[caleb105]

CNCFusion mentioned in a thread that their "Deluxe" kit had "Preloaded" ballnuts. How can a SINGLE ballnut be preloaded? I was under the impression that you needed two ballnuts, with something between them to keep the tension (see pic below). Are they just referring to using oversized balls?

Oh, and NOOK said that they can add these oversized balls at the factory for an extra $35 per ballnut. They said this reduces backlash to from .01 to .003. I think I'm going to have to get the $40 SBN series ballnut, because all other options are $300+ :-O I'm planning on doing some sort of DIY preload using dual ballnuts (ala Cadmonkey) and something between them (at least on the X, and maybe the Y).

[escott76]

If you have two sets of tracks in the nut, and you offset those tracks from each other by a set amount, you can pre-load things. This does not account for variances in the screw lead though, so in different places on the screw you have differing backlash.
I used bellville washers to pre-load my double nut setups, seems to have worked quite well. After all was said and done though, $90-$100 for a single pre-load might have been cheaper given my time in machining, but I'm happy with the result.

[project5k]

could the preloaded mean that they have the balls inside already, rather than just sending you an empty nut and some loose balls for you to put together?

my kit came with the ballnuts already on the screws, perhaps this is what was ment?

i know, just offering an alternative interpretation....

[SCzEngrgGroup]

A single pre-loaded nut is one that has had oversize balls installed to minimize (not eliminate) backlash. Another option is double-nuts, with some kind of spring or elastic tensioning means, though adjustability is really not worth much, as there is a "correct" pre-load for each size screw, based on its load rating (usually 10-15% of the rated load). Some of these are adjustable, some are not. The final option is multi-circuit nuts, which have more than one (usually two, sometimes four) separate circuits, and are also pre-loaded to eliminate backlash. These also sometimes have integral elastic tensioning means.

Regards,
Ray L.

[escott76]

could the preloaded mean that they have the balls inside already, rather than just sending you an empty nut and some loose balls for you to put together?

my kit came with the ballnuts already on the screws, perhaps this is what was ment?

i know, just offering an alternative interpretation....

No, pre-load is a specific condition depending on the fit of the balls within the grooves of the screw. There are several ways it can be achieved. Pre-loaded A/C bearings don't mean that the balls are already in place in the raceways, it means that there is some initial force applied to the bearings before any loads are imposed upon them due to actual work done. Hence PRE load. The same thing applies with a ball nut(s).

[caleb105]

Thanks for the info gentlemen.

Here's my conundrum:

Nook SBN standard no-frills ballnuts are $41. According to Andrew the Nook rep, they come with backlash of .01 (pretty bad).

I can either replace the balls with larger ones myself, or pay Nook $35 per nut to do it at the factory. I know that they will wear quicker, but according to Andrew, this would reduce the backlash from .01 down to .003 (better).

The next step up in Nook ballnuts (as far as precision) would be the pre-loaded line, which would cost me $320 each. This is ridiculous for a hobby mill.

Therefore, I am stuck with the SBN nuts, with .003 backlash at its best. I am considering doing a DIY pre-load with dual nuts similar to what CadMonkey did. However, he used BOTH larger balls AND pre-load bracket, and only got the backlash down to .0025. The trouble to make the pre-load bracket doesn't seem worth the .0005 reduction as compared to just using a single nut with larger balls.

What would you recommend as a "target" backlash to get down to? What amount of backlash can Mach3 or EMC2 successfully and accurately handle?

Thanks!

-Caleb105-

[project5k]

i know, i was trying to make a funny.. guess it didnt work.

[SCzEngrgGroup]

Thanks for the info gentlemen.

Here's my conundrum:

Nook SBN standard no-frills ballnuts are $41. According to Andrew the Nook rep, they come with backlash of .01 (pretty bad).

I can either replace the balls with larger ones myself, or pay Nook $35 per nut to do it at the factory. I know that they will wear quicker, but according to Andrew, this would reduce the backlash from .01 down to .003 (better).

The next step up in Nook ballnuts (as far as precision) would be the pre-loaded line, which would cost me $320 each. This is ridiculous for a hobby mill.

Therefore, I am stuck with the SBN nuts, with .003 backlash at its best. I am considering doing a DIY pre-load with dual nuts similar to what CadMonkey did. However, he used BOTH larger balls AND pre-load bracket, and only got the backlash down to .0025. The trouble to make the pre-load bracket doesn't seem worth the .0005 reduction as compared to just using a single nut with larger balls.

What would you recommend as a "target" backlash to get down to? What amount of backlash can Mach3 or EMC2 successfully and accurately handle?

Thanks!

-Caleb105-

Getting down to low backlash by ball-loading alone is a crap shoot, as you're limited by the dimensions of the specific screw and nut, and the availability of different-sized balls - generally quite limited. The best way is a spring pre-load, in which case ball-loading is really not necessary. Properly done, they should be *zero* backlash in one direction under all conditions, and *zero* backlash in the other direction until the load exceeds the pre-load amount. For all practical purposes, this will always be zero backlash. If you build a spring-loaded double nut assembly and don't get zero backlash, you've done something wrong. The spring pre-load should be on the order of 10-15% of the rated load for the screw, which is typically around 180 pounds for a 5/8" screw. This is easily achieved with a stack of several Belleville washers. The tricky part is coming up with a reasonable way of keeping the nuts from rotating, and, if necessary, adjusting the pre-load. If you can do it without having to machine the nut, that can be a plus, as at least some ballnuts are made from unGodly hard stuff. I can't even make a dent in mine with carbide. It is important that provision be made for a slight amount (a few thou) of axial movement, to deal with imperfections in the screws. If you don't do this, you may have binding in some positions, and/or suffer premature failure of the screws or nuts.

Regards,
Ray L.

[caleb105]

Ray,

Thanks for the great info. A few more questions. Is there enough room on an X3 to do double-nut pre-load on the Y travel? Would you need to worry/mess with this on the Z travel?

Also, I'm still not totally sure exactly what a double-nut preload setup looks like. Does anyone have close-up pictures of a setup like this? Cadmonkey's block just looked like a block with 4 holes for bolts, and a pass-through hole for the ballscrew. What do the 4 machine bolts screw into?

Also, I know nothing about these washers you speak of. Do you have any pictures of a setup like this?

ETA: I understand that the basic concept of a double-nut preload setup is to have something in the middle that pushes out against both nuts, but how do you push out in the middle, while still holding the nuts together? See pic:



Thanks guys!

-Caleb105-

[escott76]

Thanks for the info gentlemen.

Here's my conundrum:

Nook SBN standard no-frills ballnuts are $41. According to Andrew the Nook rep, they come with backlash of .01 (pretty bad).

I can either replace the balls with larger ones myself, or pay Nook $35 per nut to do it at the factory. I know that they will wear quicker, but according to Andrew, this would reduce the backlash from .01 down to .003 (better).

The next step up in Nook ballnuts (as far as precision) would be the pre-loaded line, which would cost me $320 each. This is ridiculous for a hobby mill.

Therefore, I am stuck with the SBN nuts, with .003 backlash at its best. I am considering doing a DIY pre-load with dual nuts similar to what CadMonkey did. However, he used BOTH larger balls AND pre-load bracket, and only got the backlash down to .0025. The trouble to make the pre-load bracket doesn't seem worth the .0005 reduction as compared to just using a single nut with larger balls.

What would you recommend as a "target" backlash to get down to? What amount of backlash can Mach3 or EMC2 successfully and accurately handle?

Thanks!

-Caleb105-

I used a pair of $27 (from mcmaster) Thompson nuts per axis. A stack of Bellville washers (also mcmaster, search "disc spring"). I made a block, threaded one nut into it, and put dowel pins through the same block, just enough so that the second nut couldn't rotate on the screw. There are pictures in my thread on "New Homebuilt Mill project" that show what I'm talking about. Sucky part was the $60 tap to do the ball nut mount, you can find them for less, or you can single point the thread on a lathe if you have that capability.
2x$27 gives you $54, and then I'd add $20 per for the tap, since it's getting spread over 3 holes. Another $10 (generous) for the washers and you're looking at $84 per axis.
I've got next to no measurable backlash out of this setup, and although it cost just slightly less than I've seen some pre-loaded nuts go for (especially after counting my machine time). I am happier with this solution than the single fixed pre-load. YMMV.

[cadmonkey]

Correction - I used each method on 2 different axes - I used oversized balls on Y to only use one nut and maintain max travel. I used double nuts on X. Both have some backlash but it is a small and consistent amount that is easily compensated for by my control software. (I think the .0025 was an early measurement, I thought I got it down to .001 or less, I'd have to check my .ini file for EMC to be sure).

And yes, my double nut setup didn't work like I intended as I don't have a strong enough stepper to preload the 2nd nut as much as it needs to take up the slop and still turn - I loose steps - I didn't do a spring separation setup, I compressed 2 nuts together and haven't been able to tune it 'just right'.

[caleb105]

Correction - I used each method on 2 different axes - I used oversized balls on Y to only use one nut and maintain max travel. I used double nuts on X. Both have some backlash but it is a small and consistent amount that is easily compensated for by my control software. (I think the .0025 was an early measurement, I thought I got it down to .001 or less, I'd have to check my .ini file for EMC to be sure).

And yes, my double nut setup didn't work like I intended as I don't have a strong enough stepper to preload the 2nd nut as much as it needs to take up the slop and still turn - I loose steps - I didn't do a spring separation setup, I compressed 2 nuts together and haven't been able to tune it 'just right'.

CadMonkey,

I think perhaps I am misunderstanding your setup. On your X axis, didn't you have two ballnuts facing each other, both threaded into the "Preload Block"? Doesn't this put them at a fixed distance apart? Where does the preload come from? What pushes the two nuts apart?

Also, you didn't do either method (larger balls or doublenut) on the Z axis correct? How did that work for you?

And what kind of backlash were you getting on the Y axis after switching to larger balls?

AND...(last question for this post...I promise) is there a limit to the amount of backlash that Mach3 can correct for?

Thanks!

-Caleb105-

[project5k]

i would suggest that the backlash comp be used as a last resort. Correcting it mechanically will yeild better overall results.. then whatever you cant get out mechanically, then use the BL comp.

[cadmonkey]

CadMonkey,

I think perhaps I am misunderstanding your setup. On your X axis, didn't you have two ballnuts facing each other, both threaded into the "Preload Block"? Doesn't this put them at a fixed distance apart? Where does the preload come from? What pushes the two nuts apart?

Doesn't matter is they are pushed apart or together. As long as the balls in each nut are touching opposite sides of the raceway there's preload. I have 2 blocks on my X - the second is not anchored to the saddle but is anchored to the primary nut's mounting block by 4 screws, so tighten them to remove backlash.

Also, you didn't do either method (larger balls or doublenut) on the Z axis correct? How did that work for you?

Well enough, I run the gib TIGHT so theres sticktion that makes more problem than anything. If I loosen the gib it runs nice and repeatable, but theres chatter...

And what kind of backlash were you getting on the Y axis after switching to larger balls?

I don't honestly remember, it's set in the .ini file for EMC2 so I'd have to go back and look.

AND...(last question for this post...I promise) is there a limit to the amount of backlash that Mach3 can correct for?

Don't know - I use EMC2 I'm open-source. But I think it was project 5k that said do as much mechanically then use backlash comp for the remainder. For hobby it certainly wasn't worth spending $1200 in ballnuts to get the ground nut, though I may get a preloaded double nut for the Z axis and take care of that one with that.

As for preloading yourself - there is a wide variety of ball sizes available. Get the nut and the screw first, see what your backlash is and then measure all the balls and visit the seller toolsupply (I think...) on eBay. He has steps of .0001 up from the nominal ball diameter if I remember. I should have gone one size larger for the ones I ordered for Y, but I had them and wanted it together so I just said the hell with it and was done.

[Teyber12]

stupid idea here-

but what about say on a hobby mill. you have a ballscrew thats fixed to both ends on the table and in the middle goes through the ballnut. What about using 2 cheap ballnuts, 1 of the ballnuts if fixed to the table the other is loose on the screw, but you screw and thread 4 corners of both ballscrew mounting blocks and attach (screws+nuts?) and get it very tight? this way you have a ballnut pushing in each direction of slop, would this not extremely reduce the backlash?

Is anyone following me or should i post a MS paint picture?

cheers

[Old Megawatts]

caleb105,
There are many types of ball nuts and I'm sure some are much better than others. I have a Syil SX3 machine which I have had disassembled several times so I know what kind of ball nuts it has. They use a single ball nut, however, they are in effect a double ball nut. The nut is sctually two separate nuts that are spaced apart with a two piece spacer. The nuts are screwed towards each other against the spacer and locked in place with double locks, thus eliminating nearly all of the slack. Take a look at my post of 12-1-07 on this site.

OM

[cadmonkey]

stupid idea here-

but what about say on a hobby mill. you have a ballscrew thats fixed to both ends on the table and in the middle goes through the ballnut. What about using 2 cheap ballnuts, 1 of the ballnuts if fixed to the table the other is loose on the screw, but you screw and thread 4 corners of both ballscrew mounting blocks and attach (screws+nuts?) and get it very tight? this way you have a ballnut pushing in each direction of slop, would this not extremely reduce the backlash?

Is anyone following me or should i post a MS paint picture?

cheers

That's exactly what I did - there is a section showing the X axis in the file of mine you got Sheet 3 page 7 of 31 in the PDF, the top section which is colored. I know it will work, but it's very persnickety (is that a word?) to tune and requires a stronger X axis motor than I have.

the PDF is HERE for those that are curious.

[ataxy]

picture are always fun

[HackMax]

Greg (CadMonkey) could you possibly be binding the second nut by tightening one of the bolts more than another thus skewing the nut so it rides the screw at an angle? I was looking at this before I started adding the screws to my machine and since you didn't mention any problems with it I didn't think it worth mentioning but now that I hear you're troubles I'm wondering if this could be part of the problem. I haven't added a second nut yet because I would like to see what larger balls would do but I was looking for an alternative to let the nut self-align on the screw but as of yet I haven't figured out a way to keep them tight and aligned properly to prevent binding. I got something in mind but right now its just a thought. I have to draw it up and see if it has potential.

I'm getting .0035' backlash from my $25 Roton ballnuts and when I saw that Nook has .01" for theirs I thought there has to be a mistake somewhere. It just doesn't sound right.

Rick

[SCzEngrgGroup]

Greg (CadMonkey) could you possibly be binding the second nut by tightening one of the bolts more than another thus skewing the nut so it rides the screw at an angle? I was looking at this before I started adding the screws to my machine and since you didn't mention any problems with it I didn't think it worth mentioning but now that I hear you're troubles I'm wondering if this could be part of the problem. I haven't added a second nut yet because I would like to see what larger balls would do but I was looking for an alternative to let the nut self-align on the screw but as of yet I haven't figured out a way to keep them tight and aligned properly to prevent binding. I got something in mind but right now its just a thought. I have to draw it up and see if it has potential.

I'm getting .0035' backlash from my $25 Roton ballnuts and when I saw that Nook has .01" for theirs I thought there has to be a mistake somewhere. It just doesn't sound right.

Rick

It's not a mistake. Single nuts are simply not designed to have zero backlash. Ballnuts are for low-friction drives. If you want zero backlash (there are many applications where backlash doesn't matter), you use a double nut. It is *critical* that the second nut be able to "float" a bit, or you WILL get binding in some positions and/or backlash in others, as the screw pitch is not absolutely constant. Rigidly locking the two nuts together is NOT a good idea, and will generally lead to rapid wear.

Regards,
Ray L.