SBR linear rail on LR3

So I just pulled the trigger to order an SBR10C linear rail from China. My intent will be to use it to replace the 1/2" conduit Y rail on the LR3.

I ordered the 10mm version, thinking that it certainly doesn’t need to be super beefy to replace the 1/2" conduit, given how noodly the conduit is, and that it’s really not a problem of strength.

Next step will be figuring out how to mount to the YZ plates.

My original idea when I put that rail into my Ali Express cart a couple months ago was to use 2 of the linear bearing blocks and replace the extant rail riders with those, but this seems like it may be over constrained.

It’s OK jf I assume that the YZ plates will always be square wrt the X beam, or that the beam can and should be pulled square to the YZ plate, but that seems to be maybe a little wishful in real life. It also means serious twisting force against the beam and/or YZ plate should one Y motor or the wiring have trouble. Seems like bad news…


My updated plan is to have only one bearing block on the rail, and give it a modicum of rotational freedom. This will allow some degree of freedom in the Y stops as well, allowing there to be a bit of adjustment for the stops being a bit out of square, as mine are currently. This also allows me to place the bearing just to the front of the XZ plate, to keep it out of the way of the Z axis.

Some concerns:

  • Will the one bearing allow or partially allow the beam to twist, giving some non linearity to the X axis? I think Ive got it covered if the bearing is close to the router center, so forces on the endmill have as little leverage vertically as possible, but… what do you all think?

  • Will a single bearing wear too quickly under the mass of beam and potentially spindle?

  • How much rotarional freedom is reasonable? I don’t plan on much but want to at least have 15mm or so of Y stop adjustment not stress the bearings or YZ plates.

1 Like

I am glad you are testing it, I have VEVOR Linearschiene SBR20-1200, Linearlager-Gleitset mit 2 Schienenwellen von 1200 mm und 4 SBR20UU-Blöcken, Linearlagerschiene mit Vorgebohrten Löchern, Führungsschienenset für Fräsmaschinen  | VEVOR DE in my sight for ages. :smiley:


I wish there was some easy way to make a guide or something to install one of these on both sides and keep them perfectly parallel . That would sure add some more rigidity by locking the machine down. I never take either of mine off the table so I don’t need the ability to remove like we have now.


@Tokoloshe I chose the lighter SBR10 because there isn’t much space there, and the SBR10 should only be 36mm tall to the top of the bearing surface. Also, given that the current solution only rides atop the rail, I didn’t feel the extra physical strength of the larger was necessary. I will admit though that I spent a lot of time waffling lookjng at the 12 and 16mm versions, but ultimately, the 10 is what I bought.

@Jonathjon well making a jig to set the rails parallel isn’t that hard, really, and we do it with the Primo. It actually wouldn’t be too difficult to have a little expansion accomodation on kne side that could allow for a small change. Probably a small amount of flex in the YZ plates would be enough to allow a bit of imperfection, though that then becomes less predictable for which side the machine follows… which is why I want to allow for some degree of rotational freedom in the mount design.

I’ve been trying to find a good dimensional drawing, and this is one area where the SBR10 gets tricky, as almost every dimension chart I have found starts at the SBR12 size. :man_facepalming:

I think I have the info that I need, now though. The height of the bearing surface is 36mm from the table, and it looks like 4 M5 screws at 25×20mm spacing.

I am tempted to use a 608 skate bearing as a pivot, with the bearing body and a 5/16" bolt screwed down to the bearing block, and the bolt and a nut holding the Y plate in place… I’ll see if I can get a concept drafted up today. I am a bit concerned about the bulk, but it would satisfy the requirement of holding the machine parallel to the rail while allowing it to have some pivot.

This may be overthinking things, and the bearing is likely unnecessary. Probably just the bolt would allow sufficient movement, it’s only a couple of degrees, and not constantly flexing.

Of course a solution that uses 2 bearing blocks gets a bit trickier, but I should probably work something out, in case.


To borrow Doug’s phrase: Following with interest. :smiley:


With two bearing blocks, could your intended setup be mounted onto vertical surface, or at least semi-vertical surface to help reduce footprint/space needed?

You are currently only running the rail down the one side correct? Just replacing the emt and leaving the other side riding on the table with the 608 bearings? Just making sure I’m following along and understanding correctly.

1 Like

Possibly. The rail is/will be fully captured. Though I would not try it fully vertical, it could be a pretty steep slope, and would be perfectly fine.

Well, mine is still using the 60mm skate wheels but otherwise, yes. At a bit over $100 each for 2m of linear rail, plus not wanting to over constrain the machine, one rail is perfectly adequate. It will constrain the Y axis to being parallel with the rail.

Removing the machine from the table may still be possible, but considering what happens to those linear bearings when they are off of the rails means that practically speaking, I will no longer want to do that. I may stull want to use the table as an assembly tabe though so I will retain the easy removal of the belt, at least on the wheel side.


Did you manage to di this?

Sadly, my linear rail is still sitting beside the LR3, so no. I haven’t done it. Still kind of waiting to see what the LR4 will do, but mostly I need to do something about the table first…

I’m late to the party, but have added my name to the “following with interest” category. :smiley:

Okay, since people have an interest…

The holdup is this: using the fully captured rail, I would rather not have the YZ plate fully constrained. To that end, I need a mechanism that will allow some moderate pivot for homing and a bit of XY skew allowance to take place. This will be necessary on the case where the Y end stops are not perfectly square, so the beam needs to be able to move freely to the endstops, and then adjust to a proper square position.

Of course too much freedom of movement negates the whole advantage of a captured rail in the first place, so I need to decide on an an acceptable swing angle.

It’s a bit of a pain, and I haven’t had a lot of time lately to work on my machines.

Step 1: get an accurate real-world model of the SBR10 rail and the slide bearings.

Step 2: decide on a pivot mechanism. I think I can leave the Z axis fully constrained to the rail, though some variation in Z height between Z1 and Z2 is almost guaranteed at least at power down, the span of the machine ought to keep this within the SBR10 limits on rotation. The XY plane needs to allow a pivot. Current plan is to use 2 bearings, keep the “toe” side at a consistent distance, and allow it to rotate. The “heel” bearing will allow an arc movement on XY but keep Z height constant. The alternative is to use only one bearing car, but this could allow torque to affect the beam in the YZ plane, particularly since I need to allow movement in the XY plane. The needs here are constrain the YZ plate in the YZ and XZ planes while allowing some controlled movement in the XY plane. As a bonus, a “quick release” to remove the YZ plate from the rail would be nice.

All of this though has to wait for me to fix my table. The wings I added to convert the LR2 to a LR3 have let me downand are no longer serviceable, leaving me with an effective working area of about 16" of Y to 40" of X beyond that 16" the spoilboard is too far from level now.

I think I’ll do a temporary tabletop of 3/4" MDF but that waits for the extra “play” funds for a sheet, and from there I can build a torsion box underneath it.

1 Like

If u change language to “suomi” you also find .step files

1 Like

I’ve found documents like that, but none seem to measure the rail with the bearings. The ones that measure the bearings are inconsistent, so I figured it would be better to measure what I’ve got anyway.

That’s a quick snapshot of the rail, I dont remember if I got 1.8m or 2m (I think 2m).

I just have to find where I put my calipers, then pull in CAD of the YZ plates and make rail riders that will work. I dont think I need more than a couple of degrees freedom in XY, but I’ll build for more than I will ever need. I’d love to build for a huge pivot, but of course I lose sin(θ) in usable X, so more than 10° swing isn’t happening. 5° either way is more than plenty anyway, and more than we would see with the current setup not having trouble, I think. Pretty sure Ryan didn’t think the rail needed that much freedom, and neither do I, but in this case because I am dealing with a fully captured rail, I do think it requires a more complex solution.

The more I deal with this, the more I appreciate the Primo solution to captured rails, and I sort of wish I could just do the same, but in order to keep the YZ plane stable, I’m convinced that I need 2 points of reference on the rail.

I did have a thought… mount a rail truck inside the Z motor carrier, and use mirrored wheel prints for the YZ plate. Have the wheels on the table provide the YZ reference, and the rail only acting on X. Seems kind of a step backwards though, and removes many of the benefits of a captured rail in the first place. I supose I could use one wheel and one bearing truck though… if the o e truck gets too complicated, I might do it that way…