Servoless leg design

[circuit]

I have been watching all those hexapod videos on YouTube for a year or so. During that time I've noticed that all of those hexapods are very slow. Also, servos are really expensive. So, Why not use a DC motor from mobile/pager virbo?



Lemur and tibia are 60mm in length, each joint has 100° of freedom.
I'll use DC motors from old pagers, those are 7mm in diameter. Legs will be laser-cut from 2.5mm and 1mm stainless steel. I also use 3x7x2.5mm ball bearings. No sensors drawn yet.
Hope those motors will have enough power...

Any suggestions/comments are welcome.

[MetalSkullRobotics]

There are some problems you will face.

1. The motors torque will not be enough to power the Femur and Tibia. You will need to have a extra gearing on the head of the motor before you connect these to the spindles.
2. The controlling of the motors need to be done with a separate circuit which can regulate the speed of the motor (esc of some kind).
3. The reading of the positioning of the legs have to be done with separate "sensors" which you have to monitor, this will be some more programming (where as the normal programs for a pwm output signal won’t be useful unless you have separate servo boards for each motor).

Even by a simple calculation around the power of the motors you can see that the motor has less power than a standard 3kg/cm servo. And then I’m not even talking about the bigger servo's. If you want to have more speed, and still need the same power (which you do because the servo's most hexapods use aren’t overpowered) the output power of the motor will need to be more.

Its a good concept, working with linear motion systems to move the legs. I myself also had a look around using linear motion systems as muscles for the legs, but after some calculations I decided to use the normal servo's.

[paulp]

When you say slow, I'm not quite with you. The walking pace is usually slow simply because the programmer thinks it 'looks' better at that speed.

If they are run at higher speeds they can cover distances very quickly but they tend to look a bit manic. Also higher speeds mean higher wear and tear on the servos.

By the time a design is finished, the average hexapod tends to 'acquire' weight (usually 2 to 3Kg or more) so fairly powerful servos are needed usually with some form of bearings.

I've not looked in detail at the type of motor you are talking about but I don't believe they are very powerful at all and have little or no bearings built in.

To produce any real torque you will need a substantial gear-head. you will also need some form of digital or analogue control over the motor and feedback for positioning. By the time you have perfected this design, I feel certain you would have far exceeded the cost of a reasonably priced servo.

If you dismantle the average servo, all you will find is a small motor (about 18mm in diameter) a high reduction gearhead and some form of digital or analogue control, essentially what you are designing. Someone on this site was talking about servos for $5 a piece. Probably not the best quality but you can't buy a decent drink for that much.

[paulp]

DOH, beaten to it by MSR...lol...

[MetalSkullRobotics]

Hehe, yust delete my post then

[paulp]

Nope fairs fairs, you type quicker than I do

[MetalSkullRobotics]

By the way, the servo's that paulp mentioned are probably the ones I use. 3 dollar each, shipped in from China. I ordered 60 of these and when these arrive i will certainly post my findings about them. Like paulp said the quality will be very low, and the power that they should produce (3kg/cm) won’t be precise. But like I said elsewhere, I'm just using them to test the simulation of the walking movement.

For around 10dollars each you can buy nice 12kg servo's. Have a look on http://www.bidproduct.com or http://www.unitedhobbies.com if you are looking for some affordable servo's.

[paulp]

All that aside, I love the design. It looks like industrial engineering in miniature.

I've been sketching some ideas about putting all the servos inside the main torso and using a pushrod system to drive the legs. The mechanical arrangement of your leg would lend itself to such a design.

Let us know how you get on...

[MetalSkullRobotics]

Sounds interesting. Any possibility to upload these sketches?

[paulp]

They're only on paper at the moment, When I commit to CAD I will post them...

Anyway, shouldn't hijack this guys thread anymore. Sorry......