|
|
| |
Last weekend, we took a few robots over to the Central Illinois Robotic Club's
annual Bot Brawl. We had a bunch of fun and some pretty good results.
But, as usual, they are trying to figure out how to change the rules so this
doesn't happen again...
http://www.youtube.com/watch?v=sbYyPHBVr7Y
:)
Steve
ps, the kid in the yellow shirt is my son, who just lost to his grandmother.
When asked what he thought, he said he was proud of her.
|
|
| |
In lugnet.org.us.laflrc, Steve Hassenplug wrote:
> We had a small gathering to test out stuff for our upcoming remote control Rock
> Crawling event at Brickworld.
>
> This should be fun at Brickworld (http://www.brickworld.us)
>
> Steve
Hi
My son wanted an "real RC car" so I took the RC buggy body and made a few
changes to have a final result. It is a 4x4 truck driven with 2 RC motors, and I
adapted the RC steering mechanism for the front drive.
It's very large, and quite slow (reduced 45-1), but very strong. Unfortunately,
There is no tilt point or suspension, and it has a tendency to spin if one wheel
if lifted in the air.
The good side is its manoeuvrability (right word???) both for steering and the
distance from the comand. (I can be far away and still have a response)
On the first try, I had the differential in last reduction, but the inside gear
broke... so I had to put the diff one step earlier:
http://www.brickshelf.com/cgi-bin/gallery.cgi?i=2333020
Anyhow, the whole experience has been a LOT of fun.
There is a few pics on this folder after mod:
http://www.brickshelf.com/cgi-bin/gallery.cgi?f=235560
I did 2 more funny constructions, but no pics available yet.
LMKWYT, JM
|
|
| |
At 12:04 PM 2/7/07, Peter Ehrlich wrote:
> Heres a thought. Just have the slope going in the other
> direction. That way when it lifts, the balls keep rolling the same
> direction. (Faster and efficient!) If the raising is too bumpy,
> just find a way to place a third or fourth overhead rail keeping the balls in,
One down side here, is that the "ball-lift" must be shut-off when the
bridge is raised, and the controller for the ball lift is on the
"bridge" (other) side of the opening. But, along those lines, if
there is a bin to catch lose balls, the bridge could just lift up,
and disconnect the ramp at both ends. Then, when the bridge is on
it's way up, balls just roll back across the ramp to where they
started, and into a holding area...
Steve
|
|
| |
In lugnet.robotics, Brian Davis wrote:
> Not sure what you mean here. I was just picturing a permenant gently-sloping
> rail along the entire upper superstructure of the span, fixed in place (it would
> look something like a loooong shallow diagonal spar or cable span). With the
> bridge down, the slope is right to carry the balls across. As the bridge tilts
> up it obviously wouldn't work (wrong angle; any balls that are on it would roll
> the wrong way), but that's why you'd need the RCX to stop the lift mechanism a
> little before the actual tilting of the bridge span. I admit I was picturing the
> ball stream as being "lifted" on the end that has the pivot, and the "low" end
> to the far side (left, in most of your pictures), but it should work in either
> case.
Heres a thought. Just have the slope going in the other direction. That way when it lifts, the balls keep rolling the same direction. (Faster and efficient!) If the raising is too bumpy, just find a way to place a third or fourth overhead rail keeping the balls in,
Unloading the balls at the other end take some engineering. You could have the
rail mounted near the rotation point, (and possibly lower the rotation point and
put on a heavier weight). Or you could use about a foot of flex tubing to make
it all work far away from the bridge's rotation point. See what I'm saying?
Nice work by the way :-)
--Peter
|
|
| |
In lugnet.robotics, John Brost wrote:
> I think just cutting off the lift and waiting for them
> to drain out is best. The only thing I am worried about
> is how long it would take for the sloping ball stream to
> empty out before I could lift the bridge.
Not long at all. With the 3'+ slopping double-beam track I had at BF '06, I
think it only took less than five seconds for them to clear the end... and that
was for a *very* shallow slope. Given the time it takes to lift and lower the
bridge, adding a few seconds delay shouldn't cause any real problem.
> I don't want to take down the whole GBC because my bridge
> is slowing things down.
I wouldn't worry about it. First, the original train spec was to support
multiple stations with two trains at more than 1 bps... I'm not sure we've ever
had a full GBC running consistantly at 1 bps, and since at BF '06 we removed
some of the stations, that actually speeded things up. The trains could have
taken much bigger hits in efficiency and still kept up with the balls (in fact a
few times we ran for a while with just one train, so we had a saftey factor of
at least two part of the time. Additionaly, how often are we going to be lifting
the bridge? If it really slows things down, we can just go back to ducking under
it... and it will still be an amazingly cool addition to the GBC. Heck, if it
carries the ball stream then it's "really" just a GBCm that violates the
backplane rule :-).
> How long would you be willing to stand and wait for
> the bridge to go up so you could walk through?
We're talking about the GBC here: what on Earth does the practicality have to do
with it? This thing is just remarkably cool looking - it goes in.
--
Brian Davis
|
|
| |
In lugnet.robotics, Brian Davis wrote:
> Not sure what you mean here. I was just picturing a permenant gently-sloping
> rail along the entire upper superstructure of the span, fixed in place (it would
> look something like a loooong shallow diagonal spar or cable span). With the
> bridge down, the slope is right to carry the balls across. As the bridge tilts
> up it obviously wouldn't work (wrong angle; any balls that are on it would roll
> the wrong way), but that's why you'd need the RCX to stop the lift mechanism a
> little before the actual tilting of the bridge span. I admit I was picturing the
> ball stream as being "lifted" on the end that has the pivot, and the "low" end
> to the far side (left, in most of your pictures), but it should work in either
> case.
I am 100% thinking along the same line. At first I thought of simply stopping
the ball lift and immediately raising the bridge, letting balls along the slope
drain backward into a hopper, but after watching the bridge raise and lower, I
doubt any of them would roll smoothly back into the hopper. I think just
cutting off the lift and waiting for them to drain out is best. The only thing
I am worried about is how long it would take for the sloping ball stream to
empty out before I could lift the bridge. The balls will have to travel about
3' to cross the whole span. I suppose I could use a little more than a "gentle"
slope, but at any rate, I think speed is going to be key to keep the train going
and keep the whole GBC running smoothly. I don't want to take down the whole
GBC because my bridge is slowing things down. After all, I need to keep up a
1bps rate to stay within spec ;)
Guess I'll just have to try it and see.
How long would you be willing to stand and wait for the bridge to go up so you
could walk through?
John
|
|
| |
On the other hand, why would the bridge need to carry balls at all? Just put the
bridge on the section of track that separates the reliable modules from the
unreliable modules' loop.
But I suppose if the bridge does nothing at all with balls (not directly,
anyway), then it's not technically a ball contraption. :)
|
|
| |
In lugnet.robotics, John Brost wrote:
> At the bottom it is 10-wide, but the top horizontal
> rails are only 8-wide.
Ah, OK. No problem then, I was just worrying for a moment that you'd have to
further modify the whole thing.
> > mass of the counterweight?
>
> 794 grams... made out of 1/2 used AA batteries of course.
Cool. I have an additional 7.73 kg of AA's if we need them :-)
> The thing that has been stopping me [from feeding the ball
> stream itself] is what to do when the bridge is lifted.
Not sure what you mean here. I was just picturing a permenant gently-sloping
rail along the entire upper superstructure of the span, fixed in place (it would
look something like a loooong shallow diagonal spar or cable span). With the
bridge down, the slope is right to carry the balls across. As the bridge tilts
up it obviously wouldn't work (wrong angle; any balls that are on it would roll
the wrong way), but that's why you'd need the RCX to stop the lift mechanism a
little before the actual tilting of the bridge span. I admit I was picturing the
ball stream as being "lifted" on the end that has the pivot, and the "low" end
to the far side (left, in most of your pictures), but it should work in either
case.
--
Brian Davis
|
|
| |
In lugnet.robotics, John Brost wrote:
> That was sort of my thought. The thing that has been stopping me is what to do
> when the bridge is lifted.
You'd need a hopper to store balls that arrive while the bridge is raised. You
could probably rig a gate/door onto it that is open when the bridge is down and
closed when the bridge is up. Or the entire hopper could tilt backwards with the
bridge deck so no balls can exit it until the bridge goes back down. That's
probably even easier.
|
|
| |
This weekend, I was working on a FLL robot for the Science Central
event in a couple weeks. That went well. It's not 100% reliable,
because it's hard telling how many atoms are coming off the table,
and sometimes the stain-resistant fabric just doesn't resist stains
(dirt sticks).
But, last night, after THE game (to Colts), I showed it to Heather,
and I got a perfect score in 67 seconds. I ran it five times, and my
low score was 388 (missed 1 dirt & 1 or 2 atoms). I'm not sure my
average time, because my program wasn't resetting the timer, like I
thought it was, so my time was starting sooner than I thought. But,
67 was about normal.
Now, I need to finish up my NXT-Line Follower, build an NXT-Maze
robot, and help Matt build an NXT-Sumo. :)
Steve
|
|
| |
In lugnet.robotics, Brian Davis wrote:
> In lugnet.robotics, John Brost wrote:
>
> Very slick-looking John. It seems odd that one of the neatest attractions in the
> GBC may be only tangentially related to soccer balls, but that is a very nice
> design. I especially like the electrical cutouts at both ends. I have to ask,
> have you tested it with a train?
Not yet... No train. In fact, all the train track is borrowed from Steve H and
Steve's Mom. I've got track on order, but nothing else. Someday maybe I'll buy
a train ;)
> > ...it is designed for 6-wide trains.
>
> ...especially since the GBC cars have a bump-out below the baseplate?
Actually factored that in. At the bottom it is 10-wide, but the top horizontal
rails are only 8-wide.
> What is the mass of the counterweight? And what did you make it out of?
794 grams... made out of 1/2 used AA batteries of course.
> That would be wonderful to see, although it implies a way cut off the ball flow
> (gravity driven) over the span before the bridge is actually lifted. Manually, a
> simple switch throw, wait for the balls to drain from the span, and then lift
> the span would work... but since the bridge is key-activated, if the RCX could
> be driving the ball lift, upon key insert it could shut off the ball lift and
> wait for the balls to drain away, before then lifting the bridge proper.
>
> > Oh, almost forgot, now with youtube video goodness!
> > http://www.youtube.com/watch?v=Yi9e82UC6WY
> > http://www.youtube.com/watch?v=xvHrUY5Lw0E
>
> Very cool. I think it should be pretty easy (if not as pretty) to thread a ball
> stream across it. At BF'06 I had a long (more than 3') double-rail of inverted
> Technic 16L beams that worked great, with a little sag. Feeding balls onto it
> would be the only tricky part (off the other end should by much easier).
That was sort of my thought. The thing that has been stopping me is what to do
when the bridge is lifted.
> Nice work, John!
Thanks,
John
|
|
| |
In lugnet.robotics, John Brost wrote:
> I suppose this doesn't have to really be for the GBC only,
> but that was why I decided to build it.
Very slick-looking John. It seems odd that one of the neatest attractions in the
GBC may be only tangentially related to soccer balls, but that is a very nice
design. I especially like the electrical cutouts at both ends. I have to ask,
have you tested it with a train?
> ...it is designed for 6-wide trains.
...especially since the GBC cars have a bump-out below the baseplate?
> Thanks to Brian Davis who helped me figure out the
> balance/counterweight issues I was having.
You're more than welcome; you clearly put a lot of time into it. I see what the
problem is now that I see where the counterweight is (and why it's got to be
there, of course). Nice solution. I really like the ket as well... now, we'll be
able to tell who "should" be in the GBC enclosure by the LEGO keys around their
neck (hanging fromn a chain of chainlinks, of course ;-).
What is the mass of the counterweight? And what did you make it out of?
> There's no GBC ball stream ability right now across
> the bridge
That would be wonderful to see, although it implies a way cut off the ball flow
(gravity driven) over the span before the bridge is actually lifted. Manually, a
simple switch throw, wait for the balls to drain from the span, and then lift
the span would work... but since the bridge is key-activated, if the RCX could
be driving the ball lift, upon key insert it could shut off the ball lift and
wait for the balls to drain away, before then lifting the bridge proper.
> Oh, almost forgot, now with youtube video goodness!
> http://www.youtube.com/watch?v=Yi9e82UC6WY
> http://www.youtube.com/watch?v=xvHrUY5Lw0E
Very cool. I think it should be pretty easy (if not as pretty) to thread a ball
stream across it. At BF'06 I had a long (more than 3') double-rail of inverted
Technic 16L beams that worked great, with a little sag. Feeding balls onto it
would be the only tricky part (off the other end should by much easier).
Nice work, John!
--
Brian Davis
|
|
| |
I suppose this doesn't have to really be for the GBC only, but that was why I
decided to build it. I've been playing with this idea since Brickfest '06 when
Steve Hassenplug suggested a way to go in & out of the GBC area without crawling
under the tables would be nice. This is the second attempt and still needs a
little work, but it is getting much closer. Thanks to Brian Davis who helped me
figure out the balance/counterweight issues I was having. The first version
only had a span of around 15 inches or so. This one clears 2 ft easily. Here's
the Brickshelf folder.
http://www.brickshelf.com/cgi-bin/gallery.cgi?f=231442
For the train-heads out there, it is designed for 6-wide trains. There isn't
enough room for an 8-wide.
It is powered by an RCX, set inside one of the yellow bridge towers. The RCX
powers both the bridge raise/lower function and will also power the bridge track
and a section of track leading up to the bridge so trains won't go flying into
the counterweight when the bridge is up.
On the far side, the bridge hits a polarity switch when it is lowered that can
be used to turn on/off power to the track on the far side of the bridge so
trains don't go over the edge when the bridge is up.
http://www.brickshelf.com/cgi-bin/gallery.cgi?i=2294265
http://www.brickshelf.com/cgi-bin/gallery.cgi?i=2294266
The raise/lower function is keyed, so only those with a key can operate it.
This will keep kiddies from making it go up & down all the time at shows. You
can see the "key" here.
http://www.brickshelf.com/cgi-bin/gallery.cgi?i=2294268
and here's where you put the key in. It is 3 touch sensors in an AND
configuration so all 3 have to be pressed to register the touch on the RCX.
This is embedded in the second bridge tower.
http://www.brickshelf.com/cgi-bin/gallery.cgi?i=2294245
There's no GBC ball stream ability right now across the bridge, I think this may
be a possibility, something I'll probably explore.
Oh, almost forgot, now with youtube video goodness!
http://www.youtube.com/watch?v=Yi9e82UC6WY
http://www.youtube.com/watch?v=xvHrUY5Lw0E
Comments/Suggestions please!
John
|
|
| |
In lugnet.org.us.laflrc, Steve Hassenplug wrote:
> Here's a good example of what we want to do...
>
> http://www.youtube.com/watch?v=lMv8M8s0f88
>
> Steve
Wow, those guys went all out, they even "towed" one truck with another. Pretty
nice. At some times it was hard to tell those were RC and not real trucks.
Speaking of this, when is our next get together? Since I missed the last one, I
need to see how my Red Rock Crawler is going to stand up.
John
|
|
| |
The second set of arms on Mike's robot gives it better mobility. They make its
overall tread length much longer, so it can treat stairs like a ramp instead of,
well, stairs, and that makes it faster at climbing them. The arms also let it
raise the entire chassis over certain obstacles, and it doesn't have to turn
around to use its arms because it has them on both ends.
|
|
| |
In lugnet.org.us.laflrc, Rafe Donahue wrote:
> I say we just award the prize now.
Ha! Methinks you underestimate the talents (not to mention obsession) of the
other folks in this event <grin>. Seriously, thanks for the kind words, and I'm
very happy with it, but now the rest of you know yet another way to try to do
it. Go for it: somebody make a robot that will drive over LNE. Incidently, for
RC combat this thing is fun too. Just put a blade or two on the front flippers
and it is even more fun (although I need to shift the CoM... perhaps
dynamicly... dang, this thing really does need more than three motors I fear).
> The kids will go nuts when they see the video!
Thanks. I took it to a public "science night" a couple days ago, but with faster
gearing... it ran basicly continuously for 2.5 hours, driving over kids, feet,
LEGO, and anything else they could manage. And the minifig never fell out :-).
--
Brian Davis
|
|
| |
In lugnet.org.us.laflrc, Jordan Bradford wrote:
> you've rediscovered the design of a real-world
> reconnaissance robot...
well, in all honesty I shamelessly implemented a solution I had already known
about: PackBot by iRobot. I won't remotely claim the idea originated with me, as
it didn't. I was inspired by the "real world" here.
That said, I'd not run across the link you posted - thanks. I wonder if
FirstResponseRobots and iRobot have ever gotten into infringment lawsuites (or,
perhaps, who actually holds the patents on this sort of system; it may be
neither company).
--
Brian Davis
|
|
| |
In lugnet.org.us.laflrc, Philippe Hurbain wrote:
> Brian, it would help to reproduce your outstanding
> design (and maybe CAD it!) if you could provide higher
> resolutions photographs somewhere...
Thank you both Matthias & Philo (and others!). give them an inch indeed...
first, I really feel i need to apologize for *not* CADing this up. I've tried to
CAD stuff, but the combination of my chosen platform (Mac) coupled with where I
like using my free time (building, not CADing) and furthermore the complexity of
the designs I'm actually interested in documenting (like this), have conspired
to make me a lazy slob who's not documented things in the "normal" format.
Somebody, please help me, or brow-beat me some more, and I'll try to rise to
that standard sometime.
As to higher-res photos, agreed: these were actually reduced for BS, and I've
got higher-res ones availible. I probably need to do an even *better* job
documented the construction, not just the parts but the sequence: for studless
constructions like this, I sometimes feel like I'm making one of those "slide
the pieces to assemble the cube" wooden puzzle. For instance, one of those
images I had to use my fingers to point out three red (in that "step") pins that
need to be pulled to remove a long white composite beam that is critical to
holding the two shoulders on... but is actually partially hidden within the
shoulders on the ends.
Where do you guys want more/better photos? What needs improved documentation?
--
Brian Davis
|
|
|
