Belt(s) tensioning

Hi!

I read here that having both belts equally tensioned is mandatory. Looking at this figure:



it appears to me that we can use only one long belt:

- at first, the X-carriage is not attached to the belt; the belt is first tensioned, using motors translation or so;
- then, once this is done, the X-carriage is clamped on both strands of belt. This requires a system so both strands are clamped without introducing constraints (easy to do).

What do you think? Did someone already do that?

---

Frédéric

I've only built one DBot yet, but it seems both belts automatically have the same tension.
They are routed in two different heights : one belt 20mm above the other. I don't see a chance to use only one belt.

I think you could use one long belt loop, but it wouldn't guarantee equal tension on both halves of the loop. The belt would have to pass through the extruder carriage at one point (but be clamped to it there) and then be terminated at the carriage on the second pass. The belt still forms two independent loops, so you'd adjust tension for each by moving the motors.

I don't see any advantage to using a single piece of belt.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Yes, my proposition only works with the same plan configuration...

But I don't see how 2 independent belts can have the same tension, even on a DBolt...

---

Frédéric

The loops are independent. You adjust them to have the same tension.

Think about it this way- start with two equal length pieces of belt. The path that each belt loops makes is constant length (if it isn't you have a serious problem) no matter where the extruder is located in XY space. Now think about what happens when you move the motors up and down in your diagram. Move the motor down, the path gets longer (belt tightens), move it up it gets shorter (belt loosens).

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Sorry, I was answering to o_lampe...

dentist, as said, you first have to tension the belt without having it clamped to the X-carriage. Then, once it is tensioned, you clamp it on each side, in this exact position (it requires a simple printed part, clamped on the belt, which can slide on the X-carriage while tensioning, and clamped once done).

Now, I don't know how accurate should be the tension of the 2 belts; maybe it does not matter that much?

---

Frédéric

How I tensioned the belts:

I routed both belts through all the idlers, but one pulley wasn't installed.
I clamped down the belts on the carrier without much tension.

Then I added the missing pulley and both belts were tight.

Tightening the whole loop before clamping to the extruder carriage would pretty much guarantee equal tension, though with multiple pulleys I think it's hard to say that the tension is the same in every belt segment between pulleys. I'd probably move the X axis up and down the Y axis rails a few times to try to even everything out before clamping to the extruder carriage. Are you going to build it?

In my printer the motor mounts slide on the base plate. The screw holes in the base plate are slots. I just route the belt and clamp both ends at the extruder carriage, then pull the motor/mount until the belt is tight and tighten the screws that hold it to the base plate. That makes it easy to tweak the tension on either belt loop. I have noticed that the X axis tilts a bit when I pull on one motor to tension the belt. That means that XY orthogonality is going to depend on relative tension in the two belt loops. Being able to adjust tension in the individual loops independently, such as by moving the motor, will allow small corrections in XY orthogonality.

Edited 1 time(s). Last edit at 03/04/2017 11:38AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Yes, I will design the X carriage so I can try this design.

---

Frédéric

BTW, any idea of the force (magnitude) needed to have a correct tension on the belts? I guess it depends on the length; let's say 2m for each belt...

Thanks.

---

Frédéric

Tensile force is tensile force regardless of the belt length. I haven't seen any convenient device for measuring belt tension, and without specs on it from the manufacturer, the only thing you could do is match tensions. Maybe a small spring-scale could be used. You'd hook it on the belt and pull until you get a specific amount of deflection, then read the number on the scale. It would be a trial and error type device for making adjustments. You'd read it, then tighten or loosen the belt, then try again until you got it to match the other belt.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
Tensile force is tensile force regardless of the belt length.

Yes, but different lengths may require different tensile forces? Never looked at belt specs...

Quote

I haven't seen any convenient device for measuring belt tension, and without specs on it from the manufacturer, the only thing you could do is match tensions. Maybe a small spring-scale could be used. You'd hook it on the belt and pull until you get a specific amount of deflection, then read the number on the scale. It would be a trial and error type device for making adjustments. You'd read it, then tighten or loosen the belt, then try again until you got it to match the other belt.

My idea was to insert an auto-tensioning system. Again, first loose clamps from X-carriage, let belt tension, them clamp again. The system is very easy to do; it just requires the correct spring... Note that once clamped, the spring does not pull anymore on the belts.

Edited 1 time(s). Last edit at 03/05/2017 07:29AM by fma.

---

Frédéric

For equal tension all around one long XY belt on a single plane (i.e. belt needs standard twist/cross-over), you can use a single effector pulley (say 20T) on one side of the effector, with the belt teeth engaged around the pulley. Then when you tension this long belt, by any method, anywhere along its length, you get the same tension all around the belt. (Move the effector around to equalise slight tension variations caused by pulley frictions etc.) Then you can lock the rotation of the effector pulley to isolate the identical tensions in the two halves of the long belt.

If everything for the XY movement is already bolted square (which I think is best) then you are done. Otherwise, you could square it up by introducing a deliberate differential tension in the two halves by rotating the effector pulley slightly before locking it.

SydneyDesigner, I implemented your suggestion:



The 'V' on the left auto-blocks the belts, and the part where it is engaged is used to put tension on the belts. It allows a good tension within a small space. I printed a prototype, and it works great.

As I used a single plane config, I cut a part in a PTFE sheet to put between the belts where they cross, which avoids friction:



Here is how the PTFE is maintained:



Edited 2 time(s). Last edit at 09/16/2017 01:44PM by fma.

---

Frédéric

That doesn't look right. Do you have a picture of the whole mechanism?

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

What seems incorrect to you?

---

Frédéric

The belts are sliding on the teflon. What's keeping the teflon twisted?

Anything that pushes on the belt deflects the extruder carriage. I wouldn't put anything that isn't under very solid control in contact with the belts.

Edited 1 time(s). Last edit at 09/17/2017 08:20AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Not sure to understand what you mean... The PTFE is maintained in its middle by the purple hanger, and positioned at the exact blet crossing; and the belts themselves keep it twisted.

The PTFE is flexible enough even at the fixating point so it does not push the belts.

Even if the belts where deflected, as they are always in the same position, this should have no influence on the carriage? Am I missing something?

---

Frédéric

The belt tension will vary a bit each time the mechanism reverses direction. I would watch the teflon piece while the machine is operating- if you see any movement of the teflon, I think it's going to create print artifacts, specifically, small layer registration errors that will look like z wobble.

If the prints look OK, it's working fine.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

I see. I'll check that. Thanks.

---

Frédéric

maybe it just doesnt need to be that long?

You mean the PTFE? I cut it so the belts can't touch each other. If it was shorter, it would be the case.

But this is a very flexible sheet (1mm thick), so it follows the belts, and not the opposite. The only point where it can't move is inside the holding part.

I think it could be possible to maintain it in a better way, so it is totally free and does not act on the belts at all. I'm thinking to maintain it from the corners, with very small springs... Well, but I'm not sure it would make any difference on the printing results, as there are none

---

Frédéric

To get back to the innitial question; during operation the belt, one piece or two, will operate as two independent loops. But during tensioning that is not mandatory. You could make a system with one long belt, both ends and the middle of the belt running through the extruder carriage. Tension the lot without the middle of the belt clamped to the carriage, and clamp after you tensioned. This way both loops are under the same tension.

Cheers,

Hugo

So, you advice to tighten the idler wheel after tensioning? Or can I let it free to turn?

---

Frédéric