Rectangular 3DOF hexapod

Hi people,

I have problems with tripod gait on my Lynxmotion 3DOF Hexapod with rectangular shape of body. I’ve written the code and when the robot is on top of the platform (all legs in the air), the gait seems like perfect and smooth. But when I put the robot on the ground, it makes a step or two and then looses stability and falls by it’s front part of the body. The thing is, I am using BotBoarduino (arduino duemilanove) and SCC-32. Both controllers are attached on top of the body: one is on the front part and the other is on the middle part of the body.
It seams to me that the center of mass is moved a bit because of the position of controllers. Rear legs are able to raise and ‘try to make a step’ but front legs are just unable to raise and move.

Did anyone have a problem like this? How to handle this problem?

What size servos do you have? Where is your battery?

Alan KM6VV

It sounds like the servos are not perfectly calibrated. This is done in two parts:

  1. Mechanically (which it sounds like you did)
  2. Software (this is done because even the mechanical alignment can be off a bit)

Can you confirm if you followed step 2? If not, we’ll need details as to which robot you have.

You can see the specific kit here: robotshop.com/en/lynxmotion- … -only.html
I am waiting for battery to be delivered, I don’t have it yet, the needed power was supplied from PSU.

I did calibration, mechanically. By using Servo Sequencer Utility I noticed certain offsets which I took into account while programming. Robot is able to stand still and all the legs are parallel (as far as human can notice). As I said, the gait looks fine when legs are off the ground but when the robot is on the ground, after several steps front legs bend and the body falls to the ground, while rear legs do not bend and are still able to rise.
The software calibration I did was in accordance with my observation (subjective perception) and it seemed ok. Is there a way to do it precisely?

So do you have the 645 servos? The $200 kit doesn’t have any servos.
perhaps you need to simulate the weight of the battery, as you are not yet using an attached battery?

Alan KM6VV

I do have servos. The kit I got came with servos (but the robot looks like this one). Good idea about simulation. Thanks!

Try your very best and use whatever tools you have at your disposal to get the legs 90 degrees and 180 degrees.
Can you post a video?

I am not quite sure how to help here? Not really enough information to help. Pictures help a lot. Likewise code can help…

Is it the BH3 or just looks like the BH3? Which servos? Could easily be that the servos don’t have enough capability for this. It is sort-of interesting that on the Lynxmotion site, the page talks about using the Hitec HS-475/HS-485, but I think all of the kits now come with HS-645mg which personally I think is a better choice. (Also side note: the top level talks about the control program written by Laurent Gay, which I assume is talking about PowerPod (Legacy software for Basic Atom Pro-28)…

You don’t have battery yet, but running on PSU? what PSU? How much can it output? What you describe also sounds typical of being underpowered. It has enough power to run it when not under load, but as soon as you add load, it can not handle it. Related: Wiring can also cause this. That is if you use something like the wires of the 9v connector or the like, it can not handle the current necessary to work.

Good Luck
Kurt

I will post a video on Monday, I don’t have the robot at home. Thanks!

I will film the robot on Monday and post it here, I don’t have it at home.

It is BH3, it’s just the kit I got includes servos + hardware.

I am not sure, didn’t pay too much attention since they came in the kit and should be proper for that robot. I think servos are HS-485 but I will check out on Monday and let you know.

This might be my problem. I think the PSU I used can provide enough current for servos but I used regular jumpers for connecting the power on the boards, like those:
100randomtasks.com/_/rsrc/12 … 0Wires.JPG

sub.allaboutcircuits.com/images/55016.jpg

The current taken from PSU varies from 100mA to 2A sometimes. You are probably right about this. The wiring is probably inappropriate but those jumpers was all I had available. I will try to get proper supply on Monday and see if that’s it.

Thank you!

Yeah, wiring is too light, restricts the current. Only measured 2A?

I just grabbed a LiPo battery, and it has 13 AWG wire (odd size).

18 AWG insulated (on my 6V 2800 ma 'quad battery) should be fine.

You will probably see more current being drawn with the larger wire.

Probably need the 645 servos for best results.

Alan KM6VV