Building humanoid similar to Asimo and Qrio

[tempusmaster]

indeed we'll order some alternative wire. i'm using Inventor. seems like we managed to get enough friction between the wire and the pulley for prototyping but I'll order some of that wire discussed previously and also look for other inelastic wire materials such as kevlar.
Quite difficult to find online industrial kevlar wire with strength and width parameters.

You might find high test braided fishing line to be a good alternative -

http://www.basspro.com/PowerPro%C2%AE-Braided-Spectra%C2%AE-Fiber-Micro-Filament-Line-300-Yards/product/44283/-920527

That's what's used in my RepRap printer to drive the x/y axes. It has great performance characteristics and doesn't stretch.

[limor]

Thanks for the link. I think that what I bought is similar (braided, strength, less than .5mm etc).

So googling came up with the Tantillus RepRap which indeed uses these wires instead of expensive timing belts.
Thanks for this info! It is good to know that this is has been used reliably by others.
So what is your experience with clamping, slippage, knotting and tensioning of these wires?

[Tyberius]

I sincerely apologize for that post then. It just ticks me off when people ask "why not go with the best and most expensive option?" because that's when they make the assumption that other people have a lot of free cash to spend.

My suggestion was not a blindly made one, nor was I suggesting "the best and most expensive option". I suggested a step up from the AX-12 because you don't see many 60cm AX-12 based humanoids, and for good reason.

My Giger biped is approximately 60cm, and while I could likely have done it with MX-28s and a lighter frame, the AX-12 would be nowhere near powerful enough. I'm not trying to criticize, I'm just sharing my experience with the motors & bipeds of this size.

It's not even so much the torque being an issue, it's the inherit backlash of the servos. With hips that wide and legs that long, you'll run into issues effectively lifting your feet without the entire limb slumping due to backlash. I think Limor's design is great, but focus should also be on making it as compact as possible if he wants to stick with a low cost actuator, or implementing gearing. Again, there is a reason you don't see 60cm tall AX-12 bipeds.

[tempusmaster]

Thanks for the link. I think that what I bought is similar (braided, strength, less than .5mm etc).

So googling came up with the Tantillus RepRap which indeed uses these wires instead of expensive timing belts.
Thanks for this info! It is good to know that this is has been used reliably by others.
So what is your experience with clamping, slippage, knotting and tensioning of these wires?

The Tantillus is fairly new - I have one of the prototype units. That being said, there have been zero problems using the line. It's much better for this type of application than belts. Of course clamping/slippage/tensioning depends primarily on the mechanical design of the drive, not the line itself.

[limor]

Looks like someone else has been working on AX12/AX18 servo humanoid with pully-belt system

http://www.facebook.com/marco.bonechi

[limor]

in order to be able to 3D print the AX12 top "wig" cover, it is necessary to add some support material. so this tower structure gets built up in parallel to the wig, just so that at the very top it can provide these "barely touching" triangular supports so that the top protruding screw holes don't have to be printed into mid air

[limor]

After lots of experimentation, the pivoting torso is now supported not just by a lonely AX12 shaft but by circular trenches sandwiching some lubricated coing shaped nylon pieces. The torso pivot motion is limited to about +- 30 degrees this way.



I'm not very happy with these 3 aluminum parts holding all the torso servos in place, simply because it takes a whole sheet of aluminum to make this torso which is expensive and unless i make lots of holes everyewhere in the aluminum, it will add some non negligible weight to the robot's torso. i may at some point after the prototype is done, redo the torso so that it uses smaller aluminum parts bolted together.

[limor]

Regardless of my attempts to shorten the robot, it will end up being over 60cm.
I therefore added a servo to the knee and have 2 servos actuating each knee.
The feet and knees now need some proper work as do the body shell, hands and head.

[CdRsKuLL]

in order to be able to 3D print the AX12 top "wig" cover, it is necessary to add some support material. so this tower structure gets built up in parallel to the wig, just so that at the very top it can provide these "barely touching" triangular supports so that the top protruding screw holes don't have to be printed into mid air

Have you tried flipping it so the screw holes are printed first (as the image is, but flipped forward), then the top would be made via a bridge? I'm guessing it will be internal so it shouldnt effect the end result.

I found designing my you get it looking great then you realise O.. no you cant print that ..lol

[limor]

I specifically want it built so that those screw holes are layered horizontally. this gives very high strength to these small rings. if the rings are built vertically, they have no strength at all. I posted about this couple of weeks ago http://robosavvy.com/forum/viewtopic.php?t=8150

[CdRsKuLL]

Apologies, I only joined last night.

I've tried several different reels of filament PLA / ABS and just wondered what material you did the test with ? I've found if PLA snaps to easily I increase the head temprature afew degrees and then have more success.

Anyway, looks like a very interesting and involved project looking forward to seeing more.

[PaulL]

That guy w/ the toothed pulleys needs a tensioner for his belts. That has GOT to have some adjustment to work right!

After lots of experimentation, the pivoting torso is now supported not just by a lonely AX12 shaft but by circular trenches sandwiching some lubricated coing shaped nylon pieces. The torso pivot motion is limited to about +- 30 degrees this way.

...

I'm not very happy with these 3 aluminum parts holding all the torso servos in place, simply because it takes a whole sheet of aluminum to make this torso which is expensive and unless i make lots of holes everyewhere in the aluminum, it will add some non negligible weight to the robot's torso. i may at some point after the prototype is done, redo the torso so that it uses smaller aluminum parts bolted together.

I think you can get by with a smaller thrust washer type of approach. I don't think you need to have the load distributed so broadly. I have some thrust washers that are flat and wide with roller elements in them. The heaviest thing about them are the thrust "plates" (1mm thick steel), but they do work great. I've since been considering some teflon washers. My situation is a little different, as I don't have a servo in the middle of the joint (I'm planning on using a pulley to get motion to the rotational axis). Is it possible on a Robotis servo to use the hub center screw to tighten a thrust washer against the body of the servo, with the wider thrust load distributed against the body of the servo, with teflon / nylon sandwiched between the horn and the body? Just thinking out loud...

[limor]

RN1AsOf091407: I looked at wide bearings like a lazy suzy but i don't need that range of motion and it would add complexity and cost.

We are having some issues cutting nicely the designed bearings in HDPE plastic so it is preventing us from building the structure so far. So I'm working on the foot design.
I don't want to make yet another humanoid robot with a gigantic flat foot.
It needs to have suspension in the heel so that when the robot steps, the first impact is nicely absorbed and can be simulated more reliably. We bought a few RC suspension and matching oils to play with. I'm still unsure about how to link such a structure together using hobby car ball joint linkages..

[sap1975]

Hi Limor

Thank you for a very interesting build thread.
Might i suggest you have a look at a car wishbone suspension type setup.
I think its in-line with what you're looking for.

It's basically like the suspension you get on a F1 car but the spring and dampener is mounted diagonally between the wishbones.

Best
/Stig

[limor]

Hi Stig,
Thanks for the suggestion. I have been looking at this RC car suspension solution for a few days but the problem is that in the RC world the "language" that describes things is not very useful when you are trying to find parts of particular shape/size/dimensions. So the only way is to order a bunch of spare parts for particular RC cars and hope that one kind of fits and that that spare part will be available in 6 months time.

So I'm going with yet another 3D printed piece of art inspired by the wishbone / A-Arm setup.