Arduino, and Mechanical Movement of Objects - a Newbie Question

I'm working on a prototype for a machine that splits the plies of commercial yarns. These yarns are reclaimed from hand-me-down, and thrift shop sweaters. A model of this machine can be seen here http://www.youtube.com/watch?v=AKRCNBwv4Pk, but I've made a fairly important change since the video was made. See this image.

The rod and sliding "hooks" replaced the screw eyes so as to distribute the yarn more evenly on the drum. When the yarn is even distributed, the operation works smoothly, and it didn't work well with the stationary screw eyes. At this point, I've decided that I want that distribution to happen automatically. The sliding hooks work as I hoped except they're a bit fiddly.

I am a newborn to both electronics, and Arduino. My question revolves around a joint mechanical / Arduino approach. I could use a little 12v. DC motor, and a reducing worm gear to turn a threaded rod on which these sliding hooks could ride. Alternatively, the worm gear might operate a rack directly under the rod, and the rack could be fixed to the sliding hooks. I would need for the Arduino program to reverse the rotation of the motor at specific intervals. Sometimes, 2 hooks will be used, so the motor controller would need to turn in one direction for a longer time than if 3 or 4 hooks would be used. What kind of sensor might make sense to help me accomplish that? Is this something I could accomplish, or would I be wiser to pay someone to do the eagle PCB layout, and Arduino programming?

If I've hastened through a point that needs clarification, please feel free to ask. I'll try to find a way to make it clearer.

Peace,

Dave

It's very unclear how this works from the 3D animation. Where are the hooks you are talking about? If the movement itself is basically repetative, I would stick to mechanical means and just use an oscillating mechanism to move the hooks.

Or am I not understanding? Again, the most important bit is how the hooks work and I don't see any hooks in the video.

Right. Very sorry about the confusion. The vid was "pre-hook"...there were no hooks yet when that was made.

I've made a fairly important change since the video was made. See this (image) The rod and sliding "hooks" replaced the screw eyes so as to distribute the yarn more evenly on the drum.

On most spinning wheels, there is a gadget called a flyer. The flyer has two arms, one on either side of the bobbin. Traditionally, these arms have had screw hooks, or cup hooks fairly close together. Without the hooks, the newly spun yarn would pile up in one place, rather than evenly distributed throughout the whole of the bobbin. In the last 5 years or so, small, modified spring-like devices called sliding hooks began replacing the traditional hooks because they are faster, and distribute the yarn more evenly on the bobbin.

In the image, you'll see a metal rod with 4 of these sliding hooks attached. Neither the rod, nor the sliding hooks were in the vid. They replace the row of static screw eyes, and that's a marked improvement. However, every time I need to stop to slide the hooks, I'm not cranking. The cranking mechanism needs to remain a manual operation because one's hand can feel when something needs immediate attention.

In the video, for example, there is a yarn ball hanging from an inverted U at the top of which is a sealed bearing which allows it to rotate easily so as to keep the plied yarn's twist from accumulating. The yarn travels from there to a screw eye, and then is split into it's 4 plies by the other 4 screw eyes. From there, it is wound around the drum. Plies were breaking with this method, however, because the yarn didn't wind evenly around the drum. In some places, it was thicker than others, and where it was thicker, more yarn was needed to wrap around the total circumferance at each revolution of the drum. Sooner rather than later, that ply would break. That doesn't happen with the sliding hooks.

Yes; I have thought about a number of different mechanical mechanisms to move those sliding hooks back and forth. Little levers from model railroading which are used to switch tracks could also be used to switch the rotation of the motor. Place one at each end of the rod, and you have an endless looping of rotational changes of the motor. I haven't considered an oscillating mechanism, but it sounds intriguing. I'd like to hear more of your thinking with that.

There's something I find very satisfying, and rewarding about simple, basic mechanical devices. Cars, and household appliances like sewing machines worked beautifully before printed circuit boards. Quite a number of them are still in service. Jane & John Homeowner didn't need a degree in electrical engineering to service them, either. That said, I'm looking for practical applications to learn Arduino, and about electronics in general. This contraption seems like one of those practical applications.

If you can do the mechanics, the electrical part is easy enough. You use a DC motor driver shield so you can forward and reverse the rotation. Maybe have a home micro switch so the rotation won't drift (say left rotation always gets a bit more than right rotation and that difference adds up.) Totally doable once you are done building the mechanical parts.

When you say sliding hooks, you mean from side to side or they slide around the axle? Do they need to do this independently of each other? What exactly is the determination of when they need to slide and how much (the tension?)

I'm sure you understand exactly how this works, but to someone unfamiliar with the process you are going to have to explain it like you are talking to a toddler. lol

Once we understand what exactly you are attempting, we can help. I need to know 3 main things:

  1. What determines when a hook needs to slide, and what do they do when the determination is made?
  2. Are the hooks expected to move independently of each other?
  3. Are we sliding side to side along the shaft or rotating around the shaft?

Depending on the scale of this device, you might well find that you can simply rip the drive mechanism and motor out of an inkjet printer and use it as-is, using an Arduino to replace the part of the printer that controlled it.

  1. What determines when a hook needs to slide, and what do they do when the determination is made?
  2. Are the hooks expected to move independently of each other?
  3. Are we sliding side to side along the shaft or rotating around the shaft?
  1. Ideally, they would slide very slowly but continually back and forth as the crank is turned. "As the crank is turned" suggests a purely mechanical device might be better suited. I mean, I might need to answer the phone, pour more coffee, take a leak, and forget to turn off that motor. When I move them manually, it's in response to certain visual cues; the yarn is getting deep in a particular slice of the drum, and the "V" shape of the split yarn starts to move towards the place where it's mounding more deeply.

  2. As it is right now, yes, they move independently. Whether they are moved by purely mechanical means, or with the assistance of electronics, they would slide back and forth in unison. I anticipate a nearly perfect distribution of the yarn around the drum, and zero breakage.

  3. They slide side to side.

rip the drive mechanism and motor out of an inkjet printer and use it as-is, using an Arduino to replace the part of the printer that controlled it.

I like that idea; thanks. I've also considered using the worm gear from a mechanical sewing machine, and also checking out the mechanism from a typewriter, either of the sort in which the entire carriage moves, or the sort in which the type head moves back and forth. The question I have about the typewriters is that they are designed to operate in one direction, from left to right. That's solved with the printer, though, since it operates in both directions, and I think I still have one around here somewhere.

Dave_Burrows:

  1. What determines when a hook needs to slide, and what do they do when the determination is made?
  2. Are the hooks expected to move independently of each other?
  3. Are we sliding side to side along the shaft or rotating around the shaft?
  1. Ideally, they would slide very slowly but continually back and forth as the crank is turned. "As the crank is turned" suggests a purely mechanical device might be better suited. I mean, I might need to answer the phone, pour more coffee, take a leak, and forget to turn off that motor. When I move them manually, it's in response to certain visual cues; the yarn is getting deep in a particular slice of the drum, and the "V" shape of the split yarn starts to move towards the place where it's mounding more deeply.

  2. As it is right now, yes, they move independently. Whether they are moved by purely mechanical means, or with the assistance of electronics, they would slide back and forth in unison. I anticipate a nearly perfect distribution of the yarn around the drum, and zero breakage.

  3. They slide side to side.

Ahhh! Now I understand. You want them to slide back and forth so that the yarn does not all accumulate in one spot on the spool. Duh! I should have realized that. Like a wire spooler or coil-winding machine.

Essentially, you want them to direct the yarn from one edge to the other creating a uniform layer. Actually, this makes mechanical means even more relevant. As you change the speed of the drum, you want to also increase the speed of the side to side movement. However, this will be in a ratio.

For every revolution of the drum, you want the yarn to move over one yarn width. When it reaches the end, you want it to do the same, but in reverse. This will distribute the yarn uniformly along the spool/drum. If the speed of the slides do not match the speed of the rotating spool, it will not distribute evenly. So you want these to work in unison.

So you need to determine the number of revolutions it takes to spool the yarn across one entire length. Controlling the tension will thin the yarn fibers out to make them a universal width. You want there to be tension so the yarn does not slip side to side and so it will be a constant width. A coil-winder uses a solenoid to provide this tension by rotating the axle that the eyelets are on. Too much tension, and it will break of course.

Once you determine how many revolutions it takes to spool an entire layer, you count the number of spirals. This will determine your gearing ratio.

But then, as the spool gets thicker, it will take more revolutions to wind the yarn, correct?

Yes, you do have it now. Sorry it took so long to explain it coherently. There are several variables, and you named one of them; that as the yarn gets thicker on the drum, it takes more yarn to make a single revolution around the drum. However, unless the thickness of the yarn increases or decreases, precisely the same number of revolutions will be required to move from point A to point B on layer 1 as it will take to move from point B to point A on layer 8.

Another variable is the thickness of a single ply. In GB, AU, NZ, and most if not all of Europe for that matter, ply has a specific meaning. So, the yarn we call "baby yarn", or "sock yarn" here in the states, is called "4 ply" in much of the rest of the world. We call it "worsted weight", most everyone else calls it "8 ply". Same word, same topic, but very different meaning. To further complicate things, there is no exacting standard. One mill might produce a worsted weight yarn that another mill down river a mile calls DK weight.

Folks in the US have invented a new wheel. It's called wraps per inch, or WPI. Ostensibly, it dares to set a standard. By wrapping a yarn around a ruler, and counting the wraps between any two inch marks, worshipers of the WPI standard can tell you in no uncertain terms what the weight of a yarn is. Unless they're pissed, wrap a little more tightly than usual, and cram an extra couple of wraps in there, or unless the individual who did the wraps was your demure gramma who wraps considerably more loosely than the average fiber artist because she knows not to ever wind yarn too tightly lest you stretch the life out of it.

Because precision on that level is impossible, I wouldn't mind if there was a half a yarn width gap between revolutions, or if there were an overlap of half a yarn width or more. Consistency is the goal, and it's also a very good reason to mechanize that aspect. You hit on a very important point when you said,

As you change the speed of the drum, you want to also increase the speed of the side to side movement.

Exactly right. And since the crank that turns the drum needs to be done by hand to respond immediately when something is amiss, creating an electronic component may possibly be counter productive. I wish it were as simple as adding 2 more pulleys and a belt connected to the existing power train, but I don't think it is because of the direction change that is required.

Anyway, thanks; good thinking!

Bah! Direction change is no big deal. Ever look inside an old cassette deck or VCR? The motor always moves in just one direction for forward and reverse. You simply pop a gear over based on a clutch. A center driven gear spins in one direction. You connect a gear to the left side and it rotates in one direction, you move that gear away and mesh a gear on the other side and it rotates in the opposite direction. The cam that decides which gear is meshed is just controlled by the ends of the movement (when it reaches an end, it pushes a lever which disengages one gear and meshes the other.) You don't think a car engine actually rotates in the opposite direction when you put it in reverse, do you? ;)

I will look for a picture if that doesn't make sense. I can't draw worth a crap. But it is incredibly simple.

Well, first, I tore apart a laser printer, retained all the screws, PCBs, 2 motors, a bunch of gears, cams, rods, rollers, wires, springs, switches, and bits that moved that I don't know how to name. Next, I re-read these last two posts of yours. I brought the ritual to completion by wiggling my finger in front of my lips while humming softly. "Two more printers to go", I said, and then continued my autistic mannerisms by staring at nearby lights.

Several posts up, you said:

I'm sure you understand exactly how this works, but to someone unfamiliar with the process you are going to have to explain it like you are talking to a toddler. lol

I get the idea that you know how to do what I want to do, but after numerous Google searches, I'm not finding anything to help me catch up to you. It does make sense, just not enough to jump start the old engine. A photo, or drawing, or diagram would be great.

ok. I found a few good resources for this, but none of them make it as simple as you need it to be. So this is my attempt at drawing it:

http://www.flickr.com/photos/78906572@N05/8449613492/ (apparently I don't know how to embed an image)

Bascially, a gear connected to the driving gear will rotate in the opposite direction. So for one direction, you connect the driving gear directly to the output gear. To reverse direction, we move an idler gear in between the driven gear and the output gear. You do this with a simple pivot mechanism that just rotates enough to pop the idler gear in place. This lever is toggled by reaching the end points of your sliding mechanism.

So, by default, your output gear would be driving in the opposite direction of the driven gear. The idler gear makes the output gear travel in the same direction, reversing it.

That makes sense for a simple mechanical solution, but given the presence of the microcontroller and quite likely a stepper motor I wouldn't have thought that a reverse gearing mechanism would be required.

PeterH: That makes sense for a simple mechanical solution, but given the presence of the microcontroller and quite likely a stepper motor I wouldn't have thought that a reverse gearing mechanism would be required.

I am describing a purely mechanical solution which ensures that as he turns the spool at a variable speed, the sliding mechanism will stay in sync providing a uniform layer. He already has power from the crank which he mentioned he would leave to manual power anyway. Of course, if he is going to use the microcontroller, the reversing gear wouldn't be needed.

Hey, man; thanks for having taken the time to work that out on paper for me. I think I understand the concept, or at least I thought I did. In the gizmo I'm building, there are 2 identical gears each engaged with the other. The output gear turns clockwise which means the driven gear turns counterclockwise. The driven gear is fixed to the large pulley, and via a belt, turns the small pulley. As desired, both pulleys turn counterclockwise.

In your drawing, there is a point that confuses me. Your output gear rotates counterclockwise. As drawn, it is engaged with pulley A which turns clockwise. I'm with you to this point. Then pulley A the driven gear is engaged with pulley B, the idler gear, which appears in your drawing to also be turning clockwise...and visa versa when the lever is toggled, and the idler gear engages directly with the output gear. I guess what I'm unclear about is that before the toggle, both A & B are turning clockwise, and after the toggle, both A & B are still turning clockwise.

That said, I think I found enough in the way of gears, and shafts to attempt to make something work. I'll stick close by so you can help me understand better what confuses me now.

I missed the second image at first. Is it some kind torque gaining, speed reducing doo-dad?

It looks like the upper larger gear is the driven gear, and the lower two are speed reducers? Or do I have that bass ackwards?

That second picture is another project I am working on, sorry....

Did I mess up my drawing? lol

Idler gear A is always connected to driven gear, but in one direction, it just "idles" off to the side not touching anything. When you want to reverse direction, you pivot the idler gear into connection with the output gear C and the output gear will now rotate in the same direction as the driven gear.

Yep. I did mess up my drawing. The idler gear would would turn in the opposite direction of the driven gear, always. I just drew it wrong.

Here is a tutorial: http://www.societyofrobots.com/mechanics_gears.shtml#rotdirection

But the take away is that for each gear in a chain of gears, each one will move in the opposite direction of the one driving it. By introducing a gear in between the driven gear and output gear, the third one will move in the same direction as the first.