Cheapo linear actuator using a pager motor?
- Here's the basic idea: a miniature fly ball governor-type mechanism coupled directly to a pager motor shaft. At rest the fly weights are drawn close together near the motor shaft axis. When the motor spins the weights fly out radially, and the scissor-linkage bars pull the loose collar axially along the shaft axis. The pull bar is a loose running fit so, though the flaball mechanism is rotating, the pull bar does not rotate. An extensional spring provides the restoring force when the motor stops.Has anyone tried something like this?-Bobby
- I've not seen that exact design used before, but, of course, just about ever steam power locomotive ( or steam engine in general ), over their last century of active use, used a governor just like that, and quite successfully. I think the main concern I would have would be in finding the right balance of weights, governor size, and spring tension to match the motor power and generate the desired movement with sufficient force. It could work quite well, but I'd expect it to take some fiddling.
Martin Jay McKeeOn Thu, Feb 6, 2014 at 3:39 AM, <rlnansel@...> wrote:
Here's the basic idea: a miniature fly ball governor-type mechanism coupled directly to a pager motor shaft. At rest the fly weights are drawn close together near the motor shaft axis. When the motor spins the weights fly out radially, and the scissor-linkage bars pull the loose collar axially along the shaft axis. The pull bar is a loose running fit so, though the flaball mechanism is rotating, the pull bar does not rotate. An extensional spring provides the restoring force when the motor stops.Has anyone tried something like this?-Bobby
- The restoring spring is going to subtract from the available pull. The steam engines I have seen used gravity acting on the balls for the restoring force rather than a spring and the govenor didn't do much work, it just closed down a valve for the steam.
You could calculate the available pull from the assemble but for a pager motor it won't be much at all and you would need a very low friction thrust bearing to do the pulling.
Back in the early 1960s Peter Holland in Radio Control Models and Electronics used a similar flyball design to drive the steering winch on a 6 inch long radio controlled car. The flyballs driven by a Kako-1 (IIRC) worked as a friction clutch inside a cup which turned the 1/32 inch diameter winch which wound in thread to pull on an Akerman steering rod to steer the wheels. A spring returned the wheels to the opposite lock when the flyballs disengaged. The 1/2 inch diameter cup and 1/32 inch winch gave a 16:1 mechanical advantage which was still only just enough to steer the wheels when the car was moving.
There are small linear servo using pager motors and leadscrews, see
with an 80g pull for $4.29
A simple leadscrew actuator is going to much easier to make.
Yeah, the spring is a problem. However, I have made direct-driven leadscrews with hobby motors and 6-32 threaded rod. I definitely wouldn't classify them as "easy" to build, at least without a lathe. But they are fairly cheap.
The string-twisty actuator looks like it could be a good compromise.
- Well maybe that is the trouble with robotics, it requires engineering ability. Scavenging for ready made parts is useless without the ability to learn how to do the mechanical engineering to make them all work together. If you look at and handle MWT's early bots (as I have) you will se that they are exquisite pieces of engineering without any of the 'oh we need a... - lathe, 3D-printed parts, carbon-fibre, titanium molybdenum alloy, you name any of the excuses put forward by roboticists' - being required. Cutters, pliers,a soldering iron and maybe a file plus manual dexterity and ingenuity were seemingly all he needed, even for the prototype of RoboSapien.
It is also important to know what went on in the dark worlds of mechanical and electrical engineering before everything was virtual through a computer screen. As has been said 'Those ignorant of the past are doomed to reinvent it (if they are clever enough)'. And if you have to spend your time reinventing the past you never have time to invent the future.
I've been building robots for forty years, and I have a bachelor's degree in robotics engineering. I remember the bad old days when just getting the wheels to stay on the shaft was the big issue in amateur robotics.
I met Mark at the 2nd Robot Olympics, I think it was, in 1993 (Sheesh, can it really have been that long ago? My calendar tells me so ...). I gave a talk on "ROBI" (RObot Builder Interface), an old-school standard I was flogging around North America at the time.
My overwhelming impression then was of the sheer craftsmanship of Mark's critters. I remember telling the guys back at the Seattle Robotics Society about it at the next meeting, how he wasn't so much making robots as making exquisite mobile jewellery.
(My other big impression was how ankle-breakingly dangerous the machine from Grove City College was, a machine in the walker category called "Wild Thing", if I remember right. I kept a respectful distance whenever it was on the loose. I think it got a medal.)
Seems to me there was a hexapod walker there, too, where the builders had tried to use solenoids to actuate the legs. Even though the team put a NiCd cell on the lower joint of each leg to reduce the sprung mass, the poor thing couldn't even stand under its own weight, much less walk. What that team really needed (aside from a better understanding of actuators), was a low gravity environment. So many things would be so much easier if it weren't for that pesky 9.81 m/sec**2 stuff.
Now I'm getting nostalgic. Good thing I don't have any beer on hand, or I might begin philosophizing or, worse, get all maudlin and start singing.