Remember this movie, with George Peppard? This was the movie with the cool tri-wheeled armored transport they used to escape the brutal environment after the world had a nuclear war. The vehicle had four of these tri-wheels on it, and not only did each wheel rotate, but the entire tri-wheel could rotate freely.
I just completed the design of a tri-wheel, using nothing but Lynxmotion SES parts.
You will see this tri-wheel used on a robot I will build one day. The wheel span is just a little over 7" between the centers of any two of the three wheels.
Iâve never heard of this movie before, but if you are referring to a tri-star design, then I know what you mean. Can you post a picture?
Sounds really cool.
The wheel span, center to center, between any two wheels is a little over 7" now. The wheels shown are 3 3/4" diameter and 1 1/4" wide, so larger wheels could be used on this version of the tri-wheel.
Um, how are you going to prevent the servo leads from becoming a big twisted knot? It would seem like you are going to need commutators for the power and signal leads using that approach, rather than a planetary gear drive system.
Each wheel assembly will have its own microcontroller controlled by the main controller via i2c. There will just be an i2c and power connection running through the axel to the wheel assembly which will connect to the wheelâs microcontrontroller.
I have not quite figured out how I am going to prevent the power and i2c connections from getting all twisted up. I will research commutators and see what that gives me. I could always hang a 9v battery on each wheel assembly, but that would be much less than the optimum solution. At least I will only have two connections to deal with to keep untwisted.
Interesting implementation of the Landmaster wheel configuration.
If youâre looking for cheap and dirty slip-ring style connections that can rotate freely and have no more than three conductors, I have seen 1/4" stereo plugs and jacks used to this effect before. Itâs hardly an ideal solution, but I imagine that it would last well enough to let you experiment with the configuration and decide if ytou wanted to continue, before spending a whole lot of time and effort on something that may or may not work for you.
If youâre passing power to wheel-cluster-mounted processors through a moving coupling like that, be sure that you have it well buffered, as there will almost certainly be periods of âless-than-optimal conductivityâ, and microcontrollers generally donât like that sort of thing. Data lines also donât like dropouts, and the higher the communication speed, the more susceptible your communications will be. Fortunately, photons donât suffer from twisting the way that wires do, so you might find that you can shine an IR beam down the center of your slip-ring/axle assembly, to be received by an IR detector on the other end, and achieve your data link that way.
Looking forward to seeing your progress with this one.
I will have just two connections, power (two wires) and I2C (also two wires) going through the axel down to each wheel assembly. Iâve been thinking about using a bluetooth link, but that would only work for the data transmision, not power and ground.
I am researching this now. A comutator wonât work because it depends on a constantly rotating assembly, which will not be the case here. The entire wheel assembly would only rotate when climbing over something. I just have to get power past the end of the main axel to where the wheels are. If I can get power/ground and two data lines past that point, I wonât even need a separate microcontroller for each whe
If I could just transmit power over a standard wireless link, that would be ideal. Unfortunately, I think this is just a dream and not realistic.
Something like this could work, because the two IR sets could be stationary even though one would potentially rotate at times - it would still remain in the same fixed position horizontally.
This still does not solve the power transmission problem though, but I know there is a solution as there has been a LOT of research done over the years into ways to transmit power.
I am going to switch the location of the ball bearing hub to the other end (wheel side) of the axel. It makes more sense to have it closest to the part that will actually be rotating freely.
I have found a company called Moog Components Group where I may find the solution I need. I have contacted them and requested a catalog. They have a local distributor, so I may be able to wrangle some free samples of their slip-rings if I talk it right and they are interested in what I am attempting to do.
I also just found Mercotac which could be promising for my application. I will research this further.
Itâs just a matter if whether they have a solution I can afford and/or if I can get sample parts.
I got my info and catalog CD from Moog Components Group yesterday. They definitely have the slip-rings covered with plently of options. I am not sure I can afford their potential solutions for my problem though, but maybe I can get a sample or two from them.
I still have to contact Mercotec about their products. They do not make slip rings, but have another good way of accomplishing the same thing.
Not to be contrary, but wouldnât it be a lot easier, cheaper, and more robust if you simply had a gear arrangement driving the wheels? Picture a gear in the center, right on the main axle, surrounded by three additional gears. Now extend this a bit so that the three additional gears are (or connect to) wheels. (Or you could use drive belts instead of gears⌠but you get the idea.)
This way, there are no electronics to get twisted up at all. Also, you need only a single motor for each triwheel (or two, if you want one to control when the triwheel assembly rotates rather than just leaving it to physics).
Of course you probably couldnât do this with just SES parts⌠but thereâs SES purity, and then thereâs getting the job done.
This would be fine, but I still have to get power and i2c through the part of the axle that can rotate freely (which allows the entire tri-wheel to rotate). Otherwise, there will be wires that can get all twisted up if the entire wheel assembly rotates.
I may not be completely understanding the solution you are proposing.
I know it canât be done with just SES parts, but there is always a possibility of having SES compatible adapters or whatever made at some point if this works out.
No you wouldnât, because there is nothing beyond the axle that needs power or signals.
Yes, my description wasnât very clear. Let me try a crude sketch:
Those spiky things are gears (please ignore the fact that the teeth donât actually mesh ââŹâ my drawing program is rather limited), and the dark round things are the wheels. The big gear in the center drives all three of the wheels, via the chain of small gears on each arm. No electrical power or signals go to the wheels at all; the only power transmission is of the rotational kind, via the gear chains. The gear chains and the wheels themselves would be attached to three beams (not shown) that come together in the middle.
I could probably prototype this for you in LEGO pretty quickly, if you like. (Or, quite likely, thereâs someone whoâs already done it ââŹâ those LEGO technic builders are prolific.)
hmm, iâve been following this post for a while, and your idea seems good jstrout, but wouldnât that waste power having to run wheels off a single motor? that eould strain that motor a bit, ecspecially if he is going to be going over objects with the tri wheels. plus he wouldnât have the freedom of using each motor individually
My robot, with these wheels, would be a climber and would go over things. I am also planning to use Open Servos, which would allow me to sense the load on each wheel, position, speed, etc.
The whole idea of this is to duplicate the functionality of the Landmaster wheel configuration as closely as possible. I really like that wheel design and think it would be great for a rover.
I do see the benefit of the other way of doing this, but am just not sure it fits with what I want to do. with the design. I am also not sure of all that gearing would add a lot of extra weight or not.
Oh, OK, I understand what you are describing now. I think that method would probably work, but am not sure it really fits what I am trying to do - duplicate the Landmaster wheel assembly functionality as close as possible. It seems that using one motor per wheel would cause less stress on each motor than using one motor to drive all three wheels. Wouldnât more stress on a motor mean more power being required?
I am concerned about the weight added by the addition of the gears as well as the possible added power requirement. This setup would also likely take me away from the SES orientation of my design.
I donât see how it wastes any power except for rotation of the upper wheel ââŹâ but since that wheelâs meeting no resistance, itâs not going to be consuming much power. Almost all the power from the motor will go to moving the vehicle, via the two wheels in contact with the ground.
As for the freedom of using each motor individually⌠well, what would you use that freedom for? You could eschew rotating the upper wheel, of course, but the bottom two wheels should always be turning at the same speed and direction anyway, I would think.
Note that this isnât my idea ââŹâ as far as I know, such triwheels are always built this way. I remember first reading about it in a robot book when I was a kid (which, alas, was a long time ago!). I seem to recall that at one point it was considered as a possible design for a lunar/Martian rover, but the author liked it for robots too.
then again, think of it in this way to, i mean it would have more strain on the single motor, but what it that motor gave out while you were driving it?
you would lose all power to that group of tri wheels, readering the vechile virtually useless besides a very slow speed, however, if each wheel were driven by a single open servo each, then you could simply rotate the whole tri wheel assembly and it would allow for the remaining 2 motors to continue movement of the vehicle
also, as odd as this seems, rotation of a single wheel while going over rough terrain can be essential at times
such as if you cant get 2 wheels on a surface due to extreme angles, rotation of a single wheel would help, now i realise this can be acomplished but again comes in the factor of using vaulable energy, ecspecially if its going to be a climing bot, because it will use alot of energy moving its weight with the motors that rotate the tri wheel assembly
I know this is a very old topic but I just wanted to get to know how did this design work? I saw some movies on youtube showing that such solutions enables robots to climb stairs, can yours do this to? Are there any tradeoffs of tri-star wheel system from your experience?
And finally, If I wanted to build my own tri-star wheel robot (such as Landmaster), are there any parts from robotshop you guys would recommend?
Oooh. Shiny! 
