LinkBack URL
About LinkBacks
not wanting to rain on your parade but
theres a number of problems and some rather large that will tend to differ from the animatronics methods and actual robotics
firstly animatronics designs tend to only light weight designs
and second the looks are more important than power
for the legs of an insect style robot , using mine as an example
the leverage on the leg to lift it is transfered at only 50mm offset from the pivot joint , and the lenth of the leg is 900mm , with a weight of each leg at about 3Kg thats a lot of force needed , the motor is a 1A 36v motor it has some gearing before driving a threaded bar , and each jack can provide enough force to lift about 100Kg (220lb) at its output it also and more importantly for this sort of design has a very strong holding torque of 4500N so when the motor is off
there is no possible movement or slipage of the position held (it would shear the 6mm steel pivot rods first) although 4 * 12V 20Ah (8Kg) batteries are only going to give me a 45 min run time
i am expecting some more of the jacks for my robot turning up eather today or tomorrow at the moment i only have 3 and expecting another 4 leaving another 11 to get , so i should be able to get them fitted on 2 of the legs and get a pic or two posted , but they tend to hide well ,, the design i am using would only need minor mounting point changes if i was using hydrolics
radio control servos are not realy going to handle the torque for the legs and have little holding torque also , the wire method may need to be stronger than you think when you had the weight of the robot and all its on board equipment to it also , there ideal for tiny insect style robots , but they don't have the power for long legs
if using stepper motors driving a gearbox for the pullies / spools for the wire then maybe , but
that makes it bigger , heavier and needing more power so bigger batteries and in turn needs bigger stepper's to give you the torque until you reach a working balance again , and no matter how i have looked at the problem
electric drive motors don't provide the power , if geared down to provide the power then fail to provide the speed , and wormdrives are the best option as they don't slip
if you look closely at the second design the legs seem to use a phenumatic/hydrolic type ram in them , although it would need to be a bit more substantial than the ones it shows there is scope for that
the actual size of the ram is not the point its the presure it can withstand and the force it gives that is more relivant , a smaler cc ram needs less time to fill so should move a lot quicker , even if using another servo driven ram inside the body to power it from the energy transfer is good and rather easy to do ,
another problem is 18 servos (the high torque ones ) like the futaba s9402 retail at £60 thats GBP each so £1080 for the servos alone
only provide 8Kg/cm so would limit the construction weight to rather thin aluminium or plastics for the most part
and finaly with all 18 servos positioned inside the body (40mm*20mm*37mm each) and needing suitable mountings , associated pullys / spools and routing space for the tendons its going to need a lot bigger body area than you think ,
based on 18 servos thats 3 per leg , one for hip forward/backwards one for leg lift and one for a knee joint , the 2nd design you gave the link for actualy has another 2 joints on the leg , that would require 30 servos
and could bump the cost up to £1800 (GBP) just for servos if that method was used
although i do admit those designs on that link are cool , he does not have the problem of it being able to cary the weight of the servos/drive motors as well as the power supply for them
as to cost though
a ram would just invlove the materials and machine time , and less than £5(GPB) in materials , a servo although pre assembled will cost more for ones with any real power
a simple cheep servo costs about £5 (GBP) that would have no problem driving a home machined hydrolic valve and a suitable hydrolic pump driven by eather a low cc petrol engine should do the job , its not like it would need to be a excivator grade hydrolics system
on a side note and i seem to have discused this elsware recently
quite a large percentage of the scifi robot designs around are impractical or impossible to manufacture a working design , a large percentage would not walk as there COG is in the wrong place , and most fail on basic factors like
theres no room to put suitable batteries or power source , its ideal for scifi
and using CGI for films lets them get away with not having the phisical limitations but it is no good if you want to build one
the best example is the normal "mech" design as you can see from
they had to cheat and put wheels under the feet as theres no practical way to get the balance for it to lift a leg from the ground
and as you can see from this intro
the CGI ones don't tend to comply with the laws of phisics about balance
have to add that the real full size one would make a bit of a mess if it fell over , not to mention that it would not be able to get back up under its own power
so if you go for a 800mm wide aluminium scorpion then your going to be looking at about 4 to 10kg total weight , including batteries and servos , and metalwork for a 20 to 30 min run time like as a design like say
http://www.lynxmotion.com/images/video/closeup1.mpg
that is light weight and lacks anything but the basic walking frame and it has short legs at that. but if loaded with other parts the servos would stall and it would fail to be able to walk with the extra weight
estimate you may get the tail and 2 grippers using the same construction techneque's they used for the rest , but very little extra
Originally Posted by medicontherun 
It still hurts inside, but im thankfull i had the oppertunity to build it with some friends.
