Hexrawler project

Kotobuki

Hi Again,

Well, progress is slow on the old school HexCrawler. I have got the legs adjusted, so it stands. There are two servo controllers on the way to me and I just now ordered a 7.2 V, 3200mHA battery pack and charger. I will be getting a smaller rechargeable battery pack to run the BS2 board @ 9 Volts.

Now for the sensors. To start with, I will be using a Parallax four way tilt sensor and a three axis accelerometer.

With the above, I should be able to get this to do a simple 2DOF walk, with maybe a tilt control. I am aware that the BS2 will be quite taxed, and the motion will be jerky, but this is just the start of a long process.

Now my question to the forum...

What should the next sensor be? My main goal to begin with is one very decent gait without a lot of slop. Is there anything in the way of sensors that would help this goal? BTW, what is a 'gaitPIC'?

Best,

Joe

Mike G

I wrote the original (well all of it) HexCrawler code. You should be able to find it on the CrustCrawler site. If not, I believe the files are backed up on one of my PCs.

Tilt and acceleration may not be beneficial for walking. The gaits are pretty much static due to the design. I'm sure you can find some benefits if not for the learning experience.

Distance detection using a PING))) or sharp IR sensor or both is pretty cool. The beast can walk around avoiding obstacles. Also I suggest picking up a power supply. That way you're not constantly charging batteries during development. 12 servos suck a lot of power.

Duane Degn

Kotobuki wrote: »

Hi Again,

Well, progress is slow on the old school HexCrawler. I have got the legs adjusted, so it stands. There are two servo controllers on the way to me and I just now ordered a 7.2 V, 3200mHA battery pack and charger.

What type of chemistry are you using?

The small hexapod I've mentioned uses a 2-cell LiPo. It's pretty hard to beat a LiPo for power density. If you're going to making robots, it's worth getting a LiPo charge in order to take advantages of these batteries.

Kotobuki wrote: »

BTW, what is a 'gaitPIC'?

It looks like the gaitPIC is a slave controller that will take care of the servo control. The master control just tells it which gait to use and the gaitPIC takes care moving the servos to produce the desired gait.

Mike G wrote: »

Distance detection using a PING))) or sharp IR sensor or both is pretty cool. The beast can walk around avoiding obstacles.

Ultrasound and IR really compliment each other well.

I was very pleased how well my Scribbler 2 was able to detect objects after augmenting its IR sensors with an ultrasound sensor.

It seemed like the one sensor would work well where the other one had trouble.

Kotobuki

Thank you both for the input! The batteries are NiMH and the matching charger so the battery pack will not be damaged. I was wondering if I should get the Ping and SharpIR, so I will. Next month, I am getting the pan and scan thing from CrustCrawler. I was so thrilled when they sold me the very last HexCrawler that they had... the unit that was used as their print ad model. Andy at Parallax sent me a couple of servo controllers, one a BS2 based board, the other a Propeller. I have a BOE with the BS2 that will be used just to get it walking. Then I will use a different controller, most likely a Propeller. I do want to experiment with distributed command and control however, just like an insect uses. Six tiny little controllers at the base of each leg, and one slightly larger one to coordinate the six others. Then a separate feedback loop, just like in a biological system. But that is a long way off. Right now, I just want to get it walking so that the kids do not get bored looking at it. If there is an arm that is light enough, I think that it would be hilarious to mount it on the top of the crawler, and program the thing to walk up to a dog and give it a Milk Bone doggy treat.

Thank you again!!

Best,

Joe