Help with finding solutions for getting an AT-AT walk cycle like in the movies.

[BrickBear]

46 minutes ago, Nelson said:

I made some adjustments to the cam contact areas and got more lift. This helps the inside "toe" clear the ankle stabilizer. I may make the toes hinged so they clear better and look more accurate. (I had that feature on an early model.) I'll just have to see if it compromises stability. The cadence is still a bit off. I'm pretty sure this is a result of the mechanism's geometry being imperfect and the limited range of the ankle joint. The rear foot can't flatten out when the leg comes down because the ankle only has a flex of a few degrees. This is a fine line to get right. I could try to get a bit more flexibility range, but then risk the walker falling back on its rear. (I believe @BrickBear was experiencing this phenomenon.)

I think the next step is to try and add some body panels and a rudimentary head so I can see how the design handles the weight and balance changes. 

There's also the issue of the legs which begin to split apart after a while. I'm bummed about that. I spent a massive amount of time getting these legs designed (with strength in mind) and I feel like the options of build styles I have for getting better longevity will compromise the ascetic. But hey. That's part of the challenge.

 

 

I reckon you could fix the cadence problem (i.e only one foot off the ground) not by tweaking the gearing and mechanism but by adding a spring or two to each foot. 
 

With the spring pushing the previous foot forward a little it should in theory touch the ground as the next foot lifts, giving the illusion all is in sync.

As for my AT-AT’s balance it’s more the issue is that is has the potential to fall forward left or forward right if the weight isn’t on the feet on the ground but i’m hopefully going to remedy this with the seesaw mechanism. 

[Nelson]

@BrickBear I'm not following. Are you saying the whole foot would be attached via springs? Would the overall length of the leg-foot assembly vary as the springs compressed and relaxed? I don't see how that would work. It seems like it would introduce a slew of other issues.

[BrickBear]

1 hour ago, Nelson said:

@BrickBear I'm not following. Are you saying the whole foot would be attached via springs? Would the overall length of the leg-foot assembly vary as the springs compressed and relaxed? I don't see how that would work. It seems like it would introduce a slew of other issues.

Yes, basically the (soft) spring(s) would be half a pin hole longer than the current pivot point for the foot when fully extended. This would give the illusion that the foot has touched the ground when the next foot lifts.

As the leg moves the spring could then contract to the position that you currently have the feet in (adding a stopper to prevent it contracting more) and that may well solve the problem of the foot not quite touching the ground upon its step forward, being the soft kind of spring as well it should not place too much pressure on the ground if any to affect the balance as the weight of the walker would compensate.  
 

 

[Nelson]

17 minutes ago, BrickBear said:

Yes, basically the (soft) spring(s) would be half a pin hole longer than the current pivot point for the foot when fully extended. This would give the illusion that the foot has touched the ground when the next foot lifts.

As the leg moves the spring could then contract to the position that you currently have the feet in (adding a stopper to prevent it contracting more) and that may well solve the problem of the foot not quite touching the ground upon its step forward, being the soft kind of spring as well it should not place too much pressure on the ground if any to affect the balance as the weight of the walker would compensate.  
 

 

Interesting idea, but I think I'll explore other options for now. There is only about 2.5 studs worth of lift happening between the foot and the ground. I think a system like that would negate a good percentage of that clearance. It would also be tough to keep the ankle struts' functionality working. Aside from looking accurate, those bars are critical for stability.

[Nelson]

@BrickBear I had an enlightening result from my attempt to add a bit more ankle flex. The whole design almost completely failed when I added even the slightest bit more range. Previously, the foot could only flex about 2-3 degrees at the ankle. I bumped that up to about 5 degrees and suddenly the model pitches heavily toward the lifting leg, nearly falling over. It sounds a lot like what you're experiencing in your design. Maybe try an experiment where you make the ankles of your model quite rigid, just a couple degrees of range. I'm curious to see if that negates some of that leaning you mentioned. I was not expecting that much difference...or any difference really. It was quite surprising and counterintuitive. I would not have suspected that a joint perpendicular to lateral stability would have such an effect. I guess that stiffness was a key component to my current model's success. Looks like I'll be putting it back the way it was. I'm trying to decide if I'm satisfied with the way this version walks. It's certainly not perfect, but it's closer than Lego's version. I'm going to do a few more tests this weekend, then put it on the shelf for a while. There are some other projects calling and I need a break on this. I'll keep checking in on this thread to see how things are progressing.

 

Edited January 18 by Nelson

[BrickBear]

5 hours ago, Nelson said:

@BrickBear I had an enlightening result from my attempt to add a bit more ankle flex. The whole design almost completely failed when I added even the slightest bit more range. Previously, the foot could only flex about 2-3 degrees at the ankle. I bumped that up to about 5 degrees and suddenly the model pitches heavily toward the lifting leg, nearly falling over. It sounds a lot like what you're experiencing in your design. Maybe try an experiment where you make the ankles of your model quite rigid, just a couple degrees of range. I'm curious to see if that negates some of that leaning you mentioned. I was not expecting that much difference...or any difference really. It was quite surprising and counterintuitive. I would not have suspected that a joint perpendicular to lateral stability would have such an effect. I guess that stiffness was a key component to my current model's success. Looks like I'll be putting it back the way it was. I'm trying to decide if I'm satisfied with the way this version walks. It's certainly not perfect, but it's closer than Lego's version. I'm going to do a few more tests this weekend, then put it on the shelf for a while. There are some other projects calling and I need a break on this. I'll keep checking in on this thread to see how things are progressing.

 

That is useful to know. Alas I cannot test your theory this weekend due to being ill but I shall stare longingly at my model from my bed.

[Nelson]

@BrickBear Any progress? I haven't had a chance to do much with my design lately. I did try a few other settings with the flex range of the ankle, but ended up back where I was for the test in the last video. I'm planning on eventually ordering a second pair of Powered-Up large motors and reviving my older design but with a more direct drivetrain. While this version is the best working I have built to date, I'm not 100% happy with the slightly off cadence of the steps due to the geometry of the mechanism. Hope you're feeling better and can get back to destroying Echo base!

Edited February 2 by Nelson

[BrickBear]

21 hours ago, Nelson said:

@BrickBear Any progress? I haven't had a chance to do much with my design lately.

Yes, i’ve redesigned the feet so that they have grips underneath them now. 
 

 

This prevents feet slippage.

And i’ve reduced leg wobble by placement of two plates topped by two  cheese wedges facing each other under the part of the liftarm that joins to the foot.

 

[Nelson]

@BrickBear That's a neat idea. I wonder what adding something like that to the feet of my model would do. I have a bunch of those rubber parts. Did you call them "cheese wedges"?

[BrickBear]

55 minutes ago, Nelson said:

@BrickBear That's a neat idea. I wonder what adding something like that to the feet of my model would do. I have a bunch of those rubber parts. Did you call them "cheese wedges"?

No cheese wedges are the little plastic 1x1 triangles i’ve placed both of them angled face opposite one another on two plates on the 4x4 round piece that goes on top of the AT-AT foot underneath the liftarm that connects the “ankle” joint to the lower leg. 

I just call the rubber pieces “rubber pieces”

 

[Nelson]

On 2/3/2025 at 4:16 PM, BrickBear said:

No cheese wedges are the little plastic 1x1 triangles i’ve placed both of them angled face opposite one another on two plates on the 4x4 round piece that goes on top of the AT-AT foot underneath the liftarm that connects the “ankle” joint to the lower leg. 

I just call the rubber pieces “rubber pieces”

 

[BrickBear]

Progress update, I modified the feet further, now the feet lie flat and the rubber pieces extend out to the side, still providing friction, but with much more stability.

 

secondly I have come up with a head mechanism, but it needs work. I intend to have the head act as a counterbalance and it seems to work well if it’s positioned just right however I think I want to shift the body weight left and right when walking as in a real animal too just so balance is not solely reliant on a complex head mechanism.

I estimate that once fully panelled it will be around 40cm tall.

 

note the second picture looks funny because I very badly clipped put the background and the head seems much larger than the feet cos of the angle. The right rear leg is raised off the ground by about 1cm. The head is currently acting as the counterbalance, you can probably see the mechanism by the neck, it’s a sort of seesaw that gets pushed up by the shoulders which turns a gear and that gear connects to a cam which moves the neck.

I did power the legs a little and it showed some promise of being able to move. Lets hope for further progress.

[BrickBear]

So I decided to investigate some other solutions to the wobble and i’ve come up with actually quite a neat system. Basically I move the linkage that was originally connected to the cam to it’s own separate gear. That means I have one worm gear moving two gears, the first gear makes the hips move up and down, the second moves the leg forward and back. This makes the leg have a connection above the hip joint as opposed to below it and has significantly mended leg wobble. 

The caveat however is a new knee bending mechanism needed to be introduced. I tried a few things and managed to come up with a system whereby knee bend is initiated by a spring and when the foot is on the floor, the kneebend is counteracted by the weight of the walker pressing down on the foot. This means I need to order 4 more softer springs or figure a way to modify it so the spare rubber pieces can be used.

The yellow axle would be connected to the ankle joint.

[Nelson]

@BrickBear Glad to see you're keeping at it and have made some progress. I'd love to see your model fired up, even if it just shows what challenges still need to be addressed. I've been trying to parlay the success I had with the more direct motor position of the slide piston design into a new chassis based on my earlier mechanisms. The gait was more correct with that old design. It just struggled with flex and lag in the drivetrain. Unfortunately, creating a combination of the old mechanics with the new motor position is proving very difficult. I just received a new pair of large Powered Up motors so I can keep the slide piston design intact since it's actually working pretty well. I put a short clip of it walking on a Lego Star Wars Facebook forum and have had a ton of positive response. Perhaps I'm being too picky about way that model walks and should keep fleshing that design out.

[BrickBear]

17 minutes ago, Nelson said:

@BrickBear Glad to see you're keeping at it and have made some progress. I'd love to see your model fired up, even if it just shows what challenges still need to be addressed. I've been trying to parlay the success I had with the more direct motor position of the slide piston design into a new chassis based on my earlier mechanisms. The gait was more correct with that old design. It just struggled with flex and lag in the drivetrain. Unfortunately, creating a combination of the old mechanics with the new motor position is proving very difficult. I just received a new pair of large Powered Up motors so I can keep the slide piston design intact since it's actually working pretty well. I put a short clip of it walking on a Lego Star Wars Facebook forum and have had a ton of positive response. Perhaps I'm being too picky about way that model walks and should keep fleshing that design out.

I haven’t made much progress in the last weeks. Just fiddling when in passing. Must get back to it next weekend and that Facebook reaction certainly sounds good and suggests the current iteration maybe is worth pursuing.

[BrickBear]

Progress update. I think i’ve found the final design for the spring bent leg, i’m going to attach all four and see if I can get it walking. (Probably not today though)

 

edit: on further inspection the chances of this being successful may be slimmer than anticipated 

Edited March 16 by BrickBear

[BrickBear]

Update on this, I’ve been working on the spring compression method to make it easier for the leg to straighten. 
 

I’ve made a diagram of what should happen and what shouldn’t happen.

i think i need to take a look at my first design, i’ve been developing something that seems good but i think my first one where the spring was compressed by the pressure on the foot might be the smartest idea. 

[Nelson]

On 3/22/2025 at 11:14 AM, BrickBear said:

Update on this, I’ve been working on the spring compression method to make it easier for the leg to straighten. 
 

I’ve made a diagram of what should happen and what shouldn’t happen.

i think i need to take a look at my first design, i’ve been developing something that seems good but i think my first one where the spring was compressed by the pressure on the foot might be the smartest idea. 

[BrickBear]

Somethings going down at Kurt Drive Yards:

 

[BrickBear]

Promising but spring position might need adjusting so theres just a little more leverage force pressing down on it as the legs don’t quite straighten:

[BrickBear]

Altered leg for more squish (4 wide), offset by half a beam hole to ensure symmetry.

Edited April 19 by BrickBear

[BrickBear]

Happy May the 4th!

So i’ve removed the bend from the legs for now. But i’ve been working on a mechanism to shift the weight of the walker as you can see the body is at an angle:

It works quite well, although it still struggles to walk (the front toes seem to catch so i might need to tweak them) however it does not fall over which is an improvement. In conclusion we’ve gone from nothing to vague shuffling.

Edited May 4 by BrickBear

[Nelson]

Glad to see you're still at it. You're going to have to post some video sooner or later, even if it's to see where the movement is falling short. I'm dying to see the shuffle! You hung onto the bending knee dream a lot longer than I did. I hope you're able to get that working once you solve for the other issues. I just received a part order and will hopefully get a new chassis together this weekend. It should be a bit faster than the last, have better clearance from the piston rods that were grazing the upper part of the legs, and possibly offer some better timing so the rear and front legs don't have overlapping cycles. Or it could just fall over. Who knows? I've also been working on my head design, knowing that I should get that integrated sooner than later. I'd hate to have a working version that is set back to zero because it can't accommodate the head weight. I also wonder if I could get away with only using one motor.

Edited May 9 by Nelson

[BrickBear]

15 hours ago, Nelson said:

Glad to see you're still at it. You're going to have to post some video sooner or later, even if it's to see where the movement is falling short. I'm dying to see the shuffle! You hung onto the bending knee dream a lot longer than I did. I hope you're able to get that working once you solve for the other issues. I just received a part order and will hopefully get a new chassis together this weekend. It should be a bit faster than the last, have better clearance from the piston rods that were grazing the upper part of the legs, and possibly offer some better timing so the rear and front legs don't have overlapping cycles. Or it could just fall over. Who knows? I've also been working on my head design, knowing that I should get that integrated sooner than later. I'd hate to have a working version that is set back to zero because it can't accommodate the head weight. I also wonder if I could get away with only using one motor.

I’ll get a video and I’ll definitely be attempting leg bending again, I actually found a solution to the overlapping cycles, basically if you have a worm gear powering your legs you can give it half a bush of space on the axle to slide and that provides a delay as the worm gear will slide before turning the legs.

I’ve also reduced the angle that the body shifts because it was a bit too much with the battery box, the big problem is the legs bending towards the raised leg and the feet catching, so i need to add the head as a counterweight and shore up the leg joints. Progress! 

 

[Divitis]

55 minutes ago, BrickBear said:

basically if you have a worm gear powering your legs you can give it half a bush of space on the axle to slide and that provides a delay as the worm gear will slide before turning the legs.

Brilliant solution to create an offset. I cannot wait to have the occasion to steal it!