Spectra line modification.

[daftscience]

I've bought some spectra line and have printed some pulleys and tensioners to attempt to convert to a spectra line drive system. From what I've read, the big issue is that tension will changes as the line walks the pulley.

Here is a detailed technical drawing.
[img]http://i.imgur.com/fBzIpX7.png[/img]

The tension changes is because of the angle of the line as it moves from the left to the right. (Polygonhell explains it much better in his blog.)

technical drawing #2
[img]http://i.imgur.com/ceWItge.png?2[/img]

Could it be as simple as making the pulley cigar shaped? My guess is no. Trig, Calc, Geometry, they're all a distant memory.

Snooping on polygonhell's blog I found this video:
[youtube]http://www.youtube.com/watch?v=23bBKTQygBQ[/youtube]

Since we already have pulleys by the motors it wouldn't take much to throw that on the max. Could this work?


Tom

[geneb]

Probably, but there's no reason to.

g.

[daftscience]

Probably, but there's no reason to.

g.

It seems like it would eliminate the issues of leveling and tensioning the belt. Honestly, it could be that I'm doing something wrong, but I have spent hours trying to get my belts to stop shedding. I think the fishing line could be a nice cheap alternative, especially if you need to replace the belts.

Tom

[Jimustanguitar]

It seems like it would eliminate the issues of leveling and tensioning the belt. Honestly, it could be that I'm doing something wrong, but I have spent hours trying to get my belts to stop shedding. I think the fishing line could be a nice cheap alternative, especially if you need to replace the belts.

What leveling trouble are you having that's belt related? I don't understand what problem you're trying to fix.

Tensioning is tensioning, there are lots of homebrew gadgets to do this, but I don't think you'll do away with it all together. Plus, the belts have several fibers through them, and they're very stretchproof. I think you'd actually be better sticking with the belts if you're having a stretching issue.

The advantage I see to using line instead of belts would have to do with rotating mass and jerk accelleration. I'm not sure that that's lacking with Nema 17's though.

[daftscience]

It seems like it would eliminate the issues of leveling and tensioning the belt. Honestly, it could be that I'm doing something wrong, but I have spent hours trying to get my belts to stop shedding. I think the fishing line could be a nice cheap alternative, especially if you need to replace the belts.

What leveling trouble are you having that's belt related? I don't understand what problem you're trying to fix.

It's mostly getting the belt to stay level on the idler with out it moving to the side. This is also an issue on the bottom as well. I guess tension isn't my issue. But basically, I can spend a few hours get it all calibrated and it doesn't take long before I start noticing belt shavings under the motor.

[Flateric]

I have seen the "walking" issue resolved in professional machines before by using a pulley that is free to move on a bearing in a linear form towards and away from the stepper as it is wound. The shaft of the drive is a spline much like a transmission. It cannot move or have any slop rotationally but the line winding is enough force to creep the sleave back and forth maintaining a exact position relative to the rotational driving motor/gear/etc.

................................Shaft rotation and wire windage
....................................................\/
Wire--------------------------------[--]
.......<--Desired linearmotion->>........[--]
...................................................[--] /\
...................................................[--] Sleeve on keyed shaft free to slide up and down
...................................................[--] \/
...................................................[--]
...................................................[--]--------------------------------------wire return
...................................................[...]
..............................................____||_____
..............................................|Stepper or|
..............................................|any drive |
..............................................|source etc|
..............................................___________

I hope that helps illustrate what I mean.
(And if not too bad, LOL. While type "drawing" this I was reminded of how much I hated doing it way WAY back when I was a kid using BBS services that only allowed for this type of illustration to communicate ideas)

So it will without doubt be the last time I ever attempt an ascii drawing again for the rest of my lifetime! HA!

[Flateric]

The main reasons I see line offering advantages are,

Could be less friction if done well, translating into less work for the motor and therefor easier higher speed possibilities.

Would not transfer the "cogging" effect from the belt subtly into the print during motion. But since we are still using steppers this would be of limited value. Unless we also switch over to a servo based drive system (many many hurdles with this option)

[lordbinky]

I can't see the pictures (other than ascii art.. ) right now, but replacing the belts with spectra line as it is setup in a rostock you won't see varying length or angles from the pully and spectra winding, so at most you have small firmware adjustments to make to compensate for anything.

If you are doing a continous friction drum/pully, the amount and angles aren't changing unlike a fixed drum/pully where the amount of line already wrapped affects the amount of line that is wrapped the next revolution of the drum/pully. Since on the rostock the lines are fixed to the load, you have a continous setup. Then you setup is just dependent on a few things. One is the number of wraps you have is able to maintain enough friction to prevent slipping. Then the line goes through the bearings in the base (I dunno what the guide calls them but I'm calling them alignment bearings herein),which I'd add a printed grooving sleeve over them to keep the line position constant. Same kind of guide for the bearing at the top. Then it's the same as the belt but with a (likely) different steps/mm.

Flateric nailed it on the possible advantages. being readily availible and cheap(ish) depending on the type of line is also a benefit.

[daftscience]

I have seen the "walking" issue resolved in professional machines before by using a pulley that is free to move on a bearing in a linear form towards and away from the stepper as it is wound. The shaft of the drive is a spline much like a transmission. It cannot move or have any slop rotationally but the line winding is enough force to creep the sleave back and forth maintaining a exact position relative to the rotational driving motor/gear/etc.

................................Shaft rotation and wire windage
....................................................\/
Wire--------------------------------[--]
.......<--Desired linearmotion->>........[--]
...................................................[--] /\
...................................................[--] Sleeve on keyed shaft free to slide up and down
...................................................[--] \/
...................................................[--]
...................................................[--]--------------------------------------wire return
...................................................[...]
..............................................____||_____
..............................................|Stepper or|
..............................................|any drive |
..............................................|source etc|
..............................................___________

I hope that helps illustrate what I mean.
(And if not too bad, LOL. While type "drawing" this I was reminded of how much I hated doing it way WAY back when I was a kid using BBS services that only allowed for this type of illustration to communicate ideas)

So it will without doubt be the last time I ever attempt an ascii drawing again for the rest of my lifetime! HA!

Funny you mention that, I brought this issue up with a coworker and this was his solution. The issue with the RostockMax is that we really don't have a lot of room to allow that horizontal motion. But I'm still not sure how much horizontal walk there is going to be.

...the line goes through the bearings in the base (I dunno what the guide calls them but I'm calling them alignment bearings herein),which I'd add a printed grooving sleeve over them to keep the line position constant. Same kind of guide for the bearing at the top. Then it's the same as the belt but with a (likely) different steps/mm.

That's pretty much the plan so far. I've printed these: http://www.thingiverse.com/thing:49971 and have made a mirrored copy of the threaded pulleys to try out polygonhell's solution if I need it. Then, hopefully I'll be able to get it all together.

[Jimustanguitar]

Any interest in these:

http://www.thingiverse.com/thing:87463

http://www.thingiverse.com/thing:91510

[daftscience]

Any interest in these:

http://www.thingiverse.com/thing:87463

http://www.thingiverse.com/thing:91510

Actually, I've tried both. I have the bearing flanges on now, but they push the belt out enough that it's rubbing with the acetyls. I found the tensioners harder to adjust than without them.

[lordbinky]

Funny you mention that, I brought this issue up with a coworker and this was his solution. The issue with the RostockMax is that we really don't have a lot of room to allow that horizontal motion. But I'm still not sure how much horizontal walk there is going to be.

I don't see how there would be any walking, not a necessary one at least. The angle of the line from the cheapskate to the bearing should stay the same, as well as from that point to the line of the wrapped pully, and so on to the other end of the cheapskate. So all the angles are static unlike a simpler design that had less line control and would need to account for changes in line angles.

[Jimustanguitar]

I don't see how there would be any walking, not a necessary one at least. The angle of the line from the cheapskate to the bearing should stay the same, as well as from that point to the line of the wrapped pully, and so on to the other end of the cheapskate. So all the angles are static unlike a simpler design that had less line control and would need to account for changes in line angles.

He's not crazy, mine does it too.

I have been thinking about redesigning the spacers that make up the buns of the idler sandwich If they were just a tad bigger, they might be a little more persuasive to the belt.

[theverant]

I don't see how there would be any walking, not a necessary one at least. The angle of the line from the cheapskate to the bearing should stay the same, as well as from that point to the line of the wrapped pully, and so on to the other end of the cheapskate. So all the angles are static unlike a simpler design that had less line control and would need to account for changes in line angles.

He's not crazy, mine does it too.

I have been thinking about redesigning the spacers that make up the buns of the idler sandwich If they were just a tad bigger, they might be a little more persuasive to the belt.

I'm having ongoing issues will belts moving to the side and grinding on various metal bits, as well. I had one column okay - it stayed centered, but the other two would walk no matter what I've tried. And if I got the top okay, the bottom wasn't, etc. A real pain. Any tricks in this area would be appreciated.

[Batteau62]

I had some of this at the beginning too. It seemed to help me to deburr that cut extrusion edge with a rubber grinding bullet in my Dremmel. You can't get at the inside without major disassembly though I also added belt tensioners that seem to work pretty well. They at least gave me some control over belt walking. I printed a vase that was 13+ hours with no belt chips, so I called it good. I agree that the "buns" of the idler "sandwich" needed taller flanges(for lack of a better description).
When I first assembled, I wanted to put a thin washer or something in there to keep the belts aligned. Maybe something like a washer stamped out of shim steel at just the right size?

[theverant]

I had some of this at the beginning too. It seemed to help me to deburr that cut extrusion edge with a rubber grinding bullet in my Dremmel. You can't get at the inside without major disassembly though I also added belt tensioners that seem to work pretty well. They at least gave me some control over belt walking. I printed a vase that was 13+ hours with no belt chips, so I called it good. I agree that the "buns" of the idler "sandwich" needed taller flanges(for lack of a better description).
When I first assembled, I wanted to put a thin washer or something in there to keep the belts aligned. Maybe something like a washer stamped out of shim steel at just the right size?

What kind and how are you using belt tensioners? Pics would help, if it's not too much trouble.

[lordbinky]

The angle the belts were pulled through and clamped down affected their alignment on the towers. Even after everything was tighened down I can still go back and tap or lever the "Buns" (hahaha, I can't not use the term now) around so that belt wants to go to the middle. Don't need to uncrew anything for that adjustment. Then I'd also make sure your stepper motor's pully is where you want it since that is an easily adjustable part as well.

Or you could just print solutions

[Batteau62]

Link to the tensioners:
https://www.facebook.com/photo.php?fbid ... c297b519c3
Link to my Rostock album:
https://www.facebook.com/media/set/?set ... 4f2c181935
I had the right diameter rod lying around. The corner brackets are 1" standard hardware. I think I used 6-32 socket head cap screws w/ star washer nuts. I clamped the belts to the cheapskates(guessing at tension by feeling the short length between the bottom idlers) Making sure to bottom out the top idler in its slot. Then I just try to wind the socket heads evenly to get the belt to run true. Last step is to lock down the idler axle. Adjust as necessary. You can't really get any tighter than the slot in the top bracket allows, unless you do a He-man on the belts in the initial tension. I think I might try the shim steel washer idea, I'll post results.

Just realized, sorta high-jacked this thread, sorry daftscience!

[daftscience]

I'm having ongoing issues will belts moving to the side and grinding on various metal bits, as well. I had one column okay - it stayed centered, but the other two would walk no matter what I've tried. And if I got the top okay, the bottom wasn't, etc. A real pain. Any tricks in this area would be appreciated.

I'm glad I'm not the only one. Especially with regard to the bottom.

The angle the belts were pulled through and clamped down affected their alignment on the towers. Even after everything was tighened down I can still go back and tap or lever the "Buns" (hahaha, I can't not use the term now) around so that belt wants to go to the middle. Don't need to uncrew anything for that adjustment. Then I'd also make sure your stepper motor's pully is where you want it since that is an easily adjustable part as well.

Or you could just print solutions

Hmm, the thought of the belt angle on the clamps never really crossed my mind.

I don't see how there would be any walking, not a necessary one at least. The angle of the line from the cheapskate to the bearing should stay the same, as well as from that point to the line of the wrapped pully, and so on to the other end of the cheapskate. So all the angles are static unlike a simpler design that had less line control and would need to account for changes in line angles.

I think the amount of walk on the spectra line depends on the thread width, length of Z axis and the diameter of the drive pully. I think we may be talking about two different things.

Here is a link to my mirrored pully. it fits a 606 bearing. I just need to make a spacer for it.

[lordbinky]

Oh, and I verified that moving the cheapskates with the steppers plugged in dumps enough power to backfeed (turn on) the powerlight I have attached directly to the PSU if I move them decently fast .

[geneb]

lordbinky, that's a really good way to destroy your RAMBo.

g.

[lordbinky]

Yes it is. I had just turned off the printer, reached out and move it up and down a little because it seemed like there was a funky spot and had a 'WTF was that?!' out of the corner of my eye and it was the power indicator glowing a little. After I changed my underwear I decided not to be so impulsive on the printer with my curiousity.

[camais]

Any new revelations on this? I'm looking into this as an alternative to replace my broken timing belts.

Spools and line here: http://www.tridprinting.com/BOM/Kossel-Legacy/ is $30+ shipping

vs

$20 per tower for timing belts

[geneb]

How the hell did you break the belts? That's simply not possible with normal use. (Unless of course you've got the original, all metal, sharp-tooth drive gears.)

Spectra line drive is cute, but IMO is a gimmick.

g.

[camais]

yeah, i got the "all metal, sharp-tooth drive gears" which also makes replacing the belts more expensive because I'd probably end up getting new pulleys also