| |
"Brian Davis" <brdavis@iusb.edu> wrote in message
news:JCAvs8.H94@lugnet.com...
> The more I thought about this, the more I liked this idea: I wanted to
> make a
> really high-mobility vehicle, but the problem with wheels is they are
> really
> stopped by steps greater than their own radius (well, partially, ignoring
> pushing from behind). Not to mention that the Really Cool wheels (Power
> Puller)
> are a tad hard to find, so... I thought of PackBot, and with the higher
> torque
> NXT motors, thought I might make a stab at it. The result works pretty
> well:
> handling piles of laundry, crawling over LEGO, climbing 30° wood slopes,
> uprighting itself after flipping, smashing minifigs, etc. The only problem
> is it
> needs just a *little* more length to reliably crawl over my leg, and of
> course
> treads are not the highest friction in the LEGO universe (and, I'm out of
> treads).
Have you considered using the new 5-wide treads found in the new Technic
Snowmobile? They seem a lot more suited to driving over most surfaces with
their added width, and they're relatively new, so perhaps we can find out if
they're really as good as they look.
> The first problem was transmitting the drive motion through the "shoulder"
> joint
> of the flipper arms. Then the second, bigger problem was reinforcing the
> heck
> out of the structure to keep it from tearing itself apart: the first
> version
> could easily torque the whole front off, while this one doesn't even flex.
> And,
> it's all studless*
If you are going to get more photographs, I could use ideas on how to use
the studless elements in constructions. I'm at a relative shortage of
studless model instruction books to take techniques out of so I'd appreciate
a rundown any studless techniques you can explain in failry simple terms.
> Brickshelf folder of static images:
> http://www.brickshelf.com/cgi-bin/gallery.cgi?f=229082
>
> And a few videos. I like the one where it crawls over the old Technic
> Space
> Shuttle...
> LNE vs the Empire: http://www.youtube.com/watch?v=iNXHKtnFBcc
> Studless meets studded: http://www.youtube.com/watch?v=4ik640MLj6k
> On point: http://www.youtube.com/watch?v=WwBYvs7vq_0
Okay, the first was pretty good, the second it seemed to have a few problems
getting over the Space Shuttle itself, and eventually looked like it was
staggering around drunkenly after finally having succeeded, and the final
video was not too bad but a little short, I'd have appreciated a bit of
camera zoom/pan or the model running around showing how well it can turn.
> *OK, there are six studs: I needed someplace to put the crazy minifig
> driver.
> More curiously, I found myself having the same odd issue that I think John
> Brost
> mentioned earlier: it's gotten so natural to build studless for me now,
> that it
> was hard to try to make it a "hybrid". I think I can do a little better,
> actually for a hybrid version.
I wish I could get building studless "natural" myself. I'm intrigued by the
ways you can put it together, and it's a lot more complex than the simple
system implies. Especially when you actually have to think, now, how am I
going to take the power off the NXT and drive it down the side of my model
to somewhere I can use it.
As I said earlier, I'd appreciate any ideas on studless building you can
explain in fairly simple terms.
In fact, I'm hoping someone out there has or is working on a compilation of
studless building techniques.
A comment on your model's driving rpinciples, have you thought about trying
to get the treads themselves to slide as well as turn, so that you can not
only flip them but turn them over to the other side so as to take more
advantage of the fact that one side is slightly raised near one end? It
might make the model a bit bigger, but maybe you can make a mechanism that
accomplishes it with a gear rack that runs off an idler sprocket from the
main drive.
Also, for a wheeled version, have you tried building a ganged pair of front
wheels in a triangular formation? There was a movie that featured a very
large custom-built vehicle where the wheels simply rotated backwards,
flipping slowly backwards around the collective axis if they hit something
they couldn't easily drive over, until they drove over it.
I'd imagine it would be pretty complex to build in LEGO nowadys, but doable.
:)
Cheers ...
Geoffrey Hyde
|
|
| |
"Geoffrey Hyde" <g.hyde@bigpond.net.au> wrote in message
news:JCAxx2.8KF@lugnet.com...
> Have you considered using the new 5-wide treads found in the new Technic
> Snowmobile? They seem a lot more suited to driving over most surfaces
> with their added width, and they're relatively new, so perhaps we can find
> out if they're really as good as they look.
http://www.bricklink.com/catalogItemInv.asp?S=8272-1
I forgot to include the URL of the set I was talking about. Here is is
above.
Cheers ...
Geoffrey Hyde
|
|
| |
In lugnet.org.us.laflrc, Geoffrey Hyde wrote:
> Have you considered using the new 5-wide treads found
> in the new Technic Snowmobile?
Yes, and no. The problem is they are hard plastic with ridges, so if a tread
ridge catches on a rough portion of the surface, they push fine... but if the
surface is smooth, they'll just slide (very low coefficient of friction). The
width is nice, but even better is the fact that the new Technic treads can be
made in various lengths by adding/removing tread links, and since they are wider
than the driving sprockets the support and driving mechanisms can be completely
hidden within the tread footprint, much more like PackBot. Those treads would
also allow me to make drive sprockets the same radius as the turntable
"shoulder", again much more in line with how PackBot does it (there's a number
of modifications, such as the shape of the flipper treads and their offset, to
get around this little problem).
The other problem is simply expense. I've got a bunch of those treads, because I
wanted to experiment with them, but I stil don't have the six linear feet of
tread that this *small* PackBot analog would require.
> if they're really as good as they look.
For some things, yes; the pin holes on them are really well done. They're not
quite as adjustable as the "old" chain and tread links, but that's just a size
issue (and at that small size, firm connections were... a lot more difficult).
> I could use ideas on how to use the studless elements
> in constructions.
Yes, I need to add some more detailed "exploded views".
> it seemed to have a few problems getting over the
> Space Shuttle itself...
It did, and I thought about re-shooting it... but hey, that's reality. Given
enough takes, I could always make it perfect :-). As to the "druken stagger"
after it makes it over, that's partially on purpose (to show the flexibility of
the platform) and partially because it's hard to control (remembering which way
I want to turn the flipper arms is tough when it's upside down, etc.)
> I'd have appreciated a bit of
> camera zoom/pan or the model
> running around showing how well
> it can turn.
I'll work on that: driving the model takes two hands, which doesn't leave a lot
free for running the camera :-).
> In fact, I'm hoping someone out there has or
> is working on a compilation of studless building
> techniques.
I've been asked that a couple of times, and I guess I should make up some
examples and mechanisms... but it's always surprising to me, because I feel like
I am a rank amateur at studless construction (at least every time I look at
LEGO's official models). But yeah, consider me bugged.
> have you thought about trying to get the treads themselves
> to slide as well as turn [etc.]
I'm not sure I understand what you mean: can you exaplin it better, or build a
simple model to show what you mean?
> have you tried building a ganged pair of front
> wheels in a triangular formation?
That configuration is called a "tri-star wheel", and appeared in the movie
"Damnation Alley". And yes, it's been done in LEGO, and done well. Search LUGNET
for "tri-star wheel", or just go here:
http://www.visi.com/~dc/
--
Brian Davis
|
|
| |
"Brian Davis" <brdavis@iusb.edu> wrote in message
news:JCBttB.GEJ@lugnet.com...
> In lugnet.org.us.laflrc, Geoffrey Hyde wrote:
>
> > Have you considered using the new 5-wide treads found
> > in the new Technic Snowmobile?
>
> Yes, and no. The problem is they are hard plastic with ridges, so if a
> tread
> ridge catches on a rough portion of the surface, they push fine... but if
> the
> surface is smooth, they'll just slide (very low coefficient of friction).
> The
> width is nice, but even better is the fact that the new Technic treads can
> be
> made in various lengths by adding/removing tread links, and since they are
> wider
> than the driving sprockets the support and driving mechanisms can be
> completely
> hidden within the tread footprint, much more like PackBot. Those treads
> would
> also allow me to make drive sprockets the same radius as the turntable
> "shoulder", again much more in line with how PackBot does it (there's a
> number
> of modifications, such as the shape of the flipper treads and their
> offset, to
> get around this little problem).
I see. I wonder if we'll ever see some kind of rubber plug-in outer surface
for these treads? Since they already have mounting holes, it only remains
to figure out what would best suit a tread so as to improve it's grip.
Ideally it would be like a car tire's tread pattern whereby the tread
presses harder given more weight.
> The other problem is simply expense. I've got a bunch of those treads,
> because I
> wanted to experiment with them, but I stil don't have the six linear feet
> of
> tread that this *small* PackBot analog would require.
:-) I can completely understand the expense angle. I'm hoping that since
there are even more of these treads in the bulldozer which is also due out
this year, that the expense angle can become somewhat cheaper. And I think
that the bulldozer will also have a lot of extra useful parts. ;-)
> > if they're really as good as they look.
>
> For some things, yes; the pin holes on them are really well done. They're
> not
> quite as adjustable as the "old" chain and tread links, but that's just a
> size
> issue (and at that small size, firm connections were... a lot more
> difficult).
I think there was some saying about size to usability ratio increases as you
get bigger - not sure what it was actually worded like, though.
> > I could use ideas on how to use the studless elements
> > in constructions.
>
> Yes, I need to add some more detailed "exploded views".
True, or if you have the parts in an LDRAW program derivative, you could
probably explode-assembly them there. I have to investigate a program I
have downloaded here called MLCAD, anyway. If you wanted to, you could
provide a suitable file format for downloading, or have it written out as a
text LDRAW file, that would be okay I think. But I also need to try and
update the LDRAW parts library.
> > it seemed to have a few problems getting over the
> > Space Shuttle itself...
>
> It did, and I thought about re-shooting it... but hey, that's reality.
> Given
> enough takes, I could always make it perfect :-). As to the "druken
> stagger"
> after it makes it over, that's partially on purpose (to show the
> flexibility of
> the platform) and partially because it's hard to control (remembering
> which way
> I want to turn the flipper arms is tough when it's upside down, etc.)
:-) Maybe the front arms could be made somewhat longer in length?
> > I'd have appreciated a bit of
> > camera zoom/pan or the model
> > running around showing how well
> > it can turn.
>
> I'll work on that: driving the model takes two hands, which doesn't leave
> a lot
> free for running the camera :-).
:-) All I could suggest is perhaps rework the control unit you're using,
btw, can you include some shots and explanations of that? Or are you using
a bluetooth communication to your PC/mobile to drive it?
> > In fact, I'm hoping someone out there has or
> > is working on a compilation of studless building
> > techniques.
>
> I've been asked that a couple of times, and I guess I should make up some
> examples and mechanisms... but it's always surprising to me, because I
> feel like
> I am a rank amateur at studless construction (at least every time I look
> at
> LEGO's official models). But yeah, consider me bugged.
I know, but every time I look at instructions for studless I start to wonder
which way I should orient everything. I've been experimenting with the
pieces themselves, but getting the hang of studless sometimes is difficult
if what you've built is mostly the standard out-of-the-box models. It is
certainly paying off though, I'm expanding my learning of studless the more
I work at it the more I know what I can and can't do.
Also, changing from one orientation to another to send the drive shaft where
you want it can be tricky at times, and leave you wondering which direction
you originally wanted to go in anyway. :-)
> > have you thought about trying to get the treads themselves
> > to slide as well as turn [etc.]
>
> I'm not sure I understand what you mean: can you exaplin it better, or
> build a
> simple model to show what you mean?
Well, I'm not actually sure it's possible to do, given how complex it
already is, but basically give them the ability to slide forwards and
backwards (extend and retract) along their own length. It may require more
parts, gearing, and motors than is practical, anyway. But imagine a sliding
door moving back and forth.
> > have you tried building a ganged pair of front
> > wheels in a triangular formation?
>
> That configuration is called a "tri-star wheel", and appeared in the movie
> "Damnation Alley". And yes, it's been done in LEGO, and done well. Search
> LUGNET
> for "tri-star wheel", or just go here:
A very nice vehicle indeed! I notice nobody has yet tried studless
construction. That would indeed be a true test of model building skill to
accomplish that. :-) If they were available in sufficient quantity, I'd
use the large wheels from the power puller set, as they'd make pretty good
tires for such a model and give it plenty of ground clearance to boot. :-)
Cheers ...
Geoffrey Hyde
|
|
| |
In lugnet.org.us.laflrc, Geoffrey Hyde wrote:
> I wonder if we'll ever see some kind of rubber plug-in outer surface
> for these treads?
I would like to hope so. Those mounting holes would seem perfect to secure
rubber cleats to, especially with the ridges on the hard plastic tread
underneath. That one piece woulkd change them from semi-decorative to fully
robotic functional (not that they aren't functional now).
> Ideally it would be like a car tire's tread pattern
> whereby the tread presses harder given more weight.
Tire treads are there to handle things like wet or dirty roads, not to just
increase friction when pressed down harder (there's some good basic physics
reasons friction doesn't work that way... it turns out those reasons are wriong,
but it seems the concoulsions hold :-).
> If you wanted to, you could provide a suitable (LDRAW or MLCAD)
> file format for downloading... [etc.]
Despite some very very good reasons to kump into LDRAW or similar with both
feet, I've never taken the plunge. First, I've had no luck getting them working
on my Mac (OSX), and second, if I have the time I usually want to play with real
bricks, not the virtual ones. *Someday* I'll be forced to do it.
> Maybe the front arms could be made somewhat longer in length?
Not while I'm limited to the "stock length" rubber treads. I could get them
about 3 units longer by offsetting the "front" of the arm tread from the center
of the shoulder joint, but at a minimum cost of making the whole chassis at
least 4 beams wider, and probably a lot weaker or at least a lot heavier.
> > driving the model takes two hands...
>
> perhaps rework the control unit you're using...
The problem is the robot has three degrees of freedom: left tread, right tread,
and flipper arm position. So I need to work at least three controls on the
remote. I'm planning on partially automating the flipper arms (so they can be
use din "full manual" mode, or handled by single button presses or voice
command). But for full control, you really need full constant control of three
control surfaces, and that's tough to do with just one hand.
> btw, can you include some shots and explanations of that?
The hardware for my BT remote is pretty much laid out in my Brickshelf folder:
http://www.brickshelf.com/cgi-bin/gallery.cgi?f=203384
The software, as well as a short "how to" document, Steve Hassenplug was kind
enough to host on his website (down near the bottom, although take a look at the
rest too - Steve has popped some nice content up there):
http://www.teamhassenplug.org/NXT/
> give them the ability to slide forwards and
> backwards (extend and retract) along their
> own length.
Yes, dynamicly changing the tread length would be great, and doable in a couple
of ways. For instance, here's something similar:
http://www.astolfo.com/bots/dynatrax.asp
There's also the "shape shifting tank" design:
http://news.lugnet.com/technic/?n=9672
The problem comes in at the number of motors. With three motors, I can really
only have three full degrees of freedom (OK, yes, I could use splitters if I
wanted, but then it could only move forward or something similar... not really
in the spirit of a high-mobility system) unless I want to go with 3rd party
multiplexers. And honestly, i'll probably try that once HiTechnic comes out with
something. But at least initally, I wanted this acessible to anybody with "just
LEGO", so three motors was the limit.
--
Brian Davis
|
|
|
