My Advanced Realistic Humanoid Robots Project

[xrunner]

This would sound like a construction site operating in your bedroom while you are trying to sleep.  Unnacceptable. It must be as silent as possible. 

Yea well there's one big difference between their noisy prototype and your ... nothing sandwich. It actually works pretty damn well.

Again as you protest what we say to you, you fall farther and farther behind, because you know nothing about building on small and intermediate gains towards a final goal.

 

Right right I know what's coming as always -

"I disagree"

[Alex Eisenhut]

People regularly use blenders, food processors, and vacuum cleaners that make much more noise. Why? Because it beats doing it by hand.

Noise is not a problem.

Also, that Aliexpress link has a video, you can hear the winch running at the end. It's not exactly loud.

[artbyrobot]

Also, that Aliexpress link has a video, you can hear the winch running at the end. It's not exactly loud.

no that thing would be loud.  All metal gears are loud.  Particularly when dealing with high gear ratios which means faster motors and more gear chatter.  Some of the 3:1 downgearing on a expensive pancake bldc are more reasonable noise level but even that is too loud IMO.

People regularly use blender...

Noise is not a problem.

true, a single use for 5 seconds blender that is loud is fine since its used in a kitchen during hours where everyone in house is awake and only used for a short blast of noise and everyone is relieved when it is done which is not a long wait.  Now compare that to multiple humanoid robots all around the house doing chores 24/7 at the same loudness.  Completely different.  Completely unnacceptable.  Literal torture.

[xrunner]

I guess I'm in a feisty mood right now so I' going to run with it. 

I actually gave him the benefit of the doubt and I've looked at his videos going back, and anyone can see, ten years.

Ten years.

I was thinking "surely this guy has tried, just out of shear curiosity, to control a motor of some sort with a micro controller platform, such as yea Arduino or anything at all" Surely he's tried to move a test object with any sort of pulley arrangement he wanted to try. Gosh, in all this time he must have ...

No.

If I missed it I'd like to look at it, but I think he's stated that he's actually not done that. Does this person know anything at all about programming a microcontroller? I have seen no evidence of it anywhere. Of course he's a gift to engineering and can probably just look at a controller board and understand it all, much better than we can. But right now he's concerned about noise? He's going to build a humanoid robot that can fix itself? Does anyone out there believe this? Let's all get a grip on reality here, please.

Give me a f*cking break.

[artbyrobot]

Does this person know anything at all about programming a microcontroller?

I did some extensive motor control code in arduino IDE and there should be video of this coding project.  It is non blocking and speed adjusting.  But I won't be using that code on my robot necessarily although I can reuse parts of it.  My robot will not be using DC motors that this code project was designed to control.  It will use BLDC motors.

[Kim Christensen]

It is perfect.

Well, enough talk. Lets  see a video of it in action. Lifting and lowering a 10lb dumbbell to the floor/desktop, 15 times in succession, for it's max distance of travel, without failing.

[artbyrobot]

A bit gangster compared to me but much of the lyrics to this are relevant to our situation.

[Alex Eisenhut]

For someone who doesn't like noise... noise is all you do.

[RJSV]

   It's 'Gangsta' by the way.  Not to be correctly spelled,  (that makes it incorrect,  in application).
   Or maybe that shit is real,  (but people hate it when someone puts on the slang,  like a costume to impress).

[nali]

Ten years.

I see an oppertunity for an Internet challenge... OP should set up a race with the Project Binky boys to see who finishes first 

[artbyrobot]

I wound up my 6lb test Hercules PE braided fishing line onto the previous pulley system setup only to find out that the pulley could only handle about 21 inches of fishing line wound onto it before it started to come dangerously close to overfilling the pulley. The aim is to have plenty of the plastic disc overlapping the fishing line even when it is wound up fully to one side because that plastic disc acts as the guide to keep the line in its proper channel. I want at least 32:1 mechanical advantage out of this downgearing so if I want my final output to be 1" then the first pulley has to be able to wind 32" of fishing line onto it comfortably. So I realized at least the first pulley has to be a few more millimeters increased in diameter. So I had to rebuild the thumb tacks arrangement to accommodate these changes and make that first pulley bigger.





With this increased size first pulley, I realized I'm getting what looks to be 7:1 mechanical advantage from just the first pulley alone! At least initially when it starts. As the fully wound up pulley gets winched in by the motor, the relative size differential gets smaller which means it will speed up and the torque will be less than the starting torque and increasingly so as the size differential decreases. This will create a natural sort of acceleration effect and high initial power and gradually less power. I think these side effects of this system seem to be quite good but I'll know for sure in testing. The next steps will be to wind up the reverse direction of the first pulley and start connecting the first pulley to the second pulley and so on. I may not even need all 5 pulleys but we'll see. With the first pulley being already 7:1, if the remaining 4 are 2:1 say, then we'd have 7:1, 14:1, 28:1, 56:1, 112:1 so 112:1 would be the final output. That seems quite overkill and perhaps will be too slow. Although very strong. The motor outputs about 0.42 lb on average so .42*112= 47lb! Now the lever of the joint itself makes you lose mechanical advantage due to the fulcrum location etc so it would drop down to say 15lb but my finger individual joint flexion power is only like 5-7lb so that's double mine. So a bit overkill. So I might skip using one of the 5 pulleys. Having it there is nice though just in case we wanted to trade speed for power for some of them we'd then use that one as an optional strength boost we can tap into in the future if we want to trade speed for strength so I might just leave it in the design even if I don't use it just yet. In testing I may find I prefer to use it afterall. Nice to have that option if needed.

[Analog Kid]

That whole system looks like hell. You can't be serious.

[xrunner]

That whole system looks like hell. You can't be serious.

This is incredible. I would never post such an example of such slapdash workmanship if it came out of my lab. I don't think you would either. I would only do it to show a most basic idea but in an unworkable state for an example. Maybe he will say that.

But if it isn't - the parts are just ... awful looking. I made a simple side-by-side pic. From the side view the wheels - they don't even line up in the same plane. The sides of the pulleys should be rounded to avoid interference with any lines running against them. Given what I see, and this is my opinion from what is presented because I have nothing of it in my hands, this will not work reliably over any reasonable period of time.

But here we have it presented to the public for comment, so what are we to do but give our honest opinions? My opinion is directed at the parts and quality of work only. No one forced him to present this. Even if he comes back and proclaims it WILL work, then it's up to him to prove it by a full test setup run over some hours. Do you really expect this to work under load in a robot, and with how many others scattered around the frame.

 

[artbyrobot]

as to the aesthetics, technically it is a prototype for testing.  Iterations have to be fast.  It has already been rebuilt even before it hit testing phase.  It can't be very polished nor should it be.  To make it look polished would be folly if it isn't even proven to work yet.  Consider this famous prototype of a super soaker.  My prototype pulleys look better than it IMO.

[xrunner]

as to the aesthetics, technically it is a prototype for testing.  Iterations have to be fast.  It has already been rebuilt even before it hit testing phase.  It can't be very polished nor should it be.  To make it look polished would be folly if it isn't even proven to work yet.

Yes yes and that's what I alluded to. So when will the testing begin on the prototype? Haven't we been here before?

I mean by that motor, motor interface to controller, controller (such as Arduino board or whatever you choose), lines, test weight all mounted on a test bench for examination while in operation to ensure it actually does what you want, not using imagination to believe it will do what you want. Make some test routines that move the weight back and forth in different ways or random ways to stress test the prototype system.

[artbyrobot]

everything from motor controller to downgearing method to actuator type to guide tubing type to microcontroller type has been changed mind on over and over so that progress on bringing them all together to work as a single multipart testing system has been impossible to date.  If I really do stick to this pulley type, motor type, motor controller type, microcontroller type that I currently have selected, then work timeframes can be more easily estimated.  I dunno.  Maybe a few months?  That's if I stay dedicated to this like a full time job this fall and winter which may not be doable we'll see.

[joeqsmith]

If it were one of my three businesses, I wouldn't advertise such poor workmanship.  Nor that I couldn't calculate simple geometries and was guessing.    In your example of the super soaker,  it actually worked well enough to go to the next step.  I don't see that ever happening with your business. 

Iterations have to be fast. 

Why?  It's been ten years in the works with nothing to show beyond some CAD models.  I would like to know how much money this business has brought in during that 10 years.  Enough to pay for food, shelter, college?

[xrunner]

I took some time to find some small pulleys for sale to see what they looked like. I'm not spending all day on this but I did find this site -

https://www.flinnsci.com/products/physics--physical-science/simple-machines/pulleys/

Examine the pulleys they present. See how the edges are rounded a bit? See how they have angled sides which I think help the line to go in properly? That's the finer details of the design I think you are overlooking.

[Kim Christensen]

I took some time to find some small pulleys for sale to see what they looked like. I'm not spending all day on this but I did find this site -

That's what confused me at first. The OP is not really using pulleys, but rather pairs of winch drums of different diameters and using the cable from one to turn the other. So he'll have multiple wraps on each "drum" to gain mechanical advantage. Anyone who has used a cable winch knows that this is folly and will bind into a tangled mess.
Plus, the problem comes when you need to reverse direction. Unlike a transmission using gears, where you can simply reverse the motor, with this "genius" system you'll have to pull on the string from the load end to unwind the entire chain!

[Andy Chee]

Plus, the problem comes when you need to reverse direction. Unlike a transmission using gears, where you can simply reverse the motor, with this "genius" system you'll have to pull on the string from the load end to unwind the entire chain!

Well he is replicating the biomechanics of a human after all. 

In the case of human muscles, they can only apply tensile force, thus need to operate in pairs in order to get reversible movement e.g. bicep & tricep.

[Kim Christensen]

Probably quad drums with more strings wrapped in the other direction. Easy to lose the story in the walls of text and crappy diagrams:

The reduction is explained in this photo: 

each shaft of normal size will be winching a rope that later wraps onto a shaft of 2-2.5x the normal shaft size.  This means that for every turn of the prior shaft, the following shaft only turns half as much.  This is because the circumference of one shaft is the amount of rope winched and the circumference of the doubly large shaft is twice as much circumference so that second shaft will require the first shaft to turn twice as much for the second shaft to make a single turn.  Then the next shaft in line from the 2x shaft drops back down to a 1x size shaft and since the 2x shaft only turned half as much, this 3rd shaft translates that 1/2 turn ratio into itself and so on.  If you understand gears, it is exactly the same except instead of gear cogs you are just using rope to turn the next one in line.

[artbyrobot]

I was thinking it would be a shame to have to wait 3 months to test the pulleys when testing really is needed immediately so I can find any areas that need redesign before I commit too many hours into the pulley development.  I'd in fact like to test each pulley significantly before moving onto the next pulley to ensure that first pulley is working well.  So getting into the microcontroller and motor controller electronics and programming just to test these pulleys is a daunting series of rabbit holes in my case because of the things I want to do custom in those departments.  Which would put off the testing deep into the future.  So to avoid this, I realized I can test the setup by attaching pulleys at the top of the motor shaft, one wound and the other mostly unwound and manually pull one then the other to cause the motor output shaft to rotate forwards and backwards manually.  I can then observe the pulley downgear systems at any time without having to mess with the electronics at all for now.  So that's an exciting breakthrough and I can test early and often right away now.

[Bud]

Why?  It's been ten years in the works with nothing to show beyond some CAD models.  I would like to know how much money this business has brought in during that 10 years.  Enough to pay for food, shelter, college?

Youtube income, thanks to the idiots driving it by keep posting in this thread.

[artbyrobot]

As to unwinding from the end of the system, that's not the intention generally.  The motor is supposed to rotate clockwise and counter clockwise.  When rotating clockwise it will be simultaneously winching in the flexion group of pulleys while also releasing the extension group of pulleys.  When it reverses, it will winch in the extension group of pulleys and release the flexion group of pulleys.  The joint will pull out the slack on the opposing group being released while the group being winched pulls the joint.  So in some regard the strings complete a full circuit or loop from shaft to joint and back to shaft like a belt system.  But with many intermediary steps along the way.  A leaf spring with loop the fishing line is threaded through can act as a string tensioner to remove any slack from ever hitting the motor mounted pulleys so they never tangle.  But I'll only add that if it seems prudent.  It might not be needed but I'm not sure.

[artbyrobot]

shoutout to xrunner for what I would say was some great advice he gave relating to smoothing out the pulleys and removing any sharp edges not only in terms of a perfect circle, but also beveling the corners where the outer face meets the inner face so it is rounded in all directions always.  Nothing sharp at all anywhere.  That makes sense and I think is important.  I can add all of this once it is moving using a nail sander and just hold that sander as the pulleys rotate and it will smooth everything quickly.

The term for the type of edge in question that we want to make them is "bullnose edge" - a term often used to describe countertop edge finishes that are rounded on top and bottom edges.