MGN12 from Alliexpress

Hello,
Yesterday I got my MGN12 linear rails from Alliexpress. I would not order them again but now it is too late. The overall quality looks great but the ease of the carrier is really bad. I have the same problem showed in this video.
[www.youtube.com]
After some search I found a second video.
[www.youtube.com]
So the balls is the only thing I could change the diameter of the balls I measured is about ~2.4mm but I have no precision measurement. So I think the diameter is 2.381 because that is a standard I could get from some shops.
I think I will try that and I hope that the jerk on the movement will be better after that rework.
What do you think?
Thanks kind regards
Mathias

Sometimes these Chinese rails come filled with some sort of "preservative" grease, to prevent them from rusting in transit or whatever. This grease is waaay too thick to be used in linear rails and should be replaced with proper one. But seeing that you took out the bearings and measured them, I suppose you would have seen if the grease was a problem The other thing that could help the carriages ride a bit smoother is to mount the rails on a flat and sturdy surface (e.g. Aluminum extrusion), so as to take out any bends that might be in the rail...

Can't really comment on the ball size, but I would think replacing them is a bad idea. Usually the Chinese stuff is not made to tight tolerances, and because of that each individual carriage and rail are paired together for a somewhat better fit. That is why you can't just butt two (Chinese) rails together and expect smooth movement along both of them...

Okay I have seen that but I wasn't sure what kind of cleaner I should use.
I will have a look on that. But my rails and carriage were not paired. I got 6x800cm rail and 6xMGN12 carriage and the were packed separated. I will try to pair them later.

Thanks for the heads up on the rails, think i'll give them a miss in future, You can use lighter fuel to clean grease off, I got a small IKO rail the other week, but while trying to hold the phone & carriage/rail...it slipped and popped a ball or 2 out I couldnt find them, I guessed they where 1mm bought some more, when they came, wrong size...so I took one of the balls out the LMU8 and in it went, still a little gap, put another in....and now its back to being nice n smooth.

I buy linear guides via eBay, usually from scraped industrial machines. The last three I bought were $35 each including shipping and one of the three was unused old stock. All the ones I have purchased so far have been Japanese made and extremely high quality.

I would not attempt to replace the balls in a linear guide. They are an integral part of the design and the rails should be ground to match the ball diameter. Matching the diameter with randomly purchased balls is extremely unlikely.

---

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

I would consider lapping the rails and possibly the block if you are getting that kind of stiction.

If you've never done this before, look up lapping - it's not a difficult process; but there are a few fine points (er, no pun intended) to be aware of. The youtube videos on the subject are helpful.

You are going to want a very fine lapping compound; and a soft lap; perhaps brass.

DON"T OVERDO IT!!!!

The objective is to just knock down the high spots that are binding up the block on the rail; you don't want to knock the rail out-of-tolerance.

And when you are done, be sure to THOROUGHLY clean the block and rail - you don't want any of the lapping compound to remain, and keep abrading away the metal while the rail is in use.

I wouldn't even consider lapping. If the things don't work right direct from the manufacturer, chalk it up to experience and buy a quality brand next time. As soon as you modify the guide by either trying to change the balls out lap the rails it is anyone's guess how the thing will perform. The whole point of buying linear guides is to eliminate slop in bearings and making changes is liable to reintroduce the very slop you thought you were getting rid of by buying linear guides.

---

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

Installing balls randomly picked from a big box, lapping the same, is how they build these rails in China. You would be extremely lucky to improve them by working the same way

Quote

If the things don't work right direct from the manufacturer, chalk it up to experience and buy a quality brand next time.

With all due respect...

I figured that he already didn't have much to loose, considering that they are already kinda' crusty... and "chalking it up to experience" doesn't do anything for the rails he's got in his hand now.

Also, it's a 3D printer - not a mill or a lathe. It doesn't need to hold super-tight tolerances. It's more important that it be free-moving; even slop of 0.001 or 0.002 inches (a lot for that kind of rail - and I wouldn't lap them out that far anyways) would be OK. (and by slop I mean looseness of the block on the rail; not the straightness accuracy of the rail; that's a different measurement).

_____________________________________________________________________

Quote

You would be extremely lucky to improve them by working the same way

Again, with all due respect... this is simply wrong.

Historically, hand-scraping and lapping is how all the best precision machine tools were made (and to an extent, still are); machine grinding has only taken over for reasons of industrial expediency. It is entirely possible to start with three blocks of a suitable rock, and thru hand working techniques alone work your way up to a highly precise machine tool - this is, in fact, how it was first accomplished, before there were surface plates, surface grinders, and the like.

The Chinese rails are not hand-lapped in to the required tolerances; if they were they would run quite smooth.

There are several possible causes for binding:

1) Contamination of the block and/or rail. I presume that Mathias has already cleaned them.

2) Maladjustment of the block. This would show up as general stiffness of the block on the rail, and wouldn't have any free areas. However, not all blocks are adjustable (actually, all of the ones I have here don't seem to be adjustable at all).

3) An out-of-tolerance runner in the block. Depending on how it's out-of-tolerance, this will show up as either stiffness all along the rail, or stiffness that repeats with a periodicity determined by the circumference of the balls (as the balls roll over the high spot on the runner, they bind up, and when a gap goes by they loosen up). Sometimes, you can get at the parts in question and fix the issue; sometimes you can't. The youtube videos did not seem to show this kind of binding; but without it in my hand I couldn't be sure.

4) An out of tolerance ball. This will show up as binding that lasts the length of the block, and repeats with a periodicity of (IIRC) about twice the block length; as the ball in question recirculates around and comes into the bearing area and out again. If the ball is out of round, sometimes the binding might go away as the angle of the ball changes, or sometimes it might repeat with behavior similar to a high-spot on the runner; but then go away when it rotates back out of the bearing area and recirculates back around to the other side of the block. If you can get the balls out of the block, you can measure them with a micrometer to make sure that they are all in tolerance. To my eye, the video didn't seem to indicate this; but again, I don't have it in my hand; so it could still be a possibility.

5) One or more high-spots on the rail. This is what lapping the rail would address. This would show up as stiffness at that specific location; and if you remove the block from the rail and move the balls about it will stay stiff in the same location. With a couple of ground gauge pins of the appropriate size, and a good micrometer (or even a so-so micrometer, if the error is large enough) it's possible to measure this by hand (tricky, as it kinda' helps to have four hands, but possible) - what you do is straddle the gauge pins in the ball race on the rail, and measure the width across them (somewhat similar to using thread wires); and take several measurements all along the rail (paying attention to areas that are loose, and tight).

For lapping the rail, you wouldn't run the lap across the whole rail; just the stiff spots. This will remove a small amount of metal from the ball race where it is too wide; allowing the block to travel freely across that portion. Doing this will have consequences for the straightness of the rail; however given the application (a 3D printer) and original condition of the rails (already of questionable quality), it should not be an issue.

Again - remember that this is a suggestion to make a more-or-less unserviceable rail work for the specific application. I am NOT suggesting that everybody just buy cheap rails and slap some lapping compound on them instead of a good high-quality rail system (although, if the price is right, and you don't have anything better to do with your time... ).

Of course lapping is a way to improve and "salvage" these rails but frankly, someone who knows about it and above all, knows how to do it properly wouldn't ask.
What about lapping the balls ?

Everybody has to start somewhere. For this application, it doesn't have to be done "properly" - just "good enough".

Lapping a ball is a bit more challenging, because of the geometry involved - you need a spherical concave lap of the size to fit the ball. If you have a lathe with a suitable sphere turning attachment, you can, in principal, make one. Casting one against a good ball might also work; depending on what material you are casting with. Also, you will need a way to hold the ball being lapped.

However, if it's just one or two balls that are out-of-tolerance, I'd probably just replace them; even if you can only get a ball that is somewhat undersized the rail will still work OK. Personally, if it's a majority of balls that are bad; and I couldn't easily get balls of the required size, I'd either (if possible) enlarge block by lapping down the the runner to accept the next size up of balls, or scrap the whole thing as unworkable; depending on how much money I had in it, how much it would cost to get the balls (prices for that stuff varies depending on where you are - are you next door to a good bearing supply house... or in the middle of the African outback?); how much money you have available for your project, and how much effort you want to go to.

If this was anything else and didn't work properly from the moment it was unwrapped, you'd send it back and complain. This is a part that is supposed to be precisely made and it isn't doing what it is supposed to. I wouldn't waste one minute more on it. Send it back and buy a better brand next time. I have found NSK guides to be very well made, and you can buy used ones for less than you paid for the piece of crap you got.

One of life's great, tragic truths is that quality costs. The sooner you face it, the less time you will waste trying to make poor quality things do things they aren't capable of doing.

---

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

Um... can he send it back for a refund? My starting assumption was that that was not a possibility...

Quote

One of life's great, tragic truths is that quality costs. The sooner you face it, the less time you will waste trying to make poor quality things do things they aren't capable of doing.

My only disagreements with that is that some people have more time than money, and therefore must make-due with suboptimal solutions; and sometimes (but, not always!) the point of the operation is not productivity (i.e, a hobby, or learning experience). Under those kind of circumstances, it's not a waste of time, if you either learn something, or can get it working.

If however, the point is making money and/or being highly productive, a cost/benefit analysis will often support buying the good stuff (there is a reason, after all, why high-end tool and machine manufacturers stay in business!).

But - yeah, quality is going to cost you something - time, money, effort... or more likely, some of each!

Quote
bdurbrow
Everybody has to start somewhere. For this application, it doesn't have to be done "properly" - just "good enough".

Lapping a ball is a bit more challenging, because of the geometry involved - you need a spherical concave lap of the size to fit the ball. If you have a lathe with a suitable sphere turning attachment, you can, in principal, make one. Casting one against a good ball might also work; depending on what material you are casting with. Also, you will need a way to hold the ball being lapped.

However, if it's just one or two balls that are out-of-tolerance, I'd probably just replace them; even if you can only get a ball that is somewhat undersized the rail will still work OK. Personally, if it's a majority of balls that are bad; and I couldn't easily get balls of the required size, I'd either (if possible) enlarge block by lapping down the the runner to accept the next size up of balls, or scrap the whole thing as unworkable; depending on how much money I had in it, how much it would cost to get the balls (prices for that stuff varies depending on where you are - are you next door to a good bearing supply house... or in the middle of the African outback?); how much money you have available for your project, and how much effort you want to go to.

I was kidding about "lapping balls" .

Quote
the_digital_dentist
If this was anything else and didn't work properly from the moment it was unwrapped, you'd send it back and complain. This is a part that is supposed to be precisely made and it isn't doing what it is supposed to. I wouldn't waste one minute more on it. Send it back and buy a better brand next time. I have found NSK guides to be very well made, and you can buy used ones for less than you paid for the piece of crap you got.

One of life's great, tragic truths is that quality costs. The sooner you face it, the less time you will waste trying to make poor quality things do things they aren't capable of doing.

The problem is that you never send back to CHina as it will cost you more than taking your loss. I wonder how long this "free shipping from China" will last as in fact it costs a lot to the local Post Office.
To give bad notes and inform potential customers is the way to go. Now, people have to understand they get what they pay for although these stuff from China are average but dirt cheap. So as long as you don't get dirt.

Mine were great, just saying...