 | | | | | | |
| |
|
Hello again,
I think I should have another go at this... I wanted to say "Check
this out, I found an interesting application for the old pneumatic
cylinders" What it came out sounding like was "Mr. Soh was wrong! Look
at me prove him wrong!"
Sorry about that. That wasn't my point. Anyway, If you have a look at
the links I posted earlier, you can see how I used the old red
cylinder asd a single acting cylinder with spring return:
http://www.ecf.utoronto.ca/~stehlik/images/ideas/pneumatics/old3.jpg
This type of pneumatic cylinder is very common in industry, as it
needs only one air line, and uses half the amount of air. To do this
with Lego, the only trick was putting elastics around the piston to
provide the spring return, and plugging up the unused port on the
valve. Actually, this same thing can be done with the new cylinders,
just don't connect a hose to the top port. The old style cylinders
just offer more speed and less friction. Originally, in the old
pneumatics system the air would be pumped in to extend the piston, and
the air would be sucked out to retract it. As several people have
pointed out, the suction does not work too well to pull the cylinders
back. Correct me if I am wrong, but this type of 'suction retraction'
is not commonly used in industry. I found that spring return is much
more effective, and you don't have to bother with the suction at all.
About the vacuum tanks... I won't beat this point to death, but I have
a few thoughts. According to the laws of phsics, gas flows from a high
pressure region to a low pressure region. When you compress air into a
tank, you are essentially creating a high pressure region. The air in
the tank will be at slightly higher pressure than atmospheric, so it
want to flow out of the tank. Now, lets do the opposite thing with a
suction pump: By pumping air out of the tank, you create a low
pressure region. This is by no means a vacuum, since you can not
remove all of the air with a lego pump, but describing it as a vacuum
tank seems to make sense. Since the air outside of the tank is at
atmospheric pressure, and inside the tank is at a lower pressure, the
air want to flow into the tank. Its really just the opposite of
compressing air. Now, in both cases, the air will move only as long as
the pressure difference is maintained. So if you pump up an air tank,
you can extend and retract a cylinder a few times until the pressure
in the tank drops to atmospheric pressure. With the low pressure tank,
you can suck air into it until the pressure increases to atmospheric
pressure. So... my window walker can climb happily away as long as the
tanks have a lower than atmospheric pressure in them. If I pump most
of the air out to start, the robot can climb about 6 steps before
needing a recharge. I suppose it would work without the air tanks, but
not as well:)
Rob
| | | | | | | | | | | | | | | |
> Its really just the opposite of compressing air.
True, but there is one notable difference, as others have pointed out
before:
When compressiong air, the pressure difference is limited only by
implementation details. The pressure inside the tank usually is not just
slightly higher than outside, but multiple times the outside pressure.
When removing air from the tank, the theoretical limit for pressure
difference between inside and outside is the atmospheric pressure itself.
And real implementations in the price range we are talking don't even get
near that, so the amount of energy you can store by sucking air out of an
air tank is very limited.
> If I pump most
> of the air out to start, the robot can climb about 6 steps before
> needing a recharge. I suppose it would work without the air tanks, but
> not as well:)
I suppose it would work even better if you could figure a way to use
pressure instead of vacuum. But that may not be feasible ...
Greetings
Horst
| | | | | | | | | | | | | | | | | |
| |
|
Rob Stehlik wrote:
>
> Hello again,
> I think I should have another go at this... I wanted to say "Check
> this out, I found an interesting application for the old pneumatic
> cylinders" What it came out sounding like was "Mr. Soh was wrong! Look
> at me prove him wrong!"
How coincidental, but it too struck me that we needed to backtrack as
the thread was getting pretty muddled up. It seemed to me that we may
not be working on the same assumptions. So here's the re-play...
The post that you quoted http://news.lugnet.com/robotics/?n=16764
was in response to Jennifer Clark's suggestion to use two air tanks, one
on the high pressure hose and the other on the vacuum hose between the
pneumatic diode and the valve.
The tacit assumption here is that the pneumatics were being used in the
conventional manner, that is with pneumatic cylinders being controlled
by the valve to retract and to extend the piston. It is in this context
that it was said that the air tanks would not work with the old
pneumatics. (The experimental setup is shown at
http://www.geocities.com/cssoh1/primer/oldpneu.htm).
But perhaps, it would have been clearer if I had qualified that the air
tanks would not work with the old pneumatics *when connected in the
conventional manner* (hopefully this is easily understood).
> ... my window walker can climb happily away as long as the
> tanks have a lower than atmospheric pressure in them. If I pump most
> of the air out to start, the robot can climb about 6 steps before
> needing a recharge. I suppose it would work without the air tanks, but
> not as well:)
Now, the window walker is an exception. It is not using the old
pneumatics in the conventional manner (as defined above). Indeed, it is
using pneumatics in a most irregular manner. There are none of the old
pneumatic cylinders to speak of. Therefore there is no need for
alternate compression and suction. The pneumatic valve is simply being
used to switch the *vacuum* from one pair of suction cups to the other
pair. So I reckon this critter can be left to continue carrying the
tanks on its back ;-)
--
C S Soh
CSSoh's Lego Pneumatics
http://www.geocities.com/cssoh1
... where air is power!
| | | | | | |
