[FXC]

I have my Taig mill apart and I want to replace those embarrassing nylon-tube couplings that have more backlash than an adjustable wrench. I am now pondering about what would work best in my situation:
- OR - ?

...and my situation:
I'm milling tiny acrylic parts of shapes involving many little curves and corners. Instead of bothering to figure speeds and feeds and rpm's, I just go slow and flood the thing so it doesn't melt. The #1 result that I'm looking for is an as good as possible edge finish. Would my choice of couplers affect it and if so, which type should I go with?
Thanks!

http://www.designworldonline.com/art...Selection.aspx

[YZF]

Don't use those on the left, even the slightest misalignment will break them...

[Andre' B]

Don't use those on the left, even the slightest misalignment will break them...

BS, the entire reason for their existence is to allow for a little misalignment.
Now when building a machine tool a little misalignment is 0.001" maybe 0.002", get up around 0.005" and you are well out of the little category.

Pick one with the ratings to fit the job.
http://heli-cal.com/cm/Products/Flex...ings/Home.html

[diecutter]

I used the ones on the right for my DIY machine and they worked great. The brand was Lovejoy, available through MSC, and the best part is you can buy each component individually allowing you to connect a different diameter stepper shaft and acme shaft without any modifications.

[YZF]

In theory yes. In practice not so well.

The problem is they are extremely sensitive and they stress out over time and break. I've seen them break in commercial high precision equipment and the mechanical engineers I used to work with (who designed similar equipment) avoided them. If you minimize the misalignment and work well within the spec you should be OK but why take the risk? Aligning your motor shaft to within 0.001" of the screw isn't that trivial (and there's also the parallelism). There are other styles of couplings that work better.

BS, the entire reason for their existence is to allow for a little misalignment.
Now when building a machine tool a little misalignment is 0.001" maybe 0.002", get up around 0.005" and you are well out of the little category.

Pick one with the ratings to fit the job.
http://heli-cal.com/cm/Products/Flex...ings/Home.html

[ger21]

BS, the entire reason for their existence is to allow for a little misalignment.

Yep. I've been using them on my router for about 6 months now, and they've worked great. And I definitely have some misalignment.

[YZF]

I've also seen them work on many machines and I'm sure if they didn't work at all they wouldn't exist. It's just that they are generally less forgiving so the style on the right is preferable for the hobbyist (IMHO). YMMV depending on the size, materials, misalignment, loads etc.

The linked article also says something to the same effect:
"It is able to manage all types of misalignment, but works best with angular misalignment or axial motion. It is not well suited to parallel misalignment because the single beam must bend in two different directions simultaneously, creating larger stresses in the coupling that could cause premature failure. "

Yep. I've been using them on my router for about 6 months now, and they've worked great. And I definitely have some misalignment.

[LazyMan]

Keep in mind that you cant use the Oldham style couplers on the right unless your design includes a fixed bearing support. I prefer the oldham type couplers as I do agree with YZF that they are more forgiving to parallel misalignment. They both will compensate for minor misalignment and work quite well though.

[ahren]

The helical couplings on the left are my vote as well -- that's what I have on my machine, and they have held up just fine. I also used them for years on pick and place robots, with no ill effects. They are definitely superior to Love Joys, which I have also used, but have 0.003"-0.005" of backlash typically. Check out Matt Taylor's quick backlash analysis of Love Joy couplers -- scroll down the page and you can see his test where he oscillates an axis back and forth. Coupler moves visibly, screw does not.

http://www.peakeff.com/beta/PostDetail.aspx?PostID=11

Best regards,

Ahren
www.cncrouterparts.com

[Andre' B]

In theory yes. In practice not so well.

The problem is they are extremely sensitive and they stress out over time and break. I've seen them break in commercial high precision equipment and the mechanical engineers I used to work with (who designed similar equipment) avoided them. If you minimize the misalignment and work well within the spec you should be OK but why take the risk? Aligning your motor shaft to within 0.001" of the screw isn't that trivial (and there's also the parallelism). There are other styles of couplings that work better.

To be fair asking any coupler to allow for significant misalignment while at the same time providing very high rotational stiffness and zero backlash is asking a lot.
Which is why I think spending a bit more time to eliminate the misalignment in the first place is worth doing.

On something the size of a Taig the torques involved are small enough that with a little TLC during setup they should work fine. Any a Lovejoy type coupler that allows for misalignment is going to introduce some lost motion that will get worse as it gets worn.

[diecutter]

The reason I prefer the Lovejoy style couplers are 1) they are one fifth the cost of the helical, 2) they can be selectively assembled to mate two different diameter shafts, 3) they are more forgiving of misalignment, and 5) with my 10 tpi acme rod the .003 to .005 backlash translates to only a .0003 to .0005 movement on the machine. For my diy machine that is plenty.

[YZF]

Seems to have turned into some sort of a religious war ;-)

I was looking at some data in the Lovejoy web site. They actually make all those styles of couplers:
http://www.lovejoy-inc.com/uploadedF...V7wocovers.pdf

I guess when people say lovejoy couplings they mean oldham?

And yes, the beam style couplers are inherently "zero backlash" because they are a single solid piece of material. That doesn't mean you'll see zero backlash at your tool. It also doesn't tell you anything about the dynamic behavior which is related to stiffness.
There is also a significant difference in performance between double beam and single beam couplers and it's also important how the coupler attaches to the shaft.

It's meaningless to talk about backlash at the tool in terms of distance without knowing anything else about the setup. The only thing you can talk about is angular backlash of the coupling.

Any of these couplers will perform an order of magnitude better than a nylon tube.

[Andre' B]

To be realistic for a Taig without ball screws a length of rubber air hose and two hose clamps would likely be an improvement over the stock couplers.
And if you want to smooth things out a bit of rubber between the stepper motor and the screw could help.

Kind of goes against my instincts because when I think smooth precision motion control. I start thinking collet type couplers and a torque mode servo system with 100,000 plus count encoders or an older school velocity mode servo system with analog tachometers.
A bit of an over kill for most hobby stuff.

[Mike Nash]

The "Lovejoy" couplings shown are not Oldhams. Different although somewhat similar design.

http://knol.google.com/k/steven-elli...swup37wd53q/3#

[wildwestpat]

Hi Folks

Assuming that the coupling is being operated at less than the maximum design torque there will be nothing to chose between the helical beam and the lovejoy / oldham couplings. The point to note is that the centaral element in the oldham / lovejoy is available in different materials from soft elastic to very ridgid materials. The softer materials acting as a shock absorber which can help with some applications where resonances are a problem. Unfortunately there are some low cost couplings available in all the styles. The central element should be a tight fit in the 'forks' in the metal element of the lovejoy / oldham fork. Again there are cheapo editions that exhibit a measurable clearance between the central element and its corresponding slot side walls - there has to be end clearance to alow the central disc to float and change position with the out of line angle.

Provision for maintenance as well as initial assembly order will also determine which type of coupling is best suited to the application. The helical beam type requires the two shafts to be parted suffciently for the coupling to clear the shafts as for example the motor can be removed as an entity from the assebled machine. On the other hand the lovejoy / oldham only requires the central element to clear one half of the coupling fork in the case of the cross style coupling pictured in Mike's post for placement / replace the cental element and the complete coupling only requires a gap to clear the length of one of the metal coupling parts not the entire coupling an important feature if the central element is used either as a 'mechanical safety fuse' or as a damper element. This is often a simple as sliping the coupling along the shaft leaving all the other components undisturbed saving machine down time and costs compared to the use od helical beam couplings.

Where it is vital that no backlash is added by the coupling then bellows couplings are used but these are too costly for most applications and are found in precision servo systems where some small angular alignment issues remain after fitting. For machine tools bellows couplings are limited by their relatively poor peak torque capability for a given diameter.

In my experience the three element lovejoy / oldham coupling is a firm winner in its spider form as it gives warning of problems without losing drive and makes a metal on metal noise which most machine operators report promptly even if they do not switch off!. The helical beam fractures and the drive is lost often with catestrophic consequences where varios axis motions are synchronised in a precise cyle.

For simple couplings thick wall rubber or plastic tube or even helical coil springs of suitable diameter with appropriate shaft clamps are satisfactory provided the torque transmission and elastic damping are acceptable.

My own prefference is for hard plastic inserts for servo applications and elastic (but with a high shore figure) for stepper motor drives where the slight shock absorbing effect can help quell resonance problems.

Hope this helps in the selecting of the appropriate coupling as it is horses for courses with many factors determining the most appropriate.

Kind regards

Pat

[stevespo]

I've used the helical style coupler for years and been very happy with it. According to the specs on the McMaster site, they are pretty forgiving with parallel, angular and axial misalignment.

I'm switching over to the BF30x42 style flex (spider) coupling from Zapp automation because my new motors have a 14mm shaft and I couldn't find any easy way to adapt to my existing ballscrews. The important thing is to make sure you get a style that is zero-backlash. Many of the spider type couplings are not, although they all look similar.

Steve

[FXC]

[...] horses for courses [...]

The OP mentions I'm cutting acrylic, going slow, with flood => virtually no resistance, thus no high torque requirements. I also mention that the parts have various intricacies that have to come out well because they will ultimately be fitted into something else => no room for backlash. Finally, it's also mentioned that I need the contours to have a finish as smooth as possible and here's where vibration can be harmful, am I correct? If so, then the spider option could be the answer.

I'm just wondering... how can a spider type be sold as "zero-backlash" and offer damping at the same time? My guess is that (as with damping systems in general) the rubber element gets compressed initially, then gradually extends back, ultimately resulting in no backlash - but all this happens with a little delay, which basically means a theoretical backlash for a fraction of a second, while the initial compression takes place.

But wait a second... this isn't a nanotechnology forum here, what the heck? (wedge)

[YZF]

I'm just wondering... how can a spider type be sold as "zero-backlash" and offer damping at the same time? My guess is that (as with damping systems in general) the rubber element gets compressed initially, then gradually extends back, ultimately resulting in no backlash - but all this happens with a little delay, which basically means a theoretical backlash for a fraction of a second, while the initial compression takes place.

It is zero backlash because it has no slack (and no "memory"). That doesn't mean it doesn't deform or compress under load. Everything does. Even a solid piece of steel. As long as there is force applied to overcome friction or some workload there is always something that looks like backlash. It's just that a stiffer coupling, screw, and machine will result in less of that "backlash".

In your typical machine there are always forces applied to overcome friction and to do whatever work you need to do therefore there is always something that looks like backlash even if there is no mechanical slack.

The classical meaning of backlash refers to the slack you have in gears where when you change direction you have to rotate a certain amount before you engage the other side of the "tooth".

[DIYaholic]

It is threads like this one, that I get the most out of. As a newbie, there is much that I don't know. :withstupiA "coupling is a coupling", who would have thought so much discussion would evolve. I just want to say thanks for all the information that this thread has/will provide me and other newbies.

Randy,

[Riceburner98]

I finally exploded a steel "beam coupler" the other day... 269oz motor @ 2000 RPM, 400 ipm, I guess I crashed 1 too many times! Funny thing was it exploded during a slow jog after I had been cutting aluminum for a half hour. I would have replaced it but a similar one was $80. I bought an Oldham style one rated at 5x the torque for $40.

Thing is, I got waaay more backlash from the cheap ACME nuts in my Taig than the cheap couplers unless I cranked the gibs (and nut backlash adjust screws) too tight for any decent speed use...