretty simple; just a little piece of PC glued to the rib and a 3mm nylon screw for extra strength:

I’ve also done some small changes to the battery holder, relocated the main switch and mounted a plastic band to keep the battery wire as close to the body as possible:

Also pay attention to the main fuse. When dealing with LiPo’s directly connected to the main board its always a good idea  to use fuse for limiting the max current in case of an accidental shortcut. In this case we don’t want a great ball of fire..

For replacing a battery I need to remove a 3mm hex screw. Then just slide out the battery and replace it:

The battery are fixed in place by a 3mm hex screw mounted on each end of the bottom plate. I don’t want a loose battery in “rolling mode”..

These two pictures demonstrate the effect of having an expanding body.

Closed body:

Expanded body:

Some more pictures of MorpHex in different poses:

A strange form of attack pose?

“Sphere mode”:

Some standing poses:

Standard pose with closed body:

Standard pose with fully expanded body and upper sphere sections:

In the future I plan to mount about 36 RGB Starlite LED’s as background lighting. The Starlites are made by Basic Micro:

My next post will include a new video of MorpHex.. !

As you probably can imagine, this isn’t a very stable solution. Since my goal is to make MorpHex able to roll, the upper sphere sections will be needing some additional support. I’m solving this by making six special brackets.

Some details behind the process of making the brackets:

The brackets are made of PC (Poly Carbonate). All handmade using a band-saw to cut them out. I’m using a simple technique by cutting six at ones:

Using two long 3mm screws for clamping the parts together. By making six in one cut you save time and its easier to make several duplicates.

I do wish I had a CNC machine though! Would be so much easier to make even more complex parts. The next step is to mill out a track/slot. I’m using a manual XYZ milling/drilling machine:

A simple table made of rubber coated aluminium holds the parts in place:

I’m using a guidance template of aluminium to hold the PC part in place:

Milling out the guidance track:

So far so good, only 11 more to go..

I didn’t take any photos of the next process. But I had to repeat it one more time to make an additional slot and also make a little part that hold a pair of the main parts together. The final result of the brackets looks like this:

The parts are glued together. Mounting the brackets to each sphere section was relative easy. I’m using only one nylon 4mm inset screw for mounting the bracket. As you can see the brackets makes the upper sphere section much more stable and no need for rubber bands:

So far I’m very satisfied with the result. Without even doing any changes to the code I’ve managed to make MorpHex roll! So far the rolling motion is not very controllable, more like a proof of concept. But when I’m finished writing the “roll code” it will perform much better and hopefully be easier to control in “roll mode” too.

The kill-switch.

I did spend some time to figure out a good solution for operating the servo power switch. Its operated using a simple tool for turning it ON. As you also could see in the video turning it OFF is pretty easy (KISS).

Close-up of the switch:

I’m using a simple custom tool to operate the switch:

The little hole work as a hook for turning the switch ON:

That’s it for now. I’ll post some pictures of Morphex when its complete.

Then the most exciting part of this project begins.. 

Lynxmotion A-Pod:

The original A-Pod:

The Lynxmotion A-Pod kit can be bought here.

It’s remotely operated using a PS-2 controller. Visit the Lynxmotion forum to find more information about their A-Pod.

Check out the assembly instructions for the legs, body, abdomen(tail), mandibles(gripper) and PS-2 connection, calibration and code.

A video of my original A-Pod:

In the meantime I hope you'll enjoy this video: