I have designed an application-specific, zero-backlash gears system. RFC ...

Hello guys,

I am quite unsure if this application I have will be zero backlash using the solution in the attached image. I don't really have any means to test this so asking around sounds like a good feedback chance.

Edit: the preload can take the form of a tensioned cable like in the pic below:



I believe that the tension in the cable will create a "closed loop" where at no time the motor gear will loose contact with any driven gear upon direction reversal. both driven gears will be pushing the motor gear as shown in the img. Or at least this is my theory.

PS: the gears will only turn 80-90% of a full turn before changing direction. Oh, and I am thinking about herringbone gears -if that matters at all to the sero backlash issue, but will at least prevent any axial movement-.

Thanks for your feedback.

Edited 5 time(s). Last edit at 04/21/2017 09:48AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

I think you right, as long as the preload tension is greater than the force needed to overcome friction and provide the requested acceleration of the mass of the print head or whatever else the cable or belt moves.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thanks DC for chiming in, I find myself from time to time in search of a very low or no backlash solution and this might be the closest I got so far. But I fear there is something I am missing and this might have backlash after all. I say this because it is somehow mandatory that there is a gap for any tooth to be able to engage so that gap will translate in a disconnect when reversing direction. But just maybe the tension in the cable is able to control that...

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

The gap should always be below the bottom of each tooth on the small gear that is engaged, as long as the preload exceeds the force demanded.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

And take care, that your tensioning cable is elastic enough so that i creates sufficient tension, even when your mechanical tolerances try to make it "loose"...

@DC: Ideally the tooth engagement is to have contact with both larger gear's adjacent tooth flanks or with several teeth but in practice the gears are not perfect and there is always a gap besides the tooth (pls excuse me if this looks like lecturing you, it's not):


However, because I am tensioning the cable (steel?) so that both driven gears press on opposing flanks on the motor gear's teeth (somehow preventing it to turn), even with the backlash, at each direction of rotation the motor gear will drive one of the gears with tooth-on-tooth contact and that driven gear, via the cable, will rotate the other gear accordingly, thus having no backlash at all (in theory):



Now, if all still looks good to you too, I was thinking of steel cable (can even be adapted for spectra I guess if this is to be used on a light CNC machine like a 3d printer). If the tension alone is not enough and it needs more elasticity I would add a spring along the cable but my idea is that tension in a non-elastic cable will be enough because, like I have explained above, the backlash will be a non-issue once the motor gear engages one of driven gears teeth and the driven gear is basically engaged by the cable pulling on it as opposed to the motor gear.

@powdermetal... I am afraid I don't quite get what you are saying. Could you elaborate a bit or explain so it gets through my skull

Edited 2 time(s). Last edit at 04/21/2017 02:11AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

When the drive gear in the middle is turning in one direction, its tooth/teeth are pushing one of the two larger gears because those surfaces are in contact. The other large gear is not in contact with the driven side of the drive gear tooth/teeth. You're relying on the cable tension to turn the other gear. If the load is greater than the cable tension the there will be backlash until the drive gear has turned far enough that the driven side of the drive gear tooth/teeth come into contact with that second gear. When the drive reverses direction the the roles switch and during that transition there will be backlash. I think...

For a small load it's probably backlash free. I think...

Edited 1 time(s). Last edit at 04/21/2017 07:37AM by the_digital_dentist.

---

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

@dd: Yes I rely on the tension in the cable to grant the backlash-free operation. That is the main idea because otherwise any gears will have backlash by operating the way you describe it. I even imagine a metal ribbon used to lock the loop so that at all times the driven gears contact the motor gear the way I shown in the picture. I don't see it not being able to take large loads ... unless it gets to the plastic deformation of the metal... but for that it really needs super high-torque in the motor -I believe the largest pull/jerk will be when changing directions at high speeds and only if the motor is powerful enough. Otherwise it will just skip steps. And also these parameters can be adjusted in the software.

Now, my application only needs half to a full turn to the large gears but I wonder how could I design it for continuous operation and still have the preloading work the way I sketched it. Maybe it's possible somehow... well... that's for another application altogether

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

No need for the cable, you can just stack the gears and put a spring between them.



It gets even easier if you are printing helical gears because you can just put them on a spline or keyed shaft and slip a spring washer between them. This has the advantage of not being backdrivable so the anti-backlash properties persist even under high load. Your tooth form needs to be flawless for this to be smooth though.

Edited 2 time(s). Last edit at 04/21/2017 12:28PM by 691175002.

Quote
realthor
@DC: Ideally the tooth engagement is to have contact with both larger gear's adjacent tooth flanks or with several teeth but in practice the gears are not perfect and there is always a gap besides the tooth (pls excuse me if this looks like lecturing you, it's not):

However, because I am tensioning the cable (steel?) so that both driven gears press on opposing flanks on the motor gear's teeth (somehow preventing it to turn), even with the backlash, at each direction of rotation the motor gear will drive one of the gears with tooth-on-tooth contact and that driven gear, via the cable, will rotate the other gear accordingly, thus having no backlash at all (in theory):

Now, if all still looks good to you too, I was thinking of steel cable (can even be adapted for spectra I guess if this is to be used on a light CNC machine like a 3d printer). If the tension alone is not enough and it needs more elasticity I would add a spring along the cable but my idea is that tension in a non-elastic cable will be enough because, like I have explained above, the backlash will be a non-issue once the motor gear engages one of driven gears teeth and the driven gear is basically engaged by the cable pulling on it as opposed to the motor gear.

@powdermetal... I am afraid I don't quite get what you are saying. Could you elaborate a bit or explain so it gets through my skull

Imagine, your gears are out of center a little bit and maybe they are not truly round (as always in reality). What will happen when using a quite rigid cable?
1. At some point during rotation, the rigid cable will be tensioned so much, that it will be hard to drive the gears even further.
2. At another point during rotation, the tension will vanish and the cable will get loose - no longer a zero-backlash situation.

Adding some elasticity - as your proposed for example by using a spring - will compensate these geometric errors.
Commercial anti-backlash gear solutions use various kinds of elastic elements: Rubber, metallic springs (omega-type or coil-springs), ...

I know about these scissors or split spring-loaded gears but I need the cable to be inline and the whole application of this is to have that cable to move the payload. So employing such a split spring-loaded gear will be redundant at least. The spring loaded split gears, as far as I know, are not very precise either as they settle at half backlash whereas I need predictable movement, which I believe my design offers.

@powdermetal: Ii reluctantly aggree , some room for error is needed for such situation and if the tension I put in the system is too much then the motor will skip steps and if it's too low the backlash will not have been rendered out of the system. So I believe that either the cable should be capable of minute elasticity while still providing quite strong tension to overcome backlash or a spring will be necessary.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff