[men8ifr]

No the 7202s are angular contact so I think the author meant 2 basic grade angular contact bearings with preload will still give backlash...

[RCaffin]

I think the author meant 2 basic grade angular contact bearings with preload will still give backlash...
Really?
With the correct preload?
Forgive me if I think that is ... not quite correct.
May I suggest you read some good engineering literature on the subject?

Cheers
Roger

[men8ifr]

Well that was my question -

However from the article

Even a T7 rolled screw will benefit greatly from a god set of fixed-end bearings. This is one area which beginners often skimp upon, and they then wonder why their 0.002" ballscrew performs so badly, and has 0.008" of backlash, even with a "0 backlash" ballnut installed.

I took to mean poor quality angular contact bearings - in fact he meant non-angular contact as the 'poor quality' reference - so then he is not saying quality of the set of angular contact bearings is critical - which is what I was expecting hence the earlier question...

[skrubol]

Getting proper preload on cheap AC bearings might be more difficult though. Matched AC bearings usually have the inner faces (the face pointing at the other bearing) of the races ground so that when the pair are installed tightly in contact with each other they will have proper preload.

[Mike Everman]

I concur! For DIY, for the double row bearing mentioned, use 5202 with seals. $13 ish on VXB.com. Same diameters as 7202, but a bit shorter, will need washers to get the length right if you run out of threads. Backlash is extremely low. Makes me want to go back in the lab to quantify that. We use hundreds of these.

[men8ifr]

Skrubol you are correct, it also appears that the matched pair is to control preload - at least that is what I leared from

https://www.ktaiballbearing.com/prod...tact-bearings/

Matched Pair Angular Contact Bearings, also called duplex bearings, are machined as matched sets. The pairs are furnished with the race faces ground to provide preload when installed. To accomplish this, a gap is provided between the inner races or outer race of the pair.

But the design used where you also fit a nut to the shaft - as long as you use a washer (on outer edge of bearing only I believe) then you can pre-load without a matched pair and saving vast amounts of money :-)

[men8ifr]

I concur! For DIY, for the double row bearing mentioned, use 5202 with seals. $13 ish on VXB.com. Same diameters as 7202, but a bit shorter, will need washers to get the length right if you run out of threads. Backlash is extremely low. Makes me want to go back in the lab to quantify that. We use hundreds of these.

Thanks Mike I'll have a look at the dimensions/specs of the 5202 as an alternative to 7202's - either way I will be using double angular sealed bearings with preload via nut on the shaft.

[men8ifr]

Thanks Mike I'll have a look at the dimensions/specs of the 5202 as an alternative to 7202's - either way I will be using double angular sealed bearings with preload via nut on the shaft.

Ah 5202 is a double row angular bearing - I would prefer the 7202's so the preload can be set/adjusted.

[RCaffin]

I don't think the 5202 bears much resemblance to a matched factory pair of 7202 ACBs except in superficial appearance. Certainly not in performance imhe. A good bearing, but for a different application.
Factory pairs of ACBs tend to be a shade more expensive than two stock ACBs. In a production environment the extra cost is not significant. For hobby use you can arrange the preload yourself by carefully doing up the retaining nut. Yes, it can be done by 'feel'. You will need one of those locking washers where you hammer a couple of tines over. Don't try to use a lock nut: that will wreck the preload.

Cheers
Roger

[borgsteam]

Thanks for this writeup... the time and effort taken is really appreciated.

Line on the rolled leadscrews: I would guess this is created first to give the displaced metal somewhere to go as it is forced out of the grooves? Or it might form naturally during the displacement process? (Wild guesses both).

I don't know if anyone has given any confirmation on this or not, but I have seen that all of the rolled acme I have, has this line too, and the inner area of the poorly rolled ones, from zoro tools of all places, the groove is big enough and unprocessed enough to see that it is indeed from migration of cold rolled metal building up, though I have seen nearly non-existent line rod that was advertized as rolled from wholesale tool, and luckily I had a piece here I could examine under a strong magnifying glass, to see for sure, there is indeed a very tiny line wavering around through that plateau also called a crest in some places I have looked on threading details, (valley being the groove for the mating "crest" and walls to fit into for creating pressure when rotated) so my guess is they had a secondary rolling tool that was pressing down on these crests to flatted out the two rising bits of metal till they were smooth, which is probably good to avoid loss of lubricates out of that when they are turning.

I have actually been looking at making some tooling that I may need to run coolant through, and thought of running an additional roller so I have two start that the middle additional thread is smaller and has two rollers that will finish pressing these together with enough force to seal it shut with out filling the void, so that a tube is spiraling through the rod below that crest, to press fluid or air through to cool the rod, though this is intended to be done with aluminum and use friction pressure to infiltrate it with micrograin carbide powder, to avoid wear, due to them being aluminum, so it does kinda make me wonder if rolled could be better, if like mine, a final grind is applied to force the accuracy by removing everything down to a few thousandths under where the lowest valley of pressure would be in any of the rod's threading.(also for tougher material, maybe using titanium in place of aluminum all under argon so no TiO2 or Al2O3 that forms passively is pressed into the rod. then impregnating with powdered tungsten carbide, or for more durability tantalum carbide. though the titanium is a poor thermal conductor thus why I have not thought of using it as my above mentioned custom threaded rod lol)

[borgsteam]

oh also, on the "preloading" for removing backlash, in ballscrew/nut, I was under the impression that there was pressure on the entrance and exit points that the balls were pressing against continuously much like spring loading two nuts that can not turn independently to get some preload of nut threading against the screw threading, so movement is instant as long as the pressure is lower than your spring's pressure lol, anyway, the point I would figure is something you can envision, with soda cans where if you have the rolling through a space, standing on end, and the gap is tight, the cans are in a straight line, but if the gap is wider than the cans, they start forming a sort of zig zag pattern, so that the cans are packing in the gap as tightly as they can, with less than one can width from entrance to where the second can is at, since it is rolling against the other wall, and if the walls get widened over time (wear or pressure and compression off the wall) then the cans just go into an even wider zigzag pattern so that one side of the wall or the other has cans pressed firmly against it, thus pushing on one can instantly moves all cans and creates pressure on both walls instantly, so the balls would be doing the same thing but the zig zag is in a spiral too, to make things more complicated, so that as your walls of the nut and screw are worn the pressure pushes the balls firmly into the groove to form a wider zig zag pattern so that any rotation instantly generates pressure on walls of the groove so that no backlash is noticed. would also mean that packing them is probably a pain. this remains to be seen, as I plan at some point on making my own )

[Dip1810]

Dear Swede,

Thank You For sharing information.This was really a great article for those who don't have much knowledge of Ball Screw.

I found a web page about ball screw in that they are explaining the information about ball screw and their type.

Do visit to know more information
THK Linear Guide Supplier | THK Ball Screw Distributor | Exporter in India - Seimitsu Factory Automation Pvt Ltd.

[JKAVS]

Could someone explain the difference between DFU and OFU ball nuts? DFU means double nuts but OFU means off set double nut. What does the "off set" mean here?

[dustin1706]

Could someone explain the difference between DFU and OFU ball nuts? DFU means double nuts but OFU means off set double nut. What does the "off set" mean here?

The specific design can vary a bit by manufacturer, but basically it's just two ways of accomplishing the anti-backlash. The DFU uses a spring or spacer of some sort to apply preload. If your machine force overcomes this spring/spacer compressive strength, you will end up with backlash. The OFU is actually a continuous, one-piece nut, but has a small offset in the threads halfway through that applies the preload by providing two points of contact.

The advantage of the OFU is that it's a solid piece that can withstand more force, so it's meant for heavy duty applications. The downside is that it requires a precision fit to the ballscrew (get them from the same manufacturer) and is not adjustable. Unless you have a specific use case for OFU, go with a DFU style nut.

[RCaffin]

The DFU uses a spring or spacer of some sort to apply preload.
I don't like the spring versions as they will give backlash at some stage.
The spacer versions can be hit and miss too, unless they are made with extreme precision.

I prefer the double-nut version which is manually adjusted with a very fine screw-thread. You wind the second nut out (or in) until you feel some resistance, then you check how easily the ball-screw rotates in the assembly, and fine-tune until there is just a slight drag. That seems to give close to sub-micron backlash. Then you have to lock the adjustment of course: not with a lock nut but with a toothed washer.

However, this only works with a good uniform ball-screw. If your ball-screw is worn significantly in (say) the middle, then what is tight at the ends will be loose in the middle. But then you need to replace the ball screw anyhow. You could keep it going for a while with a spring-loaded version, but with reduced accuracy. TANSTAAFL

As for a zig-zag packing - forget it. Normally the balls in a ball-screw are not jam-packed together (and should not be so!), and the grooves should definitely NOT have that sort of slop anyhow. Does this mean those cheap poor-tolerance ball-screws should not be used in a CNC? Correct. They may be good for motorised window-openers, but not in a CNC.

My 2c
Cheers
Roger

[Stevecjx]

hello friends,
This is Steve from MSM Store.We're selling various of Linear Bearings,Shaft Holders,Couplings,Handles,Manual Screw Jacks,Press Plates,etc.Mainly all kinds of accessories of machines.If you have in need of such items,please feel free to contact me [email protected]
You may also visit our site at https://www.msmstore.com/ for details.
Thanks

[rostones]

"Ballscrews? What the heck is a ballscrew, and why use one? My machine came with ACME screws, and those seem more than adequate for my CNC conversion!" This is a fairly typical question that someone new to CNC is prone to state. In this article, I will "scratch the surface", so to say, of ballscrew characteristics, and importantly, which type to use in your particular machine.

Expect no mathematics or analysis of loads; these will vary with the machine and the type of work you intend to perform! Instead, I'll propose some very simple rules of thumb regarding the sizing of the ballscrew for your machine. Take that portion of this article with a huge grain of salt. If in doubt, there's plenty of research material available on the internet from companies like Lintech.


A precision ground ballscrew is created as its name implies... exceptionally accurate and well-dressed abrasive wheels cut the ball channels in both the screw and the nut. Tolerances and overall accuracy can be phenomenal. So can the price! Two additional benefits of a ground ballscrew - they tend to run much more quietly than rolled, and due to the precision of the threadform, it is not a difficult matter to create a ballnut that exhibits zero backlash. Rolled ballscrews also can be made with zero backlash, but it is a much more involved process than a ground ballscrew. To produce this happy state in a rolled screw, usually two ballnuts are used, mounted back-to-back, with some form of powerful spring tensioning (like belleville washers) between them. Sometimes oversized balls or skewed tracks are used, but lead variation of the shaft threads makes the latter two methods much more difficult to do properly.

I believe I can say without quibbling that ground ballscrews are always better than rolled. The problem is in the cost. A 3' section of 5/8" dia. rolled ballscrew can be had for perhaps $40 new, and a nut for another $25. A similar length of ground ballscrew will easily top over $1,000 for one of good accuracy. Fortunately, ground ballscrews are available on eBay for reasonable prices; otherwise, very few of us could afford them for our home shops. They are used extensively in the semiconductor industry, and scrapped clean-room equipment can yield some very nice ballscrews.

How to identify the difference visually? It's pretty easy with just a bit of experience. To start, ignore the journals, these are usually ground for radial bearings and they all look the same. You'll want to first examine the shaft. Quite simply, ground ballscrews flat-out look better. The finish of the grooves is quite shiny, and very uniform. The "lands", those portions of the shaft which seperate the ball grooves, are finished as nicely as the grooves. They are normally plain, with no markings, lines, or other features. Now, move to the ballnut. A ground ballnut is often bulkier than a rolled ballnut, and the vast majority have a flange, meaning that the round body of the ballnut ends with a round, rectangular, or square flange perpendicular to the shaft axis. The nut's return tube may be internal or external. The finish on the nut should be very similar to the shaft, in other words, bright and uniform. Usually there is a plastic wiper installed on both ends of the nut to keep debris out of the nut's internals.

A rolled ballscrew will usually exhibit one or more of these characteristics: the shaft is often a black oxide finish. I've seen almost NO ground ballscrews other than bright. The finish, whatever the color, isn't as uniform, nor is it as pleasing to the eye as ground. The nut often has no flange, but is instead threaded to mount an accessory flange. (There ARE many exceptions to this, however.) Usually, there is some form of marking on the lands. Often, it is a light groove which spirals about the screw on the center of the land. I have no idea why that line exists; perhaps it is used to index the screw as it is being formed. Whatever it is, it is VERY common to rolled ballscrews.

By manufacturer? Sometimes. Thomson and Warner ballscrews are almost always rolled. Star, and especially THK, offer a huge mix of rolled and ground. Just because it's a THK doesn't mean it's ground! NSK screws are predominantly ground. These are GENERAL rules only!

On eBay, use caution - USUALLY the seller will identify a ground ballscrew as such, while a rolled ballscrew is usually referred to as only a "ballscrew".

Attached to the end of this article are two pictures, ROLLED.JPG and GROUND.JPG. Study them a bit.

One other method... if the seller simply describes the ballscrew with an accuracy class of C5 or better, especially C3 or better, it'll be ground.

With that, let me touch on accuracy. Ballscrews usually have their own special model #, which varies with the maker. Fortunately, the accuracy designation tends to follow a numerical format. The numbers start at 0 and go upwards, usually by twos, so you can have a C1, C3, T7, etc. The lower the number, the better the accuracy. Ground ballscrews start at 0. Usually the designation is C0; sometimes you see P0.

Some typical lead variations for different accuracy grades...
C0 - 3um or 0.0001" per 300 mm / 12"
C3 - 7um or 0.00027"
C5 - 14um or 0.0005"

The longer the stroke, normally, the larger the variation will be. In other words, it is very difficult to create a 1 meter section of C0 ballscrew, but much simpler to execute the same over only 20 cm when a C5 tolerance is indicated. Remember these values are WORST case! The average lead variation for these screws is much smaller... only perhaps in a certain portion of the travel may the variation reach these values! So you can see that ground screws can have terrific accuracy. They often come with a graph of their lead variation, created with a precision laser measurement device.

Above C5, we enter the rolled ballscrew range. The very best precision rolled screws can reach C5, but these are also quite expensive, not much less than a ground screw.

Often, you'll see the C or P designation give way to a T designation, the T standing for "transport". T-screws are often used in industry for actuating imprecise motion, like a flap on an aircraft, a gate valve, etc. But that doesn't mean they cannot be used for CNC! Just understand what you are dealing with. The majority of rolled screws you'll encounter will be T7 grade.

T7 - 52 um or 0.002" per 300mm / 12"
T10 - 210 um or 0.008"

You can see that T7 isn't bad, but above that it can get ugly, at least for a CNC machine.

Very quickly - don't get confused by repeatibility vs accuracy. Even the crummiest ballscrews usually have excellent repeatibility, which is nothing more than having the ballnut repeat to a specific point. That's wonderful, but if the accuracy is poor, a high repeatibility won't help you if you need a part of a very specific dimension.

"Great! I'll use a ballscrew. It has a number of powerful advantages over ACME or any non-recirculating screwform." Not so fast!

Hand in hand with the ballscrew is the supporting elements. To be effective, any leadscrew must be fixed axially to absorb thrust loads. Yet it must be free to rotate so that your CNC machine actually moves. The answer (I'm sure you've already guessed this) is a radial bearing, commonly called a ball bearing. Before you run off and buy some inline skate bearings, bear with me, we have more choices to deal with!

A C0 ballscrew is worthless if it is supported by a single, standard radial ball bearing. Such a setup has little ability to absorb thrust loads, and the entire ballscrew will move, along with the inner race of the bearing, under load. And it doesn't take much to move over 0.001", negating all of that expensive accuracy! The ballscrew must be FIRMLY fixed in place on one end, usually the driven end, and to do this normally requires a pair of bearings, mounted back to back. The ballscrew driven journal is physically clamped between two inner races, usually between a shoulder in the ballscrew, and a nut. When the nut is tightened, the two inner races are loaded relative to each other. If we then firmly fix the OUTER races in a block or end plate, the ballscrew is now free to rotate relative to the plate, but will not move in an axial direction. This is best achieved by using what are known as angular contact bearings. In these bearings, the inner and outer races are tapered, meaning if they are loaded relative to each other, they will no longer move axially. Rather than try and describe such a setup with words, please refer to the attached print of a typical bearing block with two angular contact bearings. Don't just glance at it if you do not understand the relationship, study it a bit and it will become clear.

For simple setups, a pair of cheaper 7200 series of bearings will work fine. These bearings can be purchased new for perhaps $12 U.S. each. For more accuracy, higher-precision, matched pair bearings can be used, but these can run to several hundred $ a pair! Ouch. For VERY light setups, two normal radial bearings, NOT angular contact, can be loaded in a similar fashion, but will handle nowhere near the axial loads that a true angular contact bearing can take.

With the driven end fixed, why not REALLY fix the ballscrew in place by using another pair at the opposite end of the ballscrew? Don't do that! The far end of the ballscrew must be free to float axially. This is due to temperature variations. As the ballscrew heats up, it will expand, and it must be allowed to do so, or binding and warping will result. The non-driven end of the ballscrew (normally referred to as the simple support end) has a simple journal turned or ground, usually to a length of perhaps 1.5 X the width of the single
radial bearing used at the simple support end. When installed, you'll want to create a modest gap between the squared-off ends of the thread, and the bearing inner race. That gap will be taken up by expansion. It is possible to simply float the non-driven end of the ballscrew, but this will greatly limit the maximum speed that the ballscrew can be driven. The simple support end of an axis is fairly easy to do, so it's best not to skip that step in your project.

Even a T7 rolled screw will benefit greatly from a god set of fixed-end bearings. This is one area which beginners often skimp upon, and they then wonder why their 0.002" ballscrew performs so badly, and has 0.008" of backlash, even with a "0 backlash" ballnut installed.

So how do we eliminate backlash? First, consider again the bearing set. If not fixed axially, you will create a backlash condition when you reverse the ballscrew under load. Assuming you have a good bearing set, all remaining backlash can then be attributed to the ballnut and its interface with the machine. I am not going to go deeply into backlash and the ballnut. Very quickly, ground ballscrews are normally fitted with a zero-backlash nut. To do this, the manufacturer loads the nut with oversized balls; or, the ballnut ball tracks can be skewed slightly relative to the shaft tracks, thus loading the system. If this is your situation, you are good to go. How it is done is not so important as the fact that it is common to mount a zero-backlash ballnut on a ground ballscrew. If you are not sure if your ballnut is zero backlash, it can be tested with a very sensitive dial indicator (say 0.0005" or less). Mount a handwheel on the screw, drive the ballnut, apply the indicator, then reverse. If the indicator needle responds to the slightest reversal of rotation, then you have zero (or close to 0) backlash.

Rolled screws are tougher to deal with. Due to lead variation, if you load oversized balls, or skew the nut tracks, the system can bind as the nut travels into portions of the screw where the threads are a little closer together, or a little farther apart. A good rolled ballscrew CAN be set up in a manner similar to ground, but far more common is the use of two nuts on the same ballscrew, mounted close together, with some form of powerful spring between them, usually belleville or stout wavy washers. Manufacturers like Thomson sell assemblies that will do this, and the price is not excessive. If you can afford it, I recommend a manufactured, zero-backlash ballnut assembly for your rolled screw.

What else? There's still PITCH. What pitch to use? With some rare exceptions, such as miniature, precision instrument ballscrews, most manufacturers' finest pitch is usually 4 or 5mm, or 0.200" travel per turn. Any of these is ideal. If the pitch is finer, smaller balls must be used, and this limits the load carrying capability. Ballscrews lend themselves well to VERY coarse pitches; even a 16mm dia. ballscrew can be made with 10, 20 or more mm per turn. These are not as desireable as a finer pitch. A coarse pitch will reduce the resolution of your system, and not transfer torque to linear motion as well as a finer pitch. Unfortunately, many of the surplus ground ballscrews available use a pretty coarse pitch. Even if the price is very attractive, honestly, I recommend passing on any ballscrew with a pitch coarser than 5mm or 0.200". Metric or imperial? It shouldn't matter. A decent control can handle metric or imperial, and output whatever product you want. Likewise, the direction of rotation is irrelevant.

Lots to think about. Where are we now? We've discussed ground and rolled ballscrews, accuracies, and the need for a good bearing set. Pitch too. Let's put this knowledge to use.

Ultimately, EVERY consideration made will correlate to your desired product. Decide what you want to produce with your new CNC toy BEFORE you begin the construction or retrofit. What follows is pure opinion. If you disagree, then feel free to completely reject what I am about to say!


Situation 1: "My goal is primarily 2D routing of hardwood. I want to build a BIG machine. I really cannot see doing any metal beyond a very occasional chunk of aluminum, and even then, the metal product can be somewhat crude so long as it is shaped correctly!"

Solution: Thomson or similar rolled ballscrews. A pair of SKF 7200 series angular contact bearings loaded into a plate or shop block. The simple support end can go into a pillow block or other simple homemade bearing block. Thomson factory zero-backlash ballnut, as routers tend to have large dimensions, and the loss of 1" of travel due to the length of a double ballnut usually isn't a problem. For an axis smaller than 24" travel, a ballscrew of 5/8" diameter, this is the cheapest new ballscrew that you can buy. Make your ballscrew bearing journal 10mm for an SKF 7200 angular contact bearing set, or 12mm for a SKF 7201. For axes longer than 24", go with 3/4" or 1" diameter, and larger bearings. This setup will give you accuracies of +/- 0.004" over 1 foot. Repeatibility will be excellent. This will be a T7 accuracy system with zero backlash, and will chop wood all day long!


Situation 2: "I want to convert a Harbor Freight or Grizzly mini-mill to CNC, using steppers. I will be making simple parts for R/C, engraving plaques, and a few other things."

Solution A: If you can handle T7 accuracy, again, I'd go with rolled ballscrews. The good news here is that while 0.004" over 1 foot doesn't sound too good, the vast majority of the parts you'll make on a mini mill will be perhaps 4" long, and over 4", you can expect practical tolerances of +/- 0.0015" or so. Problem: there's not too much space in a mini-mill for mounting the ballscrews. You'll probably need 1/2" diameter ballscrews unless you don't mind hacking a lot of cast iron out of the mill to create the necessary space. Again, I firmly believe zero backlash is very desireable, especially in a metal-cutting mill. Two nuts back to back, loaded, will do it, but that will knock about 2" off of the travel! Fitting the necessary bearings will be a challenge.

Solution B: Find a set of 12mm ground ballscrews, class C5 or better, with a zero-backlash ballnut. If they come with bearing blocks, by all means use them, and you will quickly have a killer axis, very accurate, and easier to retrofit than the rolled ballscrew. It'll be quieter, too. The problem, of course, is the cost.


Situation 3: "I want to convert a Harbor Freight or Grizzly mini-mill to CNC, using servo motors. I want a fast, accurate system for creating precision parts. I need tolerances of better than 0.001"

Solution: Now you really need C5 or better ground ballscrews, and if you can retrofit the dovetail slides with recirculating rails and trucks, like THK HSR12's, then that would certainly help, but there's a lot of work involved there. If you are going with a fairly expensive CNC control, and some nice servos, it really makes no sense to interface such a system with set of rolled ballscrews + sloppy bearings.


Situation 4: "My mill-drill will be converted to CNC. I can live with a modest amount of backlash, and 0.004" accuracy is no problem, so long as repeatibility is good."

Solution: Mill drills are heavier machines, and you will be asking your servos or steppers to move a fairly significant mass. This will call for powerful motors, and likewise, a stout set of ballscrews and bearings. With a limited accuracy requirement, I'd go with 3/4" dia or larger T7 rolled ballscrews and a fairly stout bearing set, at least SKF 7201 bearings, which are 12mm ID X 32mm OD. Larger bearings would be appropriate. When snugging the typical way gibs on a mill drill, your torque requirements will go way up. Since a small amount of backlash is OK, you can use a single standard Thomson-style ballnut, which will probably have about 0.005" of backlash; this will create a compact and low-profile interface between ballscrew and way. Repeatibility will be excellent. This is inherent so long as the ballscrews are truly fixed axially.


Situation 5: "I am scratch-building a small CNC bench mill for machining casting waxes and light metal work. Accuracy is important. I am going to machine jewelry prototypes, small components for turbine engines and R/C, and other small parts with close tolerances."

Solution: This will require a very accurate and tight system with 0 backlash. As the size of the parts go down, the need for zero backlash and a quality fixed bearing set go way up. 0.005" of backlash would ruin a fine filligree in wax for gold casting, or a turbine diffuser. I'd go with C3 or better ground ballscrews, fine pitch, servomotors, direct drive, THK/NSK linear rails and trucks. A commercial bearing block would ensure success, or you can create blocks on your own, but they'll need to be well-made. You'll also need a fast, high-quality spindle, but that is another topic entirely.

I found the information provided to be quite enlightening, and I appreciate your time. Additionally, I stumbled upon a helpful webpage that outlines the process of identifying thread screws.fmwhatsapp

[cuixiansheng]

Hello, we are a company in China that specializes in THK, IKO, NSK, NB, SBC slide rails and ball screws. If you need it, please contact us and we will give you high-quality products at the lowest price.email [email protected], WeChat:18033146782, WhatsApp：56948621436，thank you