Alt-BEAM Archive
Message #10564
To: beam@sgiblab.sgi.com
From: TINDEH@aol.com
Date: Mon, 21 Feb 2000 00:44:09 EST
Subject: [alt-beam] Re: BGM motors
Does anybody have a scan or pic of these BG motors with the specs.A scan of
the motor, preferably the ones that come out of the lens unit. You can send
it directly to me if the file is to big, Any info will be appreciated.
Thank you,
AT
10565 Sun, 20 Feb 2000 22:28:36 -0800 [alt-beam] Re: MPJA geared motors beam@sgiblab.sgi.com Bruce Robinson BUDSCOTT@aol.com wrote:
>
> I bought a couple of these motors from MPJA.
> they run fairly well under the 6 VDC. the question
> is how do i mount them?
Any way that works :)
OK, here's a site for doing research. Figure out what type of robot
you're interested in and see if you can find some examples on this site.
http://www3.telus.net/rfws/bwi/
Please don't use this site just for cruisin' because
a) it isn't set up that way.
b) I'm already way over my ISP's traffic limit for this month and
one of these days they're gonna crack down.
The walker section is reasonably complete and well organized. The rest
of it isn't (complete OR well organized). Guess what kind of robots I'm
interested in.
Regards,
Bruce
10566 Sun, 20 Feb 2000 22:32:13 -0800 [alt-beam] Re: MPJA geared motors beam@sgiblab.sgi.com Bruce Robinson Dennison Bertram wrote:
>
> These are actually very poor motors to use on a walker. The
> reason why that gearbox looks so funky is because it uses a
> worm gear. This is VERY innefecient and is just ingeneral
> difficult to use in walkers.
What's so inefficient about a worm-gear-motor?
Bruce
10567 Sun, 20 Feb 2000 23:57:10 -0800 [alt-beam] Re: BEAM: Tendency toward miniaturization beam@sgiblab.sgi.com Bruce Robinson Jean auBois wrote:
>
> Isn't it odd that people are building smaller and smaller devices
> that are often more and more fragile? If there was ever a field
> that is moving toward nanotechnology, this would be the one.
>
> ...
>
> I don't think it is because we _can't_ control bigger things with
> BEAM technology -- perhaps it is just easier to build smaller
> things that work fairly well.
Exactly right, Jean.
After reading this ever-growing list of comments and ideas, here's a
"gedunken" experiment to think about.
Imagine a solar powered walker that will just fit comfortably on the
palm of your hand. I know, it won't be going far under solar power. Put
some backup batteries in it as well, to take over when solar doesn't
work.
Now, scale it up by a factor of five; that is, every linear dimension
increases 5 times. Let's see what happens.
- 5 times wider and higher. Legs are 5 times longer.
- surface area is 25 times greater.
- volume and mass are 125 times greater.
- stresses increase 5 times (assuming identical material).
Right off the bat, we're faced with a dilema. What about the speed of
the motors? If we leave the speed the same (bigger motors, of course),
then our robot will be moving 5 times as fast (legs are 5 times longer).
Is this what we want? It probably is ... can you imagine a BIG robot
moving at the same speed as a tiny one? We EXPECT bigger stuff to move
faster. So ...
- same motor speed.
- All linear motion is 5 times as fast (because everything's
5 times longer).
- Momentum (mass x speed) is 625 times greater.
- Kinetic Energy is 3125 times greater.
So, first conclusion: Solar power becomes less and less practical as the
robot gets bigger, because the ability to collect solar energy increased
less than 1% of the demand for energy.
Second conclusion: Battery power also becomes less practical as the
robot gets bigger, because the ability to store energy (proportional to
mass) increased only 4% of the demand for energy.
Next, momentum issues. Not only does the momentum of the whole robot
increase, so does the momentum of those legs which are swinging back and
forth.
Therefore, third conclusion: we now have to worry about acceleration
(and decceleration) of the robot's limbs, because we're dealing with a
lot more momentum AND the stresses are higher. We may not want, or be
able to, simply fire those motors up at maximum speed right off the bat.
And, fourth conclusion: any rocking motion is going to cause us problems
because momentum has increased by a factor of 625. We'll be consuming
energy just to overcome the rocking motion. And we'll be damaging things
(including the robot) as it tips and bumbles along. On the other hand,
momentum may PREVENT our large robot from shifting its weight in the
manner that makes smaller robots able to walk.
Fifth conclusion: the robot will behave entirely differently, because
momentum is 625 times higher. Where our small robot would run into an
obstacle and then scrabble away trying to get over or through it, our
big robot will smash into the obstacle. In short, our large robot is far
more capable of damaging it's enviroment then a small robot.
Sixth conclusion: load-carrying bearings at joints and gearboxes will be
subject to 5 to 25 times as much force. Bearing surfaces increase by a
factor of 25, but mass increases by a factor of 125 and dynmic loads
increase by a factor of 625.
Seventh conclusion: many of our sensors will have to be a lot more
robust.
These effects will force us to design our large robots differently.
- We will contrive to make them lighter, using hollow
structural members.
- We will add more reinforcing.
- We will use larger bearings. Shaft support geometries will change.
- We will abandon solar power.
- We will look for alternatives to battery storage. Internal
combustion?
- We will design control systems that control motor acceleration,
not just speed.
- We will design more complex motion systems to keep the robots more
stable. This will probably mean more motors and more articulation.
- We will incorporate more feedback systems.
- We will have more problems tuning our control systems.
And I'm sure you can come up with many, many more.
--------------------------------------------------
Now let's go the other way. Let's take our palm size solar/battery
walker and scale it down by a factor of 10 this time. It'll fit on your
thumbnail.
- 1/10 as wide and high. Legs are 1/10 as longer.
- surface area is 100 times less.
- volume and mass are 1000 times less.
- stresses are 1/10 as great (assuming identical material).
- same motor speed.
- All linear motion is 1/10 as fast (because everything's
1/10 as long).
- Momentum (mass x speed) is 10,000 times less.
- Kinetic Energy is 100,000 times less.
First conclusion: Solar power becomes more and more practical as the
robot gets smaller, because the ability to collect solar energy
increased 1000 times more than the demand for energy.
Second conclusion: Battery power becomes more practical as the robot
gets smaller. But solar will soon surpass it. Batteries are entirely
viable for working in dark areas.
Momentum of everything decreases dramatically. So, third conclusion: we
can forget entirely about acceleration. We can even forget about the
mass of the legs we are moving.
And, fourth conclusion: any rocking motion is trivial.
Fifth conclusion: the robot will behave entirely differently, because
momentum is so low. It will now go OVER just about any obstacle it can
get a grip on. It's strength will be all out of proportion to it's mass.
There will be fewer and fewer things in its environment that it can
break or damange.
Sixth conclusion: load-carrying bearings at joints and gearboxes will be
subject to 10% down to 1% times as much force.
Seventh conclusion: we can use fragile sensors. Anything capable of
damaging them will finish off the robot.
These effects will ALLOW us to design our small robots differently.
- We can be less concerned with structural strength.
- We can eliminate reinforcing.
- We can use alternative materials for bearings, materials that
would not stand up on our palm-sized robot. Shaft support
geometries become simpler.
- We can use direct solar power.
- Batteries can be used for dark areas.
- Our control systems can become simpler. Using mechanical stops to
limit leg motion will no longer cause damamge.
- We can use a minimum number of motors. One motor robots become
feasible.
- We will need fewer feedback systems. We might not need any.
- Our control systems will be tunable by one or two adjustments at
most.
BUT
- Individually our robots will become less capable.
- Teams of specialist robots become feasible and economical.
---------------------------------------------
So why is the trend to smaller? Because it's less work. Because we get
visible results faster. Because there are fewer things to go wrong. Look
at the two words I emphasized in the two summaries. Bigger FORCES us to
design differently; smaller ALLOWS us to design differently. We don't
like to be dictated to.
But you know what? Look at those last two items. Teams of specialists.
We are tending toward smaller and smaller, and we will be forced to do
more and more co-ordination. And what we will end up with (perhaps) is a
system of very small robotic elements that must function together to
accomplish even a simple task. Which is exactly where we would end up if
we were heading toward bigger and bigger. Only on the large scale they
would be physically attached to one another.
So will the outcome be any different? Is the trend to smaller merely
giving us an illusion of progress? Have we, by good luck (or evolution)
stumbled on the optimum size to work from?
Hey, I'n not gonna answer those questions. I figure I'll get flamed
enough for some of these ramblings.
Fire away,
Bruce
10568 Mon, 21 Feb 2000 03:00:07 EST [alt-beam] Re: MPJA geared motors beam@sgiblab.sgi.com TINDEH@aol.com I have used those display motors before and they were not worm driven. The
only problem is these are not well suited for walker style robots. To much
current is needed to push these motors the only thing I have yet to try is to
use a 9-volt bat, with a 7805.
PS
If you pry the plastic casing off the motor you have metal to solder your
frame to. The other thing is you need to reattach the plastic gearhead base
with epoxy on the motor.
10569 Mon, 21 Feb 2000 00:21:24 PST [alt-beam] Re: BEAM: Tendency toward militarization beam@sgiblab.sgi.com "Timothy Flytch" >Hey, I'n not gonna answer those questions. I figure I'll get flamed
>enough for some of these ramblings.
>
>Fire away,
>Bruce
Ok Burce you asked for it... LOL... Just kidding...
But I do question your scaling practice... I have built R/C aircraft for
twenty years now and I have found that scaling up is a lot lighter than
scaling down... When you build a curtious robin for example in .40 size
then scale it up to quarter scale... it uses halve the scale power and
weighs 1/2 as much in scale numbers... Yes it is true that not the same
materials are use but you get the idea???
Timothy...
______________________________________________________
10570 Mon, 21 Feb 2000 00:56:41 PST [alt-beam] Re: Was solarbotics pager mtr+fan Is: Torque and propellers beam@sgiblab.sgi.com "Timothy Flytch"
format="flowed"
Ok dave,
I'll answer your little fame.... one more time...
Try opening up a history book... look up the "FockeAchgelis Fa 61" it was
the first true helicopter... It had two side by side rotors that were NOT
counter rotating... Anna loved it, even flew it inside a gym!
Then try the "H21C FLYING BANANA " It was a tandem (inline) production
helicopter that did not have counter rotating rotors either... Later models
were changed to counter rotation to shorten the fuselage by over lapping the
rotors...
While you are in the library grab a book on gyro copters... you will see
that there is NO full size twin rotor gyros, only models use them...
>(for those who want to know: precession is the little wobble motion in the
>vertical axis when you spin a top or a gyroscope).
What are you trying to say here???
Try this... Take a paper plate and punch a hole through the middle ... put a
pencil through it and spin it like a top... using a straw blow down on one
side of the plate ... you will notice that the plate dose not dip were you
blow but about 90' later in it's spin ... That is gyroscopic precession...
I just have to ask what is a "crap-load" anyway??? How dose that equate to
Lb/Ft???
I'll attach a couple of pictures for you ...
Timothy...
>At 01:47 PM 2/20/00 , Timothy Flytch wrote:
>>I would like to ad that the v-22 osprey uses counter rotating rotors
>>because of precession problems in the forward flight mode... not to
>>counter tork...
>
>
>Love to hear the source of this little tidbit. Propellers that HUGE are
>going to generate a crap-load of torque. I can't believe it's strictly for
>the precession problem.
>
>(for those who want to know: precession is the little wobble motion in the
>vertical axis when you spin a top or a gyroscope).
>
>-Dave
______________________________________________________
Attachment: FockeAchgelis Fa 61-(JPG - 21k)
Attachment: H21C FLYING BANANA
Home