PDA

View Full Version : What happens if you try to turn a stopped stepper?



toozie21
03-07-2014, 09:22 AM
I've been thinking about how to get a home position for a stepper motor in my animated nativity project. The most straightforward thing to do would be to have something like a micro-switch and detect when I reached a certain position. But that involves more code and more places for things to go wrong.

What I was curious about was what would happen if I turned the stepper into an obstruction (say a rod runs into something that keeps the stepper from turning anymore), would I have anyway of knowing? I wasn't sure if there was a bi-product of turning the stepper that can't turn that I could use as an indicator that I hit a stopping point (assuming it isn't damaging the motor).

Would it hurt the stepper to try to turn while against an obstruction? If the answer to that is no, then I could just have a timer, turn the stepper in one direction, and if I ran it until I know it would go from the furthest away from the stop to the stop, it wouldn't matter if it was at any point in between, I would know exactly where it was when the timer stopped (at the stop/home position).

Or, am I barking up the wrong tree and need to stick with a micro-switch or something of that ilk?

callen360
03-07-2014, 10:41 AM
I've been thinking about how to get a home position for a stepper motor in my animated nativity project. The most straightforward thing to do would be to have something like a micro-switch and detect when I reached a certain position. But that involves more code and more places for things to go wrong.

What I was curious about was what would happen if I turned the stepper into an obstruction (say a rod runs into something that keeps the stepper from turning anymore), would I have anyway of knowing? I wasn't sure if there was a bi-product of turning the stepper that can't turn that I could use as an indicator that I hit a stopping point (assuming it isn't damaging the motor).

Would it hurt the stepper to try to turn while against an obstruction? If the answer to that is no, then I could just have a timer, turn the stepper in one direction, and if I ran it until I know it would go from the furthest away from the stop to the stop, it wouldn't matter if it was at any point in between, I would know exactly where it was when the timer stopped (at the stop/home position).

Or, am I barking up the wrong tree and need to stick with a micro-switch or something of that ilk?

If it is anything like our motors on our CNC, it will fight and fight till it burns up if it hits an obstruction. We have SEVERAL limit switches and a magnetic cutoff for if it gets past that. The limit swith on mine just does a soft stop, the mag switch does a ESt

ErnieHorning
03-07-2014, 02:47 PM
It depends on if your assembly is stronger than the torque of your stepper. It’s actually a pretty common practice to do this. You always have to start out by stepping home the maximum number of steps minus one. You just have to accept the loud buzzing. Also make sure that your home position is aligned on a stopped step and not a partial step. Otherwise you might be at home or off a step.

It’s quicker to start if you can home to a switch though so that you don’t have to pound away on the stop when you’re only a couple of steps away.

Anybody that ever used an Apple computer with a floppy disk drive is familiar with the homing clatter of the floppy drive when power was first applied.

toozie21
03-07-2014, 03:40 PM
Hmmmm, does anyone have a source that sells inexpensive steppers with encoders built in (or something for home detection)? Maybe I need to switch gears a bit....

LabRat
03-07-2014, 03:51 PM
Anybody that ever used an Apple computer with a floppy disk drive is familiar with the homing clatter of the floppy drive when power was first applied.

I was going to say +1... but I think ][+ would be more appropriate. :thup2:

RobG
03-07-2014, 05:39 PM
Some driver ICs (DRV8825 for example) have current limiters which will protect both, the motor and the driver.
You could also include feedback circuit, so you know you ran into obstruction.

ErnieHorning
03-07-2014, 06:15 PM
When I used to use steppers to turn precision potentiometers for thermal simulation, I would just step them much faster so that they hardly had any torque. It got the job done much faster and there wasn’t enough torque to do any damage. I found it 100% reliable if I went 90% and then slightly overstepped the last 10% at a slower speed. No need for switches, encoders or current limiting.

kingofkya
03-07-2014, 06:18 PM
Hmmmm, does anyone have a source that sells inexpensive steppers with encoders built in (or something for home detection)? Maybe I need to switch gears a bit....

In the free or almost free price range.
Any pc with a floppy:)
OR MFG printers have a ton.

LightUp
03-08-2014, 02:51 PM
My understanding is that encoder data is volatile, meaning that if the power turned off you lose the absolute position. The CNC machine I was involved with (Excellon) used encoders but also used the HOME position of the X & Y axis as a software calibration point to know where it is physically. I would suggest you use a similar approach.

If you don't use this method then you will periodically need to manually home the machine to re-calibrate the physical co-ordinates in your software as a reference point. What if someone manually turned your motor with the power off? How would your system know it was turned? Most, if not all, encoders are relative in position accuracy, but not absolute in position.

ErnieHorning
03-08-2014, 03:52 PM
You could save the encoder value in EE memory and compare the two when you start up. Not 100% guarantee but the likelyhood of an exact match would be somewhat low.

You could also home and watch that the encoder changed each step and 99% certain that you're actually home. Switches are cheap though.

Sent from my DROID RAZR using Tapatalk

toozie21
03-10-2014, 07:59 AM
When I used to use steppers to turn precision potentiometers for thermal simulation, I would just step them much faster so that they hardly had any torque. It got the job done much faster and there wasn’t enough torque to do any damage. I found it 100% reliable if I went 90% and then slightly overstepped the last 10% at a slower speed. No need for switches, encoders or current limiting.

Not a bad idea, that would save on the motor and probably still accomplish what I need. I wasn't thinking that I could control the speed.


My understanding is that encoder data is volatile, meaning that if the power turned off you lose the absolute position. The CNC machine I was involved with (Excellon) used encoders but also used the HOME position of the X & Y axis as a software calibration point to know where it is physically. I would suggest you use a similar approach.

If you don't use this method then you will periodically need to manually home the machine to re-calibrate the physical co-ordinates in your software as a reference point. What if someone manually turned your motor with the power off? How would your system know it was turned? Most, if not all, encoders are relative in position accuracy, but not absolute in position.

What I was thinking of, was something like this:http://www.usdigital.com/assets/datasheets/E5_datasheet.pdf?k=635297895452977206. The index position is at a set location every time, so it could be used to find home.

LightUp
03-10-2014, 08:39 AM
...What I was thinking of, was something like this:http://www.usdigital.com/assets/datasheets/E5_datasheet.pdf?k=635297895452977206. The index position is at a set location every time, so it could be used to find home.

I suppose that depends on your hardware design. If your prop uses only one revolution of the motor shaft then the index pulse may be configured for a given physical position. Often motor projects will use many revolutions, hence many index pulses, before resting at the HOME position. If so, you will need to count the number of index pulses before coming to the HOME position, and thus need to store this count for reference of the axial position.
It will be interesting to see how you integrate your sketch with your hardware.

ErnieHorning
03-10-2014, 09:42 AM
If your prop uses only one revolution of the motor shaft then the index pulse may be configured for a given physical position.Nothing say's this has to go on the motor shaft. I could be attached to anything that makes one revolution. It could be a gear reduction solely for position sensing.

A micro switch for home and end position is still the easiest. These could even be in parallel so you only use one input. One step would tell you which one you’re at if you just powered up.

LightUp
03-10-2014, 09:51 AM
Nothing say's this has to go on the motor shaft. I could be attached to anything that makes one revolution. It could be a gear reduction solely for position sensing.

A micro switch for home and end position is still the easiest. These could even be in parallel so you only use one input. One step would tell you which one you’re at if you just powered up.

I realise that. I should have remove the word "motor". :)

Edit: Ernie your private maibox is full. I was going to ask you for clarification on your "reduced" torque technique while still retaining STEP count accuracy without missing pulses. Or are you relying on an encoder for that?

toozie21
03-10-2014, 09:58 AM
I appreciate everyone's helps. As I muddle through the issues and options, I think I might need to rethink it all from the beginning anyway.

What I was planning on doing was have large (2-3ft tall) "pages" of a book made out of coro (framed in some wood for stability). As my blowmold nativity lit up to tell the story, the pages would turn to have some text on it as well. When everything goes dark to reset the story, the book would flip back to the beginning to start over.

What I've been trying to do is figure out how to do the book portion without over complicating the process. Sadly I cannot come up with a slick solution that doesn't involve multiple moving pieces and animations (mostly the resetting of the book is what throws things off). Probably the easiest thing to do would be to have a stepper that rotates a movable arm (a linear actuator, rack/pinion, another servo, etc) that moves to a page, extends the arm, catches the page, turns the page, retracts the arm, and repeats. At the end the arm gets extended and the motor runs in reverse to put all the pages back in one fell swoop. What I worry about with this approach is the torque on the stepper, as well as the amount of sensors needed to accomplish the goal.

Simpler solutions could be had with spring loaded "catches" on a fixed arm, using the park feature on a parking motor, etc. But I cannot seem to wrap my head around the turning of individual pages, and then the resetting of the whole book (which doesn't necessarily need to be done in one fell swoop).

If this rambling background of my issue helps spur a different track of thinking in you guys, I am all ears!

Zeph
03-10-2014, 12:28 PM
How many pages do you have in mind flipping?

Leaving an arm that flips pages aside at the moment, how are the pages hinged? I mean how do you allow them to stack with page 1 behind page 2 (from the viewer's perspective) at the start, then page 2 behind page 1 at the end? It can be fairly tricky if you wanted stiff to lay flat, but you could move each page from say 30 degrees to 150 degrees (where 90 degrees points at the viewer), if the spacing of hinges arranged linearly perpendicular to the view was twice the page thickness (including wood bracing).

The mechanical mounting of your pages can affect the best solution. For example, if there are just a few pages, a relatively inexpensive geared hobby motor aligned on the hinge axis of each page might be the easiest - give it a timed burst of power whichhas been calibrated to be enough to move the page to the other position plus an error margin.

toozie21
03-10-2014, 12:49 PM
Well, I was imagining turning 5 pages (the cover and 4 pages). There is a fifth page, but it would be the end of the book.

I was imagining the pages would be hinged by having a rod through the page vertically and would extend out the top/bottom an inch or two. That extended piece could set into holes in a piece of wood and would sort of pivot on that. The pages would then sit one next to the other (sort of like a kid's board book). I was imagining that the book would not be opening 180 degrees (though I am open to whatever would work), so the 30deg to 150deg would be fine. And I was imagining that the book would be tilted back some (at say a 15deg angle) so that I could use gravity to help the pages fall into place when turning if need be.

Nothing is built, I've just been mentally spit-balling everything before getting started on it to try to work out all the kinks and catch potential pitfalls earlier. So everything is negotiable.

I'm not sure if that helped clear anything up, but I would be interested in your thoughts. I am trying to picture your geared hobby motor idea, but I think I might need a little more of a description of that.

LightUp
03-10-2014, 01:49 PM
I gave it some thought and could only come up with this method of hinging. 23470
Each page needs to lower so that the next page can lie on top of it. I've used a 1/8" pin, or screw, fitted in the slot. Obviously, some clearance is need to rotate properly. Each book page is 1/4" thick to allow a wooden frame of some sort.
Like Zeph, I'm tackling one problem at a time and then move on to the next.

ErnieHorning
03-10-2014, 01:59 PM
Here’s the vision that I see; all measurements are hypothetical and should be changed for best appearance:

Imagine a single page. Now if you cut a notch an inch into the page at the top, from the hinge side to half way. Now with a rotating lever with a center that’s offset to the right, you lift the page and at a point slightly more than half a page turn, the lever falls through the open notch and the page free falls. Do this for each page.

You should be able to reset all pages to the beginning by doing the same thing but more toward the center of the book instead and your only interested in the last page.

The example said an inch but you should be able to make it smaller depending on the distance from the viewer.

toozie21
03-10-2014, 02:17 PM
I gave it some thought and could only come up with this method of hinging. 23470
Each page needs to lower so that the next page can lie on top of it. I've used a 1/8" pin, or screw, fitted in the slot. Obviously, some clearance is need to rotate properly. Each book page is 1/4" thick to allow a wooden frame of some sort.
Like Zeph, I'm tackling one problem at a time and then move on to the next.

Yeah, I think that that is what I was thinking (or pretty close), pictures are a lot easier to imagine than words :). Wood could also be replaced with 1/4" (1/8") steel rods if needed as well (could be welded and would be thinner and less conspicuous).


Here’s the vision that I see; all measurements are hypothetical and should be changed for best appearance:

Imagine a single page. Now if you cut a notch an inch into the page at the top, from the hinge side to half way. Now with a rotating lever with a center that’s offset to the right, you lift the page and at a point slightly more than half a page turn, the lever falls through the open notch and the page free falls. Do this for each page.

You should be able to reset all pages to the beginning by doing the same thing but more toward the center of the book instead and your only interested in the last page.

The example said an inch but you should be able to make it smaller depending on the distance from the viewer.

OK, I think I am understanding what you are saying, but am missing something. The un-notched section allows the lever to pull on the page, then when the rotation of the page has turned enough, the lever will hit the notched section, and just pull through (meanwhile the page will continue falling on its own), that about right? What I don't see is how the second page wouldn't have gotten in the way of grabbing the first page. I am guessing the point I am missing has something to do with the significance of starting the notch an inch from the hinge and that being adjustable.

kingofkya
03-10-2014, 04:11 PM
Keep it simple guys:)


What about a small string with knots tied in it. the --- are string and the ++ are knots and = is a page. Note the last page is tied on both sides this lets it reset the entire stack.
-+=-------------------+=-------------------+=-------------------+=---------------+-----

Then just a small dc moters will pulleys on them. string is red. You could do it with one moter but you need to keep tension in the other side somehow. So it can unwind. I am assuming your pages are ridged.

23474

ugh nvm physics are in the way. It would need some way to un hook the page on the other side. Or maybe multiple pulleys for each page?

LightUp
03-10-2014, 04:25 PM
... And I was imagining that the book would be tilted back some (at say a 15deg angle) so that I could use gravity to help the pages fall into place when turning if need be. ...

Did you also indicate that there may be some text on the page for the audience to read? If so, would it not require a greater angle than 15 deg to read properly?

toozie21
03-10-2014, 05:13 PM
I had thought of the string approach as well and came up with the same issue as well.

The 15 degrees was tilting the book back so it wouldn't be straight vertical.

Sent from my DROID BIONIC using Tapatalk

kingofkya
03-10-2014, 05:32 PM
I think you could do it with 1 string per page but you would need some kind of clutch on each wheel. Or multiple moters.
There must be a simple mechanical way...

I think i got it if you cut 2 holes like this, then get some bouncy balls or something simmaler maybe a round bead. Then when the page is on the right the round thing is in the lower side that it cant fit in and once the page gets near the end the round thing then moves to the top and pops out the other side. You woudl have to place the pulley on the right above a bit higer then the other to do this.
23478

kingofkya
03-10-2014, 06:00 PM
OR going of the same though process a belt maybe made from http://www.uline.com/cls_09/Strapping
then have a pieces glued to it that just looks like and upside down T so it could pull though a slot in the page when it gets to the other side.

You would probably need to have some separation between pages so on reset the pages end up with the T tabs on the right side(hide in the box below maybe)

23480


Then one more this is the pages would have to have an extra tab so you could use it as a stop when the pages go below. Like a file folder. Each tab would be in a different spot so it could stop at the bottom.

http://www.staples.com/Staples-Manila-File-Folders-Letter-3-Tab-Assorted-Position-250-Box/product_221689?cid=PS:GooglePLAs:221689&srccode=cii_17588969&cpncode=33-269674710-2

Roterhead
03-10-2014, 09:30 PM
Here's a different take on page turning..

Consider the use of Model Servos. One servo per page, so 4 servos needed. Adjust size of servo to handle torque needed to turn the pages. Page turning speed and slow / fast can be easily adjusted.

Mechanical: Use spring steel wire (very high strength, bends but will spring back, used a lot in Model aircraft) arranged in a brass tube acting as a bearing.

Take a look at the attachment and see what you think....

toozie21
03-10-2014, 10:44 PM
I had a potential idea this evening, but I don't know how well I will be able to articulate it in words, so bear with me (and I definitely don't have the drawing skills to show you).

My thought starts with each page having the tabs sticking out of the bottom like kingofkya's manila envelope idea. These tabs will be at different points on different pages (say the first page has the tab 6" from the hinge, second page is 5", third page is 4", etc.). Next I have a disc underneath with pins sticking up vertically from it. As the disc turns, the pin would catch the tab sticking out the bottom of the page and rotate it (to turn the page). If I do the math right, the disc would never need to rotate a full 360 degrees, so the pins would be spaced around the disc. With me so far?

This is something that I've considered in the past, but what I thought of that would be different this time is to not have the disc parallel to the bottom of the book, but to have a downward incline so that the part of the disc that is closest to where the pages will land when opening the book is closer to the ground (sloped downward). By doing this, the tabs will be out of the way as the disc turns around, yet will also not be in the way as I reverse the disc and reset the book (also where my previous plans failed).

What I would need is one special pin that when the book is resetting, it catches the first page (even though there is an incline, so it would be taller) and pushes it (which pushes all the other pages as well) back to the starting position. If I put a microswitch to detect when the special reset pin is in a position that all the pages must be at the beginning, I will have a "home" position for the stepper, as well as a way to know that I need to stop the motor and not break anything.

OK, now is where you can poke holes in my idea. Seems solid to me, but so did the previous 108 ideas I had for this project...... One of my big concerns with this idea is, would a stepper have enough torque to handle this, in particular, the turning of the pages back to the beginning (since it has to turn all the pages at once).

LightUp
03-11-2014, 08:25 AM
.... One of my big concerns with this idea is, would a stepper have enough torque to handle this, in particular, the turning of the pages back to the beginning (since it has to turn all the pages at once).

Once you build a prototype you then can pull the pages using a "fishing scale" to approximate how many oz. pull it takes to turns all the pages. You then compute a rough torque number. If your stepper is too low you could use gears, or pulleys, to increase the torque from the motor.

toozie21
03-11-2014, 08:29 AM
Once you build a prototype you then can pull the pages using a "fishing scale" to approximate how many oz. pull it takes to turns all the pages. You then compute a rough torque number. If your stepper is too low you could use gears, or pulleys, to increase the torque from the motor.

Interesting idea on the fishing scale. A low-tech way to get an answer, something I would have NEVER thought up on my own!! I believe would would need to be done is take the measurement of the rotation point from the axis in inches, and multiply that by the value on the fishing scale (in oz), this will give me a oz-in measurement that I can use to compare to the ones on the stepper (like a 70oz-in stepper). Of course, any wind on it when trying to turn will add to that value, so I would probably need to pad my overhead quit a bit.

The one catch I've been able to come up with since my idea last night was that I might have an issue with the disc since it is a radial motion around a central access, yet the pages all have different axis points. By picking the center axis, it might be close enough (as long as I leave some fudge factor with the tabs), but I will probably have to play with it to figure it out.

LightUp
03-11-2014, 08:39 AM
Interesting idea on the fishing scale. A low-tech way to get an answer, something I would have NEVER thought up on my own!!

Neither would I, except that when I did some work on reel-to-reel tape recorders, many years ago, we needed to make tension adjustments on some parts and we had special little "fishing scales" to do the job. Hence the idea stuck. :)