Alt-BEAM Archive
Message #10448
To: beam@sgiblab.sgi.com
From: "Timothy Flytch" flytch@hotmail.com
Date: Fri, 18 Feb 2000 23:11:59 PST
Subject: [alt-beam] Re: electric helicopters are getting close...
Hay David,
Ever fly a cricket??? know why you don't see many like them ??? they were
just too stable... you could take you hands off for long periods of time(
like 20 to 30 seconds at a time! I've seen it personally!)... the down side
is you can't make one fly inverted at two inches!!! so no one wanted a
boring to fly model I guess...
Timothy...
>being a model helicopter pilot (who can manage to make it fly for a few
>seconds than land again) of a kyosho methenol/oil/nitro powered beast ( i
>have way too many hobbies), i have to say that they are incredible
>difficult
>to fly and are really expensive, even electric ones. For instance - the
______________________________________________________
10449 Fri, 18 Feb 2000 23:30:13 PST [alt-beam] Re: Cutting Perf Board beam@sgiblab.sgi.com "Timothy Flytch" Over kill??? I use my band saw and then clean up the edges with a belt
sander... It is nice to have tools...
Timothy...
>greetings,
>
>It may seem like overkill but I use a Dremel with a fiberglass
>reinforced cutting blade. This is very useful for small cuts, but
>if you need to make a long straight edge cut, you should use a
>really sharp razor blade and a steel ruler, make a few passes
>then snap. I normally like to file down the cut sides so they
>are as smooth as the factory cut sides.
>
>
>Andy
>fco@total.net
>
>*********** REPLY SEPARATOR ***********
>
>On 18/02/00, at 1:11 AM, TurtleTek@aol.com wrote:
>
> >Greetings BEAMers,
> >
> > To try something new and perhaps even find a building method I prefer
>to
> >free-forming, I've built the usual photopopper photovore on a Perf board.
>I
> >knew I'd need to cut it so I bought a X-acto Razor Saw. Guess what- that
> >razor saw doesn't work so well. Not well at all. I had no idea perf board
>was
> >so hard to cut.
> >The question is: How would I go about cutting this stuff? What methods do
>you
> >use?
> >
> >Thanks
> >-Brien the TurtleTek
______________________________________________________
10450 Saturday, 19 February 2000 6:18 Re: electric helicopters are getting close... beam@sgiblab.sgi.com Timothy Flytch
>Hay David,
>Ever fly a cricket??? know why you don't see many like them ??? they were
>just too stable... you could take you hands off for long periods of
e(
>like 20 to 30 seconds at a time! I've seen it personally!)... the down side
>is you can't make one fly inverted at two inches!!! so no one wanted a
>boring to fly model I guess...
>Timothy...
>>being a model helicopter pilot (who can manage to make it fly for a few
>>seconds than land again) of a kyosho methenol/oil/nitro powered beast ( i
>>have way too many hobbies), i have to say that they are incredible
>>difficult
>>to fly and are really expensive, even electric ones. For instance - the
>
>______________________________________________________
>
10451 Fri, 18 Feb 2000 23:57:56 -0800 (PST) [alt-beam] Re: On making a better RJP...(webcams) beam@sgiblab.sgi.com Darrell Johnson Have you seen this site?
http://KhepOnTheWeb.epfl.ch/
lets you control a little Khepra robot from the web..
you can look from above, or from the bot's point of
view and drive it around using your web browser.. very
cool...
have fun,
darrell
--- Jim Taylor wrote:
> would be neat to see a webcam attached to a bot
> inside the RJP, or even
> place the web cam inside the RJP.
>
> |____|
> -------O()O-------
> James Taylor
> URL: http://fly.to/springmeadows
>
=====
____________________________________
BICOREEOS... They're BEAMtastic!!
http://www.geocities.com/beamtastic/
__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com
10452 Sat, 19 Feb 2000 02:48:36 -0700 [alt-beam] Re: solarbotics pager mtr+fan beam@sgiblab.sgi.com Dave Hrynkiw At 11:38 PM 2/18/00 , Timothy Flytch wrote:
>Only if the axels or coaxial... if you separate the blades then you do not
>have to use counter rotating blades to counter act tork... you use an
>offset or tilt of the blade disk to counter tork...
I disagree. If this were true, then many twin engine aircraft wouldn't be
going through so much effort to have props spinning in opposite directions.
Engine torque is a (counter)useful tool to turn aircraft in one direction
harder than the other.
Just like in any other physics system, the sum of the forces must be zero;
if there isn't an equal and opposite torque from the motors, it's going to
come from with resistance with the surrounding air, and produce rotation
effect.
-Dave
---------------------------------------------------------------
"Um, no - that's H,R,Y,N,K,I,W. No, not K,I,U,U, K,I,_W_. Yes,
that's right. Yes, I know it looks like "HOCKYRINK." Yup, only
2 vowels. Pronounciation? _SMITH_".
http://www.solarbotics.com
10453 Sat, 19 Feb 2000 05:47:13 -0500 [alt-beam] Biggest BEAM ap. yet? alt-beam@egroups.com "R. Martin Keen" If you remmber a few weeks ago someone had written an email about the
possible uses of beam tecnology. They included things like sun tracking
for solar panels and other practicle stuff. One thing wasn't
mentiontioned!!!!!!
Saving lives!
I don't know how many of you watched ABC news last night, but still no
emails written about it.
It's a 'neural program' that 'knows' how a plane should fly. It uses
feedback loops, can tell when the plane is upside down and can control
the wings, tail, rudder to cause the plane to fly normally. In tests they
disconected control to one of the wings, and instead of nosediving, the
plane rolled over and righted it's self.
Is it just me or does this sound very similar to the walker circuit? The
walker is neural, has feedback loops, and when you snip the motor wires,
or remove the leg, it keeps on going. It 'knows' how to walk (the lazy
way).
Interesting project for anyone who has a RC model plane should try this.
I suggest using an RC sailplane, since the gas would go until it crashes
or runs out of fuel.
Give a microcore the servo controls, and make a tilt sensor (weighted
pots for each axis). To make it self stabilizing, cause the tilt senors
to be negative feedback.
________________________________________________________________
YOU'RE PAYING TOO MUCH FOR THE INTERNET!
Juno now offers FREE Internet Access!
Try it today - there's no risk! For your FREE software, visit:
http://dl.www.juno.com/get/tagj.
10454 Sat, 19 Feb 2000 07:34:36 -0500 [alt-beam] Neural Plane: transcript alt-beam@egroups.com "R. Martin Keen" A followup to the biggest BEAM ap. yet:
A Plane That Adapts
Neural Net May Keep Aircraft Aloft When Damaged
By Michael J. Martinez
ABCNEWS.com
April 29 — When an aircraft is damaged, it takes a lot of skill, and a
great deal of luck, for a pilot to avoid crashing. Ailerons fail, rudders
and flaps become useless, making it nearly impossible to find the right
balance to keep the plane aloft.
But what if a plane could regain balance on its own, with little
action on the pilot’s part?
NASA is experimenting with software that can adapt to aircraft
malfunctions or combat damage. By checking an aircraft’s status six times
per second, the program can quickly adjust and keep the craft flying long
enough for the pilot to land.
“The challenge here is to flight-qualify a software program,” says
project engineer Mike Thompson of NASA’s Dryden Flight Research Center in
California. “We’re teaching it first to fly the airplane, then to fly it
when it’s damaged.”
Flight Data Re-Order
The new software, called a neural net, contains data on the aircraft’s
abilities and the way it’s supposed to fly. Then, in flight, it checks
that information against actual flight data. Using advanced mathematics,
the neural net can figure out what the plane can do — which flaps are
left working, and how to move them — to offset any damage.
That ideal scenario is still a long way off, Thompson acknowledges,
but within the realm of possibility. First, the neural net has to be
trusted to fly a perfectly functioning plane. Tests of that ability, in a
specially modified F-15, were successfully completed earlier this month.
“Takeoff and landing were normal, but once the plane got up to
around 15,000 feet, we engaged the neural net and let it handle the
flying,” says Thompson .
In a normal plane, directional maneuvers are handled mechanically.
The pilot pulls back on the control stick, which triggers the elevators.
With a neural net, the pilot is still in control, and would still pull
back on the stick to pull up. But that command would go first to the
neural net, which would then transmit the order to the elevators.
“It went as we expected,” Thompson says of the tests. “The plane had
fairly good handling, though the nature of the control system took a
little getting used to.”
Learn As You Go
The next step will be to simulate some kind of failure on the airplane,
and see if the neural net can compensate for the loss of a flap or a
rudder, for example.
Capt. Jim Smolka, project pilot. (NASA)
“This technology has the capability to reconfigure the plane in an
optimal way in order to give you the best flying qualities for either
continuing the mission or bringing the aircraft home for a safe landing,”
says NASA project pilot Jim Smolka, one of the pilots actually flying the
neural net-enhanced planes.
Smolka says he’d like to see neural net technology in commercial
aircraft someday.
“Civilian aircraft don’t have ejection seats, so you have to set the
airplane on the ground,” Smolka says. “There’s no other option.”
“The technology could be applied in many different areas where
controls are used,” adds NASA researcher Chuck Jorgensen. Factories,
power plants and other vehicles could all benefit from neural nets —
anyplace accidents happen and could be avoided.
If the predictions of the NASA experts are borne out in further
tests and research, the possibilities for using adaptive software as a
safety tool may indeed be endless.
S U M M A R Y
NASA thinks neural net software can keep damaged planes in the air long
enough to land safely.
“The challenge here is to flight-qualify a software program.”
Mike Thompson, NASA project engineer
Copyright ©1999 ABC News Internet Ventures. Click here for Terms of Use
and Privacy Policy applicable to this site.
________________________________________________________________
YOU'RE PAYING TOO MUCH FOR THE INTERNET!
Juno now offers FREE Internet Access!
Try it today - there's no risk! For your FREE software, visit:
http://dl.www.juno.com/get/tagj.
10455 Sat, 19 Feb 2000 09:13:53 EST [alt-beam] Microcore controller for Bicore aps... beam@sgiblab.sgi.com StickG@aol.com Is it possible to have multiple bicores regulated by say a four neuron
microcore? Is it useful to do such a thing? Would the microcore still need an
auto PNC circuit, because the bicores can regulate themselves? Has anyone out
there tried this?
stick
10456 Sat, 19 Feb 2000 10:30:57 EST [alt-beam] Re: photodiodes beam@sgiblab.sgi.com Blumojo13@aol.com on the top of these PDs there is a notch. Is the notch the anode or cathode?
blumojo13
10457 Sat, 19 Feb 2000 10:34:45 -0500 [alt-beam] Re: More fun questions beam@sgiblab.sgi.com Bob Shannon Sathe Dilip wrote:
> Using the ohms range of your multimeter, measure the resistance of the
> solenoid coil. Divide the voltage you want to apply by this value in
> ohms. You will get the value of current required to drive the solenoid
> from a DC source.
>
> e.g.,
>
> You want to drive the solenoid with 3 volts.
> You measure the coil resistance as 6 ohms.
>
> You will draw 3/6 = 0.5 Amps for the solenoid.
>
> Dilip
This is not quite right. The DC resistance of the coil
neglects its high inductance. This inductance will keep
the current down well below the value calculated above.
Of course the cap voltage will be falling rather quickly
as well, so the peak current will never reach about half
the value you might expect from the DC resistiance.
I've driven small bots with soliniods (early prototypes
of LEM did not use motors at all). They are not as inefficenct
as you might think.
For best performance, raise the SE trigger voltage rather than
add more capacitance. The higher the applied voltage, the
greater the current delivered. In a soliniod, the magnetic field
is proportional to the current.
Home