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

[BrickBear]

Fair enough. I should be back up and running next weekend so hopefully I can continue contributing to the effort in the irl.

[Nelson]

A quick update. I tried several different "wheelbases", (the length between the front and rear legs), along with subtle leg angle changes, (pitching them slightly more forward or back), but nothing improved upon what I had in the most recent videos. Those changes invariably made the balance and movement worse.  It's all just so touchy. I'm going to try figuring out a way to get the head moving. I hinged it up in testing and manually moved it to counter the shifting weight and it seemed promising. It would be great if the head movement was in phase with the movement of the liftarms, so I could link them, but alas, they are not. There's no way to access the big gear that drives the crank either. Time to put the old thinking cap on. I'd like to use some kind of linkage rather than things like string or rubber bands, but those may be the best solution. 

[BrickBear]

On 1/15/2026 at 1:15 AM, Nelson said:

A quick update. I tried several different "wheelbases", (the length between the front and rear legs), along with subtle leg angle changes, (pitching them slightly more forward or back), but nothing improved upon what I had in the most recent videos. Those changes invariably made the balance and movement worse.  It's all just so touchy. I'm going to try figuring out a way to get the head moving. I hinged it up in testing and manually moved it to counter the shifting weight and it seemed promising. It would be great if the head movement was in phase with the movement of the liftarms, so I could link them, but alas, they are not. There's no way to access the big gear that drives the crank either. Time to put the old thinking cap on. I'd like to use some kind of linkage rather than things like string or rubber bands, but those may be the best solution. 

One option for the head might be to make it slide along a beam with the lifting of a leg, so essentially it moves by its own weight on a tilting beam.

[Nelson]

@BrickBear I may indeed do something along those lines. I have a few ideas. We'll see which one actually works and can fit in that tight space. Even if I reach Echo Base, I won't have room for any troops in there!

[Nelson]

This head articulation is a tough nut to crack. I've tried about a half dozen ideas this weekend. This is the most promising, but still needs a lot of refinement. Or perhaps some other epiphany will hit me.

 

 

 

Edited January 25 by Nelson

[BrickBear]

9 hours ago, Nelson said:

This head articulation is a tough nut to crack. I've tried about a half dozen ideas this weekend. This is the most promising, but still needs a lot of refinement. Or perhaps some other epiphany will hit me.

 

 

 

I’ve still not made any progress myself. Various art opportunities in the way. Maybe this weekend I can get back to things. In the meanwhile have you seen the official Lego motorized AT-AT head movement mechanism. It’s simple yet effective (albeit timed for the inaccurate walk cycle. Maybe check it out and tweak the timings so it works for the actual AT-AT gait? 

[Nelson]

@BrickBear Yes. I built that kit with my girlfriend's son back when it came out. I don't have a way to feed off the drivetrain to make that kind of mechanism work. At least I haven't found a way yet. 

Edited January 26 by Nelson

[BrickBear]

1 hour ago, Nelson said:

@BrickBear Yes. I built that kit with my girlfriend's son back when it came out. I don't have a way to feed off the drivetrain to make that kind of mechanism work.

I see, well I think the one you’ve got is very good and worth me stealing if I may. 

[Nelson]

@BrickBear  Steal away!

Now you've got me thinking of ways to connect to the drivetrain. That really would provide ultimate control over the timing. Another fine evening of ripping the whole thing apart!

[Nelson]

It took several attempts, but I finally figured out a way to run a gear for the head articulation off the main drivetrain. There is so little space to work with up there, it is a miracle I found a way to rig it up. This will provide complete control over the timing of the head movement, which is awesome. I'm working on ways to make the mechanism smoother. There is also the question of aesthetics. The place where the neck joins the body might be a bit uglier than I'd like, but of course, I'll keep refining it. The motion in the video is a bit wonky as I haven't balanced things out weight-wise yet.

 

Edited February 8 by Nelson

[BrickBear]

2 minutes ago, Nelson said:

It took several attempts, but I finally figured out a way to run a gear for the head articulation off the main drivetrain. There is so little space to work with up there, it is a miracle I found a way to rig it up. This will provide complete control over the timing of the head movement, which is awesome. I'm working on ways to refine this mechanism to be smoother. There is also the question of aesthetics. The place where the neck joins the body might be a bit uglier than I'd like, but of course, I'll keep refining it. The motion in the video is a bit wonky as I haven't balanced things out weight-wise yet.

 

Noice noice noice. Very impressed. It’s neat how the head starts moving back towards the other side right before the other leg lifts.

Been working on the hip mechanism myself. As always trying to get that specific movement in. 
 

Looking forward to see progress!

[Nelson]

Okay. It may be time for a break. (Yeah, right.)

Another update on the Abridged-Terrain-Awkward-Transport. It's getting heavy on the awkward ("heavy" also being an operative word.) I'm struggling to get this MOC to the finish line. Much of this project has involved overcoming the limits of the materials, and this phase is no different. To get the aesthetics I desire, the weight just keeps piling on, and the plastic bends, flexes, and occasionally gives up the ghost. The "pinning" technique I worked out for the legs did a LOT to sturdy those up, but it's not enough to wrangle so much weight. I'll try to get an ounce or two off the body panels and chassis, but I feel like something needs to happen with the legs themselves to get things working better. It may mean giving up on the faux leg bend and adding another solid liftarm inside the leg structure. That would suck, but it may be the best option considering I'm also struggling to figure out a way to articulate those upper leg portions. On the plus side, I'm really happy with the way it's looking. I just received a parts order and altered the upper legs to get the width proportions more accurate. It's tricky to pull off widths like 5 studs when dealing with curving areas. Many of the panels are not yet completed in this video. Anything sitting on top of the head and body is literally sitting due to gravity alone. I may take a little breather and do some diffuse thinking rather than focused thinking on it. Unless I have an epiphany of course!

Update: I made some adjustments to the ankles to limit their rotation a bit more and that helped somewhat. I think the rods that drive the mechanics need to be bulked up or doubled or something. They flex too much when struggling to move so much weight. I'm sure pondering a solution will invade my sleep for days to come.,

Before I put the MOC on the shelf, I did a quick comparison to some online photos of a film model to see how my proportions and details were stacking up. If you ignore the missing head and rear sections, I think it's pretty close for a motorized model of this scale. (The head is a bit droopy because it's barely held together at the moment.) There are definitely a few tweaks that could be made.

 

 

Edited February 25 by Nelson

[BrickBear]

On 2/22/2026 at 8:08 PM, Nelson said:

Okay. It may be time for a break. (Yeah, right.)

Another update on the Abridged-Terrain-Awkward-Transport. It's getting heavy on the awkward ("heavy" also being an operative word.) I'm struggling to get this MOC to the finish line. Much of this project has involved overcoming the limits of the materials, and this phase is no different. To get the aesthetics I desire, the weight just keeps piling on, and the plastic bends, flexes, and occasionally gives up the ghost. The "pinning" technique I worked out for the legs did a LOT to sturdy those up, but it's not enough to wrangle so much weight. I'll try to get an ounce or two off the body panels and chassis, but I feel like something needs to happen with the legs themselves to get things working better. It may mean giving up on the faux leg bend and adding another solid liftarm inside the leg structure. That would suck, but it may be the best option considering I'm also struggling to figure out a way to articulate those upper leg portions. On the plus side, I'm really happy with the way it's looking. I just received a parts order and altered the upper legs to get the width proportions more accurate. It's tricky to pull off widths like 5 studs when dealing with curving areas. Many of the panels are not yet completed in this video. Anything sitting on top of the head and body is literally sitting due to gravity alone. I may take a little breather and do some diffuse thinking rather than focused thinking on it. Unless I have an epiphany of course!

Update: I made some adjustments to the ankles to limit their rotation a bit more and that helped somewhat. I think the rods that drive the mechanics need to be bulked up or doubled or something. They flex too much when struggling to move so much weight. I'm sure pondering a solution will invade my sleep for days to come.,

Before I put the MOC on the shelf, I did a quick comparison to some online photos of a film model to see how my proportions and details were stacking up. If you ignore the missing head and rear sections, I think it's pretty close for a motorized model of this scale. (The head is a bit droopy because it's barely held together at the moment.) There are definitely a few tweaks that could be made.

 

 

Very nice. I understand what you mean about the weight. I guess it’s why people tend to say it’s an impractical concept. Like obviously when it comes to things like the elephant or paracetherium we know as animals they would naturally be able to distribute the weight via micro adjustments in their body. With a mechanical vehicle it’s a bit different and therefore frustrating and fiddly.

That all said, I’ve done some work recently on the hips. You can probably hazard a guess as to how they work from the image. Basically the orange liftarm rotates pushing the L liftarm and the white curved liftarm upon contact. Creating a forth and back movement with a short stride inwards than outwards as in the real thing.

The universal joints allow for the hips to move up and down while powering the orange cams. 

[BrickBear]

More progress. Same gearbox connector to front and back as in other prototypes.

Edited February 28 by BrickBear

[Nelson]

@BrickBear

I'm going to try and replicate your image to see what you've got going on. While it may not be a design conducive to my concept, I've found inspiration from your ideas in the past. It looks like you may have had some bushings on the ends of the orange pieces in the first photo, but they're just tipped with raw pins in the second photo. Any reason? Or am I not seeing it right?

[BrickBear]

32 minutes ago, Nelson said:

@BrickBear

I'm going to try and replicate your image to see what you've got going on. While it may not be a design conducive to my concept, I've found inspiration from your ideas in the past. It looks like you may have had some bushings on the ends of the orange pieces in the first photo, but they're just tipped with raw pins in the second photo. Any reason? Or am I not seeing it right?

I removed the bushings as it was a bit tight. When I tested the whole thing I found it quite a friction-y mechanism.  Might be due to the closeness of the 16 tooth gears against the 8’s. I bet you’d figure a better way.

I decided the next step is to use the concept of the cams to push and pull the legs on my previous design which might prove less, as I say…friction-y

[BrickBear]

 

Weekend achievements.

Outcome cycle:

 

The gaps between the red dots represent speed, yellow indicates slowest, the more separation between the dots, the faster it was moving, thanks to the spring I’ve achieved the different speeds the leg achieves in lift and forwards step vs ground contact. I just need to adjust the cycle shape.