Syndicate This Blog |
Going high risk steampunkTuesday, January 20. 2009
In which your narrator finally gets stuck into exploring other prime movers
for future reprap machines...
Some time ago, Tigertaler, one of our more creative Reprap fanatics, did a
series of design charrettes for printable stepper motors
which he published on Youtube. His designs {example shown above} looked at using
solenoids as a motive force for cleverly designed kinematics to achieve stepping
motion in a highly heterodox manner.
His logic is impeccable. It is a lot easier to wind a solenoid coil than it
is to do the windings and rotor construction for virtually any kind of
electrically powered motor you'd care to name. I'd mooted this notion around the
Reprap Core Group before. Nophead noted that the frequencies that such a
solenoid might have to reach to provide a useful combination of step size and
rotational speed could well be beyond what could be achieved. Given that Nophead
has likely forgotten more than I'll ever know about electronics, this tended to
dampen my enthusiasm at the time.
Last weekend, however, a concatenation of events caused me to revisit
Tigertaler's work. The previous week I'd very successfully milled quite a
handsome anti-backlash assembly for a tin-can stepper powered linear drive. I
built the assembly around a 1/4-20 threaded rod with the intention of driving it
with a Kysan
25BY2414 rated for 12v and 0.2 amps/winding. When I got the assembly
together it became obvious that the Kysan was not going to have adequate torque
to propel the thrust collar at meaningful speeds.
I had two choices at that point. I could redesign the assembly for #6-32
threaded rod or I could buy a beefier tin can stepper. Anaheim had
beefier tin can steppers at beefier prices for one-off. Both options were going
to take more time while going for a beefier stepper was going to increase the
price of the assembly. None of that pleased me to any great extend so I make a
third choice and decided not to decide for a few days.
The whole experience had brought one lesson home to me. When designing
something that takes power it is not always that obvious, to me anyway, just how
much power it will take to make it go around. For testing purposes it's always
nice to have a lot more torque than you need just to test concepts.
That was when the notion of using pneumatics came back to me. Pneumatic
actuators power output depends on the pressure of the air supply. I'd bought
myself a compressor for Christmas with something like that in mind. To date it
has been doing yeomanry duty keeping swarf out of my milling jobs. It is,
however, capable of providing up to 150 psi (~10 atm) of air pressure and can
easily drive nail guns and pneumatic wrenches. Pneumatic actuators are very
small for their power output. As well, I should be able to mill them out of HDPE
without a great deal of drama.
At that point I began to give Tigertaler's most ambitious printable stepper
a very close look.  While it looked quite complex at first glance, in fact it was merely a repetitive presentation of a relatively simple mechanism. At that point I set about to develop a script that could do a slice description for milling.  Now it remains to do a few modifications to my slice and dice app to accept a slice directly instead of insisting on generating it out of an STL file. I hope to be milling my first Tigertalar Steam Pumk stepper motor in the next day or so. Trackbacks
Trackback specific URI for this entry
No Trackbacks
Comments
Display comments as
(Linear | Threaded)
"Nophead noted that the frequencies that such a solenoid might have to reach to provide a useful combination of step size and rotational speed could well be beyond what could be achieved. "
No what I actually said was "Well if you are moving at say 16mm/s in 0.1mm steps you would be buzzing away at 160Hz.. Not impossible but pretty fast for a mechanical system."
It is certainly possible to have fast solenoids. The old dot matrix print heads are a good example. They used 24V and get very hot but they moved a lot faster than 160Hz.
It was the mechanical issues of wear and tear that worried me, not to mention the noise.
Could you have multiple solenoids and wheels. Each one pushing it just a fraction. kinda like a V8 pertol engine.
So from the example
16mm/s in 0.1mm steps you would be buzzing away at 160Hz
or two at 80hz
or 4 at 40hz
or 8 at 20hz
you get the idea
Stephen
I hear what you're saying but I can't visualise it. Maybe a quick drawing and a builder's blog entry, perhaps? :-D
I see a few ways of doing it.
depending on what you are trying to achieve.
Let's work with two solenoids working on a single wheel/cog which has teeth on either/both sides
Each sol will work on one side.
idea 1
You put half the groves on one side and half the groozes on the other.
Sol 1 fires and moves the disk
Sol 2 is waits in its current position
then
Sol 1 is fixed in current position and sol 2 fires.
In this way you spead the wear and tear and load.
Idea 2
The cog as the same amount of teeth on each side but they are 180 degrees out of phase.
As one sol is changing direction the other is in the middle of it's "power" stroke.
Grin
Hope that helps.
Stephen
I'm going to need some visual aides to make this work in my stupid head. :-(
Look at the animation on this web page. You will see that a large part of the time the cog is not moving. just imagine another cog with another pushy upy downy thingy attached to the first cog which is moving when the first cog is not.
like a radial piston engine? http://commons.wikimedia.org/wiki/File:Radial_engine.gif
Superficially, anyhow. :-)
I'm excited to see where it goes!
My inner nitpicker says "extent" not "extend" and "actuators' power" instead of "actuators power"
...But that irritating comment just shows how interested I am, right?
Interesting!
I kind of doubt this will be practical as a motor, but on the other hand I might have said that for Ultrasonic motors too (http://en.wikipedia.org/wiki/Ultrasonic_motor).
That's why I said "high risk". Milling on a light frame positioning system like Tommelise wasn't supposed to work, either. :-D
A google search for:
Pneumatic stepper motor
shows there is a lot of interest in this concept to make devices that can work in operating MRI machines.
You know pneumatic wobble motors are also worth considering:
http://www.act.sys.okayama-u.ac.jp/kouseigaku/research/okamoto_wobble_06/english.htm
Bloody fascinating! Thanks for the link! I've got to think about this. :-D
I bet there's a better PDMS bladder design than the one they used...printing the pneumatic bladder with a paste extruder, and the gears from polymer, you should be able to print the whole motor.
Also, I want to float these two ideas: 1. If solenoid valve frequencies make stepping this impractical, the wobbling could open and crimp channels printed into the PDMS, pressurizing and exhausting the various chambers of the bladder in sequence.
2. Solenoid valves that connect the wobble motor to compressed air inlet and the exhaust can be set up in a push-pull arrangement, so that gas can flow through in either direction.
Oddly enough, if these ideas were used in parallel and the system could be rigged to allow only a set volume through after each idle period, the motor would have the signal properties that Tigertalar wants.
this design is similar to the micro motors used for autofocus on high end optics (USM and the like)
Shoot. That's just reminded me of something. There was an old IBM robot (the 7565) that used 'linear hydraulic motors', which were four cylinders actuating rollers against a waved surface. By operating the cylinders in the right order the robot would move along the track kinda like a linear stepper motor.
In order to extend the work envelope all you have to do is bolt down an extra length of sinusoidal track.
I don't know if the principle is useful here, perhaps with different actuators?
Best picture I can find is here:
( http://www-03.ibm.com/ibm/history/exhibits/robotics/robotics_2.html )
The wavy track for horizontal and vertical axes is visible.
Please note that (IIRC) this robot used separate magneto-restrictive linear transducers for feedback, but I'm not sure if that was to increase accuracy or because they couldn't rely on the hydraulic system as an open-loop actuator.
I think youre all missing my point here. this and several other unpublished designs is NOT an intent to design a great motor but to rid of stepper electronics. the controll specs for all my design are as follows two bit input: 00 idle. 10 step up. 01 step down, 11 backlash lock. I have clear resons for this but thats another part of my total reprap plan
Sorry if I misrepresented your intent. I knew about your intent to get rid of stepper electronics, which your designs do quite nicely. I was seeing your designs from a rather different angle, however. :-)
Normal air compressors are terribly noisy, like sleeping with the hoover bag on. How can we deploy this in a home environment?
You are absolutely right about the noise. I came up with the idea primarily to make it possible to deploy such milling/printing Repraps in more remote, off-the-grid locations in developing nations. You find small, engine driven air compressors quite commonly there, in areas where a 12v battery is about the only reliable source of electricity there is. 12v auto batteries will easily give you enough electricity to run control circuitry but is largely impractical for running things like stepper motors for days at a time. Compressed air is another matter, though. :-)
Returning to the noise problem. Presently, air compressors are designed to be used in workshops and on construction sites and as such use a very cheaply designed compressor. There is no reason by a quiet, multi-piston compressor could not be designed and deployed. I figure that if there is a need for a quiet air compressor for the home environment it won't be long before some factory in China starts making one for sale in the big box stores. :-)
Hi Forrest!
I'll admit, I'm a complete addict to neat mechanical devices such as this. However, if you think about this working with a load pushing the wheel the other way, it becomes obvious that the motor will not make progress.
So, you'll need at least two actuators. Then again, it is easy to add multiple actuators in this design. Also, as shown in another video by Tigertalar, this motor can directly be unrolled and used as a linear motor. Having multiple actuators at slight offsets allows for smaller step size. Of course, as the Tigertalars stated goal is getting rid of stepper electronics, this wouldn't help much.
Question is: why get rid of stepper electronics. My main reason to want to get rid of stepper motors is that they are very inefficient (certainly the way we drive them) and make the "run off of a car battery" idea unrealistic. Also, we'll never be able to make useful steppers with printed parts and standard materials.
Mercifully, the x and y axes don't impose a counter force and there is enough friction in the z axis to counteract it. :-)
|
