Foot Mechanism Revisited

[Brian..]

WARNING: If you enjoyed John's high tech builds, then this isn't for you. This is seriously LOW TECH!

 

I haven't been building for some months now. My son and daughter both have new homes and I'm the free DIY guy. My wife also wants to make improvements, so I've not had the time to do much. Slowly I'm beginning to get through the mountain of jobs and can think about something new.

 

I've always been frustrated by the complexity of feet. I find it difficult to get the wheels turning in one direction. So I use four ratchet wheels with eight tyres and four pawls (the pawl is the tiny bit that stops the wheel turning backward.) The feet end up looking like Swiss cheese with holes for the multitude of pins that I need. It works fine, but the moment I look at a new design I have to, literally, reinvent the wheel. The B movie and pulp robots that I like to build have very small feet. There's bugger all room inside them to fit the two pawls that I need. The tyres rub, the pawls jam and it takes ages to get the design fine-tuned.

 

Then there's the problem of assembling them. I can do the job blindfold now, but it's taken years and it's bloody tedious.

 

So when John dismantled the new Robby and showed the foot design I decided to try to simplify the whole business. The new Robby wheels have no traditional ratchets. Each wheel sits in a recess that is slightly oversize, so that it can slide backward and forward an eighth of an inch or so. That's just enough to jam the wheel against the foot when the leg moves backward, and allow it to turn freely when the wheel moves forward. 

 

Well, I failed. It works 99% of the time but once every now and then the wheels would misalign and the action would go haywire. Just not reliable enough.

 

But I did find an alternative. 

 

The waste plastic that you have to remove from certain printed parts has interesting properties. It's very thin (1/64 inch), very flexible, very robust and it's free. I'd toyed with the idea of using it before, but I'd invested months in the other design and it was working. Now in desperation I tried using it on the new wheel project and it worked a treat. All I have to do is lay a short piece of this flexible plastic on the base of the foot so that it's touching the ratchet. Position it correctly and it works like a charm. The front wheel is easy, because there's plenty of room to glue down the strip. The rear wheel is slightly more awkward but by using a natural "kink" in the flexible strip it hooks over the rear ratchet.  I'm sure that I could do with just one ratchet on each foot but it's really no trouble fitting two, and it provides belt and braces.

 

So now from each foot I have eliminated eight holes, two pawls, and two pins. The design is easily adapted for any future build and I won't have to spend hours reworking the foot to give enough clearance. The foot also can have a solid base that can be clipped into place. The strips take just a minute to fit.

 

There's one final advantage. The feet act like little sounding boards and the plastic strips make a lovely clacking noise. 

 

 

 

BEFORE:

 

 

 

AFTER:

 

 

 

 

 

 

 

 

 

[ROBODUDE]

GENIUS !!!!!!!

[K-9]

Very clever solution and the sound effect is a nice plus...

[Golddalek]

Interesting...I commend you for your research.

I'm also looking at such a foot rachet mechanism...But I am copying the mexican robot

files from Robothut 'Thingaverse' kit and adapting it for my robot design.

 

still having parts printed and tested.

[robothut]

Brian I have a new project that I will call Walking Trash Master Robot. It uses the sliding foot wheel system of the new Robby, and that system is used in countless other vintage toys as well. But it is simple and prints and works every time. The Teeth on the wheels do not need to be saw tooth "pointed" in the new design so one ratchet wheel will work on any foot and in any direction, meaning no need to look when putting it together. Waiting on some hard ware bits to show up before I shoot the video. But also there is a 3d printed one way bearing that a you tube guy I like to follow has made that could make for a nice foot wheel in these Slide walking type robots if it can be made small enough. Interesting any way. I used to use 1 way bearing in the high speed duplicating equipment back in the 1990's and metal one come in very small sizes, but they are not cheep.

 

[Brian..]

John, I look forward to the walking Trash Master. I'll be "borrowing" left right and centre.  I notice that I've been using crudely cut 2mm nails as axles, while that new Robby seems to use smooth, rounded lengths. Perhaps I need to tidy up my cutting. 

 

I recall something similar to that second design on Angus' site - is that the one you mean? There looks to be a lot of friction on that example in the video, and the tolerances for those printed inner pieces look unforgiving, particularly if you need a small wheel.

 

To be fair, my low tech solution has given excellent results on the test robot I ran. Easily the most positive action I've managed - there was zero backward slip.  You also avoid that problem where gravity leaves the pawls upside down! You do need the directional ratchet teeth, but since I've already got them then it's out of the way. 

 

Removing or improving that inner leg strut is high on my list. Later Horikawas use interesting solutions. 

 

I think I'm going the way of Hong Kong robots - increasing simplification of the basic design. Hopefully I'll stop before I reach that piss-poor action of the later robots. 

[robothut]

Yes, I was not suggesting you use the 1 way bearing , just thought it was interesting that there was a small 3d printed version. Smaller than angus's version.

Ok here are some pictures of what I am using on the Trash Master. The ratchet wheels as you can see is very small and yet very robust, no small thin or pointy parts to wear or break and no wrong way to install it. The parts print with no supports standing on end.

For an axle I use 2mm cut rod and sometimes that means finding a finishing  nail and cutting it down. I just file the ends so they are not ruff. So there are only 2 parts to the foot. The main body of the foot "top part" and the bottom cap. Never mind the shape and cut out section in the center of my foot as they are part of what I am up too and have nothing to do with how the simple and robust ratchet wheel , foot design works. The wells and axles sit in cut outs in the foot bottom cap. This bottom cap can be glued or screwed to the main foot body. In my case I am screwing them on. The part that the ratchet wheels lock on to is part of the top main foot and it prints flat with no supports . The cool part is that the part that the wheels lock on to is solid and robust . Any way its fool proof and works great and will not break like thin part can do over time.

[Brian..]

Thanks John, very clear. I'll adjust my first attempt based on your design. I think I can see where I went wrong.

 

I'll have a two robot run-off to see which of the feet work best.

 

The two-tyre wheel gives a lovely stable robot, and two ratchets guarantee good grip. The Japanese got away with a broad tyre and a single ratchet, which made for economy but they don't always walk reliably. 

[Brian..]

The no-ratchet wheel finally works. John's detailed photos were invaluable. I cut the pins more accurately and cleaned up the ends. I also used those non directional wheels.

 

On the plus side, it's easy to install.  It's also going to be a breeze to adapt it to suit any future build. The Devil Girl Robot has very strange horse-like feet and it was always a problem squeezing pawls into the confined space.  On the down side, it's silent, so I lose the delicious clickclickclic.  On balance I like the new wheel! I'm retro fitting it on a Dynamo Robot to give it a good airing. 

 

I've had a look through my robots and can't see it in use on vintage models or on recent reproductions. It's such a labour saving feature that I wonder when it was first used. The new Robby was the first time I'd noticed it - following John's autopsy. I'm not up on recent robots but it would be interesting to find out more. Those early robot designers really struggled with ratchet wheels. 

 

If you're designing your own robots it's the way to go. I cut my teeth on the CAD program by struggling with those blasted ratchet/pawl designs so it taught me a lot, but it's also potentially demoralising for a newbie. 

 

I think I may add a clicker to the wheels just for old time's sake. 

 

 

[robothut]

Nando lil telocomando from the 1950's uses this type slide ratchet wheel design. Also Mr. Robot JR. uses it. The list goes on.  I also found that it is easy to design a version with a single rubber O ring in the center and the non directional teeth on each  side of the rubber O ring.

[Brian..]

Much older than I thought. Which makes me wonder why it wasn't used as standard. Later Horikawas and clones took every opportunity to shave cents off the build price. 

 

One down side appears to be an extra lag in the operation of the foot. It's not as crisp as the usual fine tooth ratchets, so slow-walking robots move even slower! I think I'll finally have to abandon my white motors and move over to the 1:120 ratio ones. 

 

 

 

[robothut]

Yes I covered the sliding ratchet wheel down side when I did the thread and video on the Nando robot repair some years ago. The leg/foot movement is so small that almost all the movement is used up before the ratchet wheels can lock, so the steps are very small. But with robots that have plenty of leg / foot movement "no matter what the speed" the design works great. It has been used on many robots. I guess I could go out and start talking pictures of all the ones that come to mind. Or maybe a new thread on the robots that used the sliding wheel ratchet design, if any one is interested. Sounds like work and it is very cold in the hut. I need motivation.

[robothut]

The Nando video was 6 years ago. Wow, how time flies.