MINDSTORMS BIG ROBOT V2

INTRODUCTION
Recently I posted a topic about the End of a Reviewing Era. More details can be found in the corresponding topic. In this topic I mentioned that I was planning on picking up building Mindstorms. A while a go I started an Omnidirectional Robot project, using EV3. I have been working on this project on and off for several years. Of course, more off than on hehe.

Some issues I had with this project:

• Difficulties daisy chaining with the original software
• Python on EV3 boots very slowly (a couple of minutes)
• Replaces EV3 hubs with BrickPi's
• BrickPi solution wasn't easy to integrate into the build
• Bulky design
• Lots of non-LEGO parts
• Etc

When Spike Prime was released I wasn't very enthusiastic. You can read more in my review of Spike Prime. After the release of the Robot Inventor (see my review here) set I got really interested in the new hardware. The new software works like a charm and has the possibility of broadcasting messages to other bricks (this basically replaces my needs for daisy chaining multiple hubs). I have been in contact with Laurens Valk quite frequently this year, so I got really excited about Pybricks as well.

I decided to start working on yet another big robot. Making a big robot, possibly humanoid-ish, child sized, is kinda my dream project. My Magnus Opus hehe. Since I hate the color scheme of Spike Prime, I decided I will use the Robot Inventor color scheme, black and white with Teal (Dark Turqoise) accents. Currently I have four large Technic Hubs and two of them are in Teal. I can use the other two as well, but when push comes to shove, I will replace the yellow ones with Teal ones, unless they will be mounted somewhere internally. I have no clue. what the robot will look like or what kinda functions it will have, but I will start building the base. This will be challenging enough. When I am happy with the base, I will contine with the upper structure. First things first.

Some criteria I have in mind:

• Color scheme is black, white and teal
• 100% LEGO
• Software written in blocks or Python (using the official software or Pybricks)
• Using the big BMW motor tires (I already ordered 4 on Bricklink)
• ...

I am enjoying Stud.io (or Studio) more and more these days. I was hardcore LDD user, so I was disappointed that I can't use it anymore, but I am starting to like some of the features of Stud.io. Whenever possible, I will share some concepts drawn in Stud.io.

DESIGNING THE CONCEPT
Since my physical bricks are still stored due to our renovation, I had to start digitally using Stud.io. I came up with this as my first concept (the final version will obviously have four wheels). I will be using two motors per wheel for propulsion and one for rotating the wheels. This means three motors per unit, which boils down to using a single big hub for one side of the robot. Having two teal hubs, this will cover driving the robots (two hubs, 12 motors).

Unfortunately, there aren't any extension cables, so I have to be careful with the placement of the motors. I will place the hub in the upper structure, between both wheel units. The left version will most likely result in not having enough cable to reach the hub. So I have created the second version, which will probably suffice.

@Mr Jos was kind enough to send me instructions for his 6 Degrees of Freedom crane. The image below is how he solved the gears. I will probably use a slightly different approach, but I really do like looking at his solution for inspiration. I also downloaded the Bucket Wheel Excavator instructions, but I am not sure whether they will help me a lot.

I am glad that we merged Mindstorms with Technic, so I can post it here without scrupules. As usual I would love to hear your opinion. Do's or don'ts. Hints or tips. I am not an experienced MOC builders, so any help is appreciated. 

I am aware that the bracing is insufficient, but I am trying come up with an initial design and iterate from there. And obviously, this will be easier when I can build it physically.

I already know this will be a topic I will be following closely, and await updates from!

You could use the inner 60T turntable to steer the wheel, and the banana gear racks as a bearing only, to allow smooth turning. When both are connected I think it should work pretty good.

You just will need to put very good cross bracing everywhere to support the high weight + torques for moving.

Based on your initial design, I imagine this will really be huge. I sure hope it will be stable enough to stand and “walk”.

I too have a big robot moc in mind that I’ll be working on once I get more free time. I think I now have enough parts and electronics to make it work. And now having hub-to-hub officially supported is great!

I will take what I learned from upgrading Blast to a biped and see how it works with a bigger version.

Will watch out for your updates and maybe can learn from each other. 😉

21 hours ago, Jim said:

(the final version will obviously have four wheels)

Just asking, but why is that obvious? I mean, it's a sensible design choice, especialy in the world of rectangular Lego bricks, but one advantage of having 3 wheels is that you're sure your model is always stable and all wheels always touch the ground. With 4 wheels, how will you handle uneven terrain? Not saying you should do 3 wheels always, but I did want to make you think a second about this in an early stage of the project, so that your choice to do 4 wheels is not "obvious", but supported with some kind of reasoning :) Maybe it's easier to program, I don't know.

 

Anyway, the wheel units look nice and the choice of wheel is great for this, as well as the turntable. I assume that with the programming you can later control all wheels (drive and direction) independently for smooth movement :) I would try to minimize the vertical distance between the wheel axle and the turntable ring, to reduce unwanted forces on non-horizontal undergrounds. The lower the ring sits, the better the forces can be distributed. So, personally. I'd make things a tight fit as much as possible here. Also remember the turntable doesn't have to be at the same height as the ring. Also, curious as to how you will brace the turntable, as the ring in the digital picture isn't a whole-number height.

PS For strength: don't forget that 11x15 frames exist :) (and 7x11)

In any case ... I'll be following this project with great interest :D

18 hours ago, Mr Jos said:

You could use the inner 60T turntable to steer the wheel, and the banana gear racks as a bearing only, to allow smooth turning. When both are connected I think it should work pretty good.

I thought about that, but given the potential weight, I do think it's better to drive the 11x11 curved gear racks. I could use a worm wheel on the inner turn table, but then it would be better to use the old turn tables. They work better with worm wheels. For now, I will try driving the big ones.

12 hours ago, ninoguba said:

I will take what I learned from upgrading Blast to a biped and see how it works with a bigger version.

Will watch out for your updates and maybe can learn from each other. 😉

I'm sure we will. Please create a WIP topic if you have some design concepts or whatever. I'm interested to see what you come up with.

5 minutes ago, Erik Leppen said:

Just asking, but why is that obvious? I mean, it's a sensible design choice, especialy in the world of rectangular Lego bricks, but one advantage of having 3 wheels is that you're sure your model is always stable and all wheels always touch the ground. With 4 wheels, how will you handle uneven terrain? Not saying you should do 3 wheels always, but I did want to make you think a second about this in an early stage of the project, so that your choice to do 4 wheels is not "obvious", but supported with some kind of reasoning :) Maybe it's easier to program, I don't know.

Hmmm, that is actually an interesting thought. A very interesting thought. I automatically went for a four wheel design....just because. I started with a 3-wheel triangular design for my V1 robot. That didn't really work out. Partially because the triangular design resulted in a lot of stress on the omnidirectional wheels. And partially because I wasn't comfortable with building a triangular shaped base. But for this design, it might be feasible. I don't necessarily need to build a triangular base.

This is the reason I am posting my concepts early on. So people with more experience and insights can give me tips like these 
I will give this idea some thought.

9 minutes ago, Erik Leppen said:

Anyway, the wheel units look nice and the choice of wheel is great for this, as well as the turntable. I assume that with the programming you can later control all wheels (drive and direction) independently for smooth movement :) 

Yes, indeed.

10 minutes ago, Erik Leppen said:

I would try to minimize the vertical distance between the wheel axle and the turntable ring, to reduce unwanted forces on non-horizontal undergrounds. The lower the ring sits, the better the forces can be distributed. So, personally. I'd make things a tight fit as much as possible here. Also remember the turntable doesn't have to be at the same height as the ring. 

True that. I actually considered placing the rings below the middle of the wheels, but I quickly dismissed that idea. I can indeed play around with the vertical placement of the center turntable.

12 minutes ago, Erik Leppen said:

Also, curious as to how you will brace the turntable, as the ring in the digital picture isn't a whole-number height.

I am not entirely sure how to brace the two rings. I figured using the center turntable, so it won't fall off, but that will probably not be enough to brace the rings themselves. Maybe I will try a solution like Mr. Jos used. Not sure yet.

14 minutes ago, Erik Leppen said:

PS For strength: don't forget that 11x15 frames exist :) (and 7x11)

Absolutely. I have placed them in my Stud.io palette and I will definitely be using them.

15 minutes ago, Erik Leppen said:

In any case ... I'll be following this project with great interest :D

Cool. And as always; don't hesitate to question my questionable decisions 

31 minutes ago, Jim said:

True that. I actually considered placing the rings below the middle of the wheels, but I quickly dismissed that idea.

Why not both? My gut feeling is that a ring below the axle would be tricky to bear load, but it may help stabilise the wheel assembly if it's locked in between rings above and below.

15 minutes ago, pleegwat said:

Why not both? My gut feeling is that a ring below the axle would be tricky to bear load, but it may help stabilise the wheel assembly if it's locked in between rings above and below.

I guess they will be tricky to brace. But I did consider turning the wheel assemblies into "tubes" using lots of them.

This is a very interesting topic I will be following closely.

One thought I had is that, as you already mentioned, will be limited by the length of the wires on the PU motors.  I worry in this case that the wires won't be long enough for the two motors attached to the wheels.  As the wheels turn, the wires will twist together getting effectively shorter and stressing the wires.  One thing I might explore is to put the drive motors in the chassis and send an axle through the turntable.  This would allow unlimited steering rotation with no worries about the wires.  The downside, of course, is that it certainly be less powerful than your proposed two motor design attached directly to the wheels.  I wonder if the two motors may be overkill?

One other thought on driving the steering.  I have a similar design challenge with my 6 axis robotic arm.  In my latest incarnation (which I haven't published yet), I found a way to fit 4 12t gears driving the 60t turntable in the middle.  This ends up with a very solid feeling rotation axis that can turn a very heavy robotic arm using maximum acceleration on the motor.  I did this instead of driving the outer ring because I am using a bearing similar to the one used in the Rough Terrain Crane which interferes with any gear on the outer ring.  This makes disassembly and reassembly much, much easier than having a bunch of 1x1 round tiles as a bearing.

 

41 minutes ago, Glaysche said:

One thought I had is that, as you already mentioned, will be limited by the length of the wires on the PU motors.  I worry in this case that the wires won't be long enough for the two motors attached to the wheels.  As the wheels turn, the wires will twist together getting effectively shorter and stressing the wires.  One thing I might explore is to put the drive motors in the chassis and send an axle through the turntable.  This would allow unlimited steering rotation with no worries about the wires.  The downside, of course, is that it certainly be less powerful than your proposed two motor design attached directly to the wheels.  I wonder if the two motors may be overkill?

That is indeed something I have considered. But given the potential weight of the robot, I would like the number of gears between the motors and the wheels. I have also thought about a solution with two tires per "wheel unit" and a differential in between, so I can drive the differential with a 20T gear. Having unlimited would be very cool, although not necessary. 

Two motors might be overkill, but I'd rather have some extra power than not enough power. From what I understood the Large Angular Motors aren't that strong, so having two per wheel feels like a safe bet.

41 minutes ago, Glaysche said:

One other thought on driving the steering.  I have a similar design challenge with my 6 axis robotic arm.  In my latest incarnation (which I haven't published yet), I found a way to fit 4 12t gears driving the 60t turntable in the middle.  This ends up with a very solid feeling rotation axis that can turn a very heavy robotic arm using maximum acceleration on the motor.  I did this instead of driving the outer ring because I am using a bearing similar to the one used in the Rough Terrain Crane which interferes with any gear on the outer ring.  This makes disassembly and reassembly much, much easier than having a bunch of 1x1 round tiles as a bearing.

Hmmm Big Red. I totally forgot about that one. I do prefer your design over a gazillion 1x1 round tiles. Will it be a strong? I will download the instructions for Big Red and take a look. Thanks.

And please post a WIP for your robotic arm. I always love to see robotic arms. And I am interested to take a look at driving the inner turntable. Are you using a worm wheel or a gear/gears? I reread your message and you mention 4x 12T gears. I would like to see a pic of your setup.

Here's a couple quick and dirty pictures:

 

 

The top picture is of my robot base with the previously mentioned bearing.  The gear train is in the base but the motor is in the unit that sits on top of that (bottom picture).  I am using a medium angular motor as you can see there.  The 98585 pieces you see there are attached to the output of the motor.  I use that because it is a very rigid way to transfer torque. This setup acts a bit like a planetary gear.  If you notice the blue pins is both pictures, when the two pieces are attached together, 2L beams are put on those pins to lock the 2 pieces together.  This is critical in my application because when the arm is fully extended, it tries to pull this joint apart.

Once I make the Studio file, I intend to put it up on Rebrickable for free download (like my last version of the robotic arm already is).

That is one fine looking base (and upper structure). I can mount my wheel hubs to it and I will have a perfectly fine robot 

Using the 98585 pieces is what I have in mind as well (between the wheels and the motor, there is 1L space). I probably need another coffee, because I am not entirely sure how the motor and 98585 are driving the black ring.

I would love to see the Studio file. All these posts are really helping me out and I am enjoying them a lot.

Where can I get the new 12T gears? They are not on Pick a Brick and it's not currently listed on Bricklink (if I am not mistaken).

1 hour ago, Jim said:

Using the 98585 pieces is what I have in mind as well (between the wheels and the motor, there is 1L space). I probably need another coffee, because I am not entirely sure how the motor and 98585 are driving the black ring.

This is the interesting part of this new design.  Normally, one would have a gear train in the upper part with the motor  I didn’t have enough room up there.  Instead, I put the gear train in the lower base and sent an axle through the middle of the turntable to turn the gears in the lower part.  The gears in the lower part take power from the driving center axle and drives 2 gears, each of those drive 2 more gears which drive the 4 axles where the 12t gears are attached.  Once I have a Studio model, I should be able to make a good rendering.

1 hour ago, Jim said:

I would love to see the Studio file. All these posts are really helping me out and I am enjoying them a lot.

hopefully, I can finish that soon.

1 hour ago, Jim said:

Where can I get the new 12T gears? They are not on Pick a Brick and it's not currently listed on Bricklink (if I am not mistaken).

I ordered a bunch of those and the 20t gears while they were on Bricks & Pieces.  I received them a little over a week ago.  I don’t know where to get them now..

Note, for your application, I would put the gear train and motor that drives the turntable up in the body and not do the thing I described above.  My main point is that I think you can make a really solid large turntable by just driving the 60t turntable in the middle with 4x 12t gears and the right gear train.

19 minutes ago, Glaysche said:

This is the interesting part of this new design.  Normally, one would have a gear train in the upper part with the motor  I didn’t have enough room up there.  Instead, I put the gear train in the lower base and sent an axle through the middle of the turntable to turn the gears in the lower part.  The gears in the lower part take power from the driving center axle and drives 2 gears, each of those drive 2 more gears which drive the 4 axles where the 12t gears are attached.  Once I have a Studio model, I should be able to make a good rendering.

The 98585 is driving the four 12T gears? That's a nifty solution!

19 minutes ago, Glaysche said:

Note, for your application, I would put the gear train and motor that drives the turntable up in the body and not do the thing I described above.  My main point is that I think you can make a really solid large turntable by just driving the 60t turntable in the middle with 4x 12t gears and the right gear train.

Roger that.

19 minutes ago, Glaysche said:

I ordered a bunch of those and the 20t gears while they were on Bricks & Pieces.  I received them a little over a week ago.  I don’t know where to get them now..

Hopefully they will be available on Bricklink soon-ish. I am not in a hurry, so hopefully next month or so. Provided, I will need them for my solution. Maybe I don't need them at all. But I would like to have some in my collection.

2 hours ago, Jim said:

That is one fine looking base (and upper structure). I can mount my wheel hubs to it and I will have a perfectly fine robot 

And thank you for the kind words.  I was thinking this structure could make good shoulders for a humanoid robot.  I haven’t shown off the tilt axis yet.  It uses two large angular motors to drive 8x 12t gears around 2x 60t turntables which ends up being very smooth and powerful even under high load.

1 minute ago, Glaysche said:

And thank you for the kind words.  I was thinking this structure could make good shoulders for a humanoid robot.  I haven’t shown off the tilt axis yet.  It uses two large angular motors to drive 8x 12t gears around 2x 60t turntables which ends up being very smooth and powerful even under high load.

If we can find a way to mount my wheel units, I would love to turn your concept into a robot. That way I can fool around with the software etc. I am sure this robot will have many iterations, so I would love to test temporary concepts.

18 hours ago, Glaysche said:

If you notice the blue pins is both pictures, when the two pieces are attached together, 2L beams are put on those pins to lock the 2 pieces together.

How do you attach the 2L beams? From the pictures, it seems impossible to do as I don’t see a way a finger could push those beams to the pins once the upper part is in place.

 

It is a very clever way nonetheless!  I may borrow this one someday 😉

Edited April 26, 20222 yr by ninoguba
Added a compliment…

1 hour ago, ninoguba said:

How do you attach the 2L beams? From the pictures, it seems impossible to do as I don’t see a way a finger could push those beams to the pins once the upper part is in place.

You can place them from the bottom of the base.  You first bearing on the base, then snap the “shoulder” unit on.  This engages the axle for the gear train.  Finally, flip the whole thing over and reach in with relatively skinny hands and place the 2L beams.  It’s a bit tricky and takes a couple tries but not too bad. I might be able to use longer beams that are easier to handle.  I haven’t tried that.

1 hour ago, ninoguba said:

It is a very clever way nonetheless!  I may borrow this one someday 😉

Thank you!

Comes to mind that I can also take a look at the Liebherr for inspiration. That model uses a geared down 20T gear directly on the turntable.

I found this example and this is more or less what I am aiming at. A child sized robot, consisting of a base, body and head. And some sort of arms. Of course, my version will be much more rudimentary but this robot does reflect my ideas.

Here is my new version of my robotic arm on Rebrickable:

https://rebrickable.com/mocs/MOC-109607/glaysche/6-axis-robotic-arm-mk-2/#details

 

It has a freely downloadable Studio file and a bunch of renders of the interesting bits.  To finish the previous conversation, here are a couple renders of how the rotation axis works:

 

 

I plan to do a writeup on my Eurobricks topic on this and my previous version but I want to get some videos and pictures first which may take me a while.

Edited April 28, 20222 yr by Glaysche

Sweet! I will definitely be taking a look.

The other day I received the BMW wheels. They sure are impressively big and equally light. I was amazed by the weight, or lack thereof. They are much lighter than I expected. Hence I have been thinking about the overkill of two large angular motors and the possibility to drive the wheels from the upper structure, so they can rotate 360 degrees. This will obviously be a more convenient solution, because I don't have to worry about the wires.

Below is the concept I drew up in Studio. The space between the tires needs to be small, but this might be two studs too small. It might be hard to properly mount and brace the frame. Thoughts on this setup are welcome. I have ordered some extra tires and the hubs en brake discs etc, so I can use 8 wheels (4 double) in my robot. I will decide on 3 or 4 wheel units later.

Due to the high price, I haven't bought the large shock absorbers (yet), but they might be a nice solution for the arms. I will be needing 8 so I will wait till my final design before I push that button. The wheels are expensive enough already 

This is turning out the be a BMW M1000RR K66 Robot 

PS In my first renders I accidentally used the wide tires, but I am using the narrower ones.

And the front view (bit dark, sorry)

Oh, by the way, you can use "rubber black" instead of "black" in Studio so the tires don't render so glossy.

I'm not sure what the motorcycle hubs on the outside would attach to.  If you are willing to go wider, you could use different hubs and do something like this:

 

 

Edited April 30, 20222 yr by Glaysche

35 minutes ago, Glaysche said:

Oh, by the way, you can use "rubber black" instead of "black" in Studio so the tires don't render so glossy.

Sweet! That's an awesome tip. I'm still learning the Studio ropes, so these tips are more than welcome. There are still some nuisances, but I am definitely starting to appreciate this program. I love to try out quick concepts by drawing them in a digital design environment to get an idea of how to fit everything. And the render option is awesome.

35 minutes ago, Glaysche said:

I'm not sure what the motorcycle hubs on the outside would attach to.  If you are willing to go wider, you could use different hubs and do something like this:

Wow, thanks for building this example. It's not that I don't want to go wider, but I am/was afraid that wider would make it more difficult to turn the wheels. Do you think it matters a lot, if there is a big difference between your setup and mine? 

I can do both btw. Using your setup for the basic construction, which is better than mine in term of attachment points and bracing etc, and still attach the outer hubs to add some aesthetic stuff. Or only attach the disc brakes on the outside.

This does indeed look way better (black rubber) 

Here's a comparison. Looking at your image, I couldn't resist and I did order the wider tires as well.

Will a liftarm going straight up touch the tire? This is actually a question I should have asked before ordering the tires, but I wanted to have them in my collection anyway.

Edit: Never mind, I can test this in Studio