Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
I compared the original Prusa bearings to the Misumi LMU8 and the new Vesconite polymer bushings.
The original bearings were installed with the original oil and ran for 16 days. After that, I wanted to decrease the noise and ordered polymer bushings from vesconite. They were much more quiet, but had a huge play compared to the original bearings. Since I already disassembled everything, I also installed the much praised Misumi LMU8 bearings for comparison.
You can see my findings here in these short videos. I tried to rotate the Y-slider with roughly the same force every time.
Prusa original bearings (oiled):
They run without any noticeable play. The only apparent play comes from the elasticity of the rods and the other structural parts. They run very smooth with little friction due to the low viscosity of the oil inside
Misumi LMU8 bearings (greased):
They feel exactly the same as the original Prusa bearings, maybe a little bit smoother, but not much. I filled them with grease as recommended by the internet. Due to the higer viscosity of the grease, they run a little harder, but might last longer.
Vesconite polymer bushings:
They run very quiet, but have some play. The play feels like I can rotate the slider freely between two endstops, so it is definately not the elasticity of the rods as before. I used the more expensive pink version (LM8UU style, OD15 x ID8 x 24mm long, Superlube HS3926.9090)
I also tried to adjust the tightness of the nuts on the U-bolts. I started very loose (where I had a lot of play) and slowly tightened them down evenly. At a certain point the friction increased dramatically, but there was still significant play.
I don't know how the play will influence the printing results, but I think that less play is always better.
I will leave the Misumi LMU8 bearings installed since I alreads have them. If I had known these results before, I would have stayed with the original bearings, as I don't see any significant improvement of the Misumi over the original ones. The results might have been different after more than 16 days of print time.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Your Vesconite results are appear about the same experience I had with Drylins. They are working fine on my X and Z rods, but on the Y there was just too much play. One thing I noticed was that the play appeared worse when I had everything bolted down and printing. I'm guessing the extra weight gave it more mass and when the Y would start and stop moving it seemed more jerky than with the stock bearings. I tried all kinds of adjustments and using various compression holders, but either they were tight enough to cause movement failures or loose enough for the carriage to jerk about. The results when printing was the worst Y ghosting I've yet seen (on a 20x20 cube there was close to 10mm of Y ghosting!).
I cleaned out my stock bearings (filled a ziploc with enough 99% IPA to submerge the bearings, sealed it, and ran it through 8 4 minute cycles in an ultrasonic jewelry cleaner) and then repacked them with white lithium grease. They are messy (which is what I hopped to avoid with the Drylins), but the Y carriage is now super quiet and smooth.
On the Misumi bearings I thought the thing with them was also getting matched rods too?
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
On the Misumi bearings I thought the thing with them was also getting matched rods too?
I thought that was the case if someone had the "old" rods from prusa before they switched to the hardened ones?
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Are those the Hilube or Superlube (pink) ones on the bed? I've got bed play as well with the latter, but I can't see any on the extruder where I have the Hilube (white) ones
I have not noticed an impact in print quality; I suspect the bed has enough weight on it that it doesn't lift up and down. spreading the rods slightly to pretension them would also help reduce play.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Chris Warkocki said that he contacted Misumi after he noticed some rattling with the Misumi bearings. According to Chris, Misumi said that for LMU8 bearings you need matching rods.
I ended up biting the bullet and paid $80 (shipping included) for the matching rods for my X and Y axis. I had noticed that there was an increase in resistance but not in a negative way. They felt more firm on the matching rods kind of like how the Prusa bearings feel on their rods.
I do not feel the value with Misumi bearings and rods after spending nearly $120.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
I ended up biting the bullet and paid $80 (shipping included) for the matching rods for my X and Y axis. I had noticed that there was an increase in resistance but not in a negative way. They felt more firm on the matching rods kind of like how the Prusa bearings feel on their rods.
Did you pack the Misumi's with grease before installing them onto the rods? If you did, what grease did you use?
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Are those the Hilube or Superlube (pink) ones on the bed?
Pink superlube, as written above the video
I suspect the bed has enough weight on it that it doesn't lift up and down.
I also don't think it will lift up and down, but it will try to rotate if the belt is placed in the middle and the center of gravity is placed outside of the middle due to objects printed on the side of the plate. That's why I try to rotate it in the video.
spreading the rods slightly to pretension them would also help reduce play.
Of course it would, but this would also increase friction and wear. And a little wear will then lead to play because I suspect that the pre-tensioned rods will not distribute the force evenly in the bushings, but might touch them only in one spot. Since I don't want to exchange the bearings on a regular basis, I chose to go the safe route.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
According to Chris, Misumi said that for LMU8 bearings you need matching rods.
Who says that the stock rods aren't matching? On my rods, the LMU8 bearings slide perfectly smooth, no play, no binding, no noise. So I'd say it's a match.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Who says that the stock rods aren't matching? On my rods, the LMU8 bearings slide perfectly smooth, no play, no binding, no noise. So I'd say it's a match.
My stock rods weren't all quite the same diameter. On some even the original bearings were tight, on others loose. Going to Misumi rods solved that.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
I ended up biting the bullet and paid $80 (shipping included) for the matching rods for my X and Y axis. I had noticed that there was an increase in resistance but not in a negative way. They felt more firm on the matching rods kind of like how the Prusa bearings feel on their rods.
Did you pack the Misumi's with grease before installing them onto the rods? If you did, what grease did you use?
No, although thick grease is recommended for the bearings I did not pack the bearings with a thick grease because using these bearings in 3d printers will not get the grease hot enough to provide the proper lubrication. These bearings were intended to be used in industrial applications where the bearings would traverse miles of smooth rod at a time. I used a high viscosity oil from SuperLube.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
According to Chris, Misumi said that for LMU8 bearings you need matching rods.
Who says that the stock rods aren't matching? On my rods, the LMU8 bearings slide perfectly smooth, no play, no binding, no noise. So I'd say it's a match.
This is 3rd hand information that I got from one of Chris Warkochi's YouTube. He had contacted Misumi and Misumi said that LM8U rods were recommended with LM8U bearings and LM8UU rods were recommended with LM8UU bearings. I know the U only means that it has one seal vs 2 seals but I can tell you that from my personal experience the Misumi rods matched with the bearings feel completely different from the Misumi bearings on stock rods. The Misumi bearings on the stock rods were too smooth, almost no resistance, and you want some resistance with your bearings or you'll get slop.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
No, although thick grease is recommended for the bearings I did not pack the bearings with a thick grease because using these bearings in 3d printers will not get the grease hot enough to provide the proper lubrication. These bearings were intended to be used in industrial applications where the bearings would traverse miles of smooth rod at a time. I used a high viscosity oil from SuperLube.
I wanted to expand on my post and I can't edit it because you have a short window to edit posts on these forums. Since the bearings wont reach the intended temp the grease will end up clogging the bearings and they'll end up gliding instead of rolling.
I used this oil: https://www.amazon.com/gp/product/B000UKUHXK/ref=ppx_yo_dt_b_asin_title_o04_s00?ie=UTF8&psc=1
With this oiler bottle placing 3 drops on each row of balls: https://www.amazon.com/gp/product/B00EXP082S/ref=ppx_yo_dt_b_asin_title_o03_s00?ie=UTF8&psc=1
I degreased the bearings with: https://www.amazon.com/gp/product/B0009W4MX0/ref=ppx_yo_dt_b_asin_title_o04_s00?ie=UTF8&psc=1
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
I can tell you that from my personal experience the Misumi rods matched with the bearings feel completely different from the Misumi bearings on stock rods.
Since prusa doesn't give any specification on their rods, it is quite possible that they change them quite often. So maybe you got completely different rods than me. I'm going to measure mine to see which diameter they have.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Something that occurred to me is this may not be a completely valid comparison. The bearings have seals on each end which can impact the appearance of slop, since those are typically rubber coated brass and quite stiff. The vesconite bushings do not have a seal.
Unfortunately, a better comparison would require sacrificing bearings by trimming back its seal to see whether there is slop in the actual bearing elements that is being masked by said seals. (They cannot be removed completely as that would make the innards fall out, and is usually destructive of the seal)
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
I can tell you that from my personal experience the Misumi rods matched with the bearings feel completely different from the Misumi bearings on stock rods.
Since prusa doesn't give any specification on their rods, it is quite possible that they change them quite often. So maybe you got completely different rods than me. I'm going to measure mine to see which diameter they have.
they *should* measure somewhere between 7.95 and 8.00 mm, if memory serves. (This is from a conversation I had with Juan at Vesconite on said slop. Their bushings are machined to precisely 8.05mm, and so can be a bit loose if your rods measure on the lower end of the range.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Well, the names are LMU8 for Misumi bearings and LM-8UU for IKO bearings. The Misumi LMU8's are specified for any axle with 8mm outer diameter with a g6 tolerance - that means diameters from 7.996mm to 7.986mm. The original Misumi rods meet that spec, of course. IKO has the same requirements in their LM-8UU (Japan) model. There are no specific LM8U or LM8UU rods. But if you get quality bearings made to specific tolerances, it makes sense getting quality rods for them, made to matching tolerances, too.
Pretty much all the LM8UU's on the market today, including the ones Prusa ships, are chinese clones of IKO with varying quality. And the accompanying rods vary in exactly the same way. In the end, when you use the cheap LM8UU's, you get what you pay for and you need to make sure to find which rod matches which bearing when you get them by trying them out. Even in a set of a single supplier.
References:
- IKO https://www.ikont.com/catalogs/needle-roller-bearing-series/2936E-1.pdf
- Misumi https://us.misumi-ec.com/pdf/fa/2012/p1_0265.pdf
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Apologies to Tim for going off the deep end again 😉 :
I have spent quite some time researching linear roller bearing lubrication. Even went so far as to consult all the graphs and tables to match the lubricant to the conditions on a 3D printer - next to no load, medium to high speeds, temperature range from room temp to mildly elevated.
Before I go into what came out, let me talk a bit about greases. A grease has four components:
- Base oil. Typically either high grade mineral oil or a synthetic (ester) oil.
- Gelling agent. Typically some form of soap, like lithium or aluminum soap. Possibly silicone gel.
- Solid particles. Graphite ([C]ₙ), Molybdenum sulfide (MoS₂), or PTFE ([C₂F₄]ₙ)
- Extreme pressure additives.
The base oil is what provides rolling lubrication. It needs to have low enough viscosity to be able to flow and spread over the ball surface quick enough. This depends on the expected speeds and ball diameters. LMU8's have tiny balls and need a low viscosity base oil. Viscosities below 200 mm²/s seem appropriate.
The gelling agent is there to prevent dripping. It provides a molecular matrix that turns the oil into a gel, a non-newtonian liquid that has different properties when idle (behaving almost like a solid, not flowing) and when under shear stress when the balls roll over it when it behaves very liquid and low viscosity. The thickness is measured in a NLGI consistency numbers. NLGI 000 is almost liquid, NLGI 6 is solid like cheese. Thickness of NLGI 0 and lower is appropriate for LMU8's because higher numbers (most greases are 2-4) will make the balls not move well in the recirculating raceways, increasing a risk of sliding instead of rolling. Aluminum or Lithium soap are the soaps of choice for linear bearings - compatible with many materials and not expensive. Aluminum soap's advantage is resistance to washout, not needed on a printer.
The solid particles are there to aid sliding. They're added to greases intended for bushings and low speed sliding elements, like pistons on heavy machinery. So MoS₂ and PTFE is not recommended for linear roller bearings, because they increase the risk of sliding.
Extreme Pressure additives or EP are not needed. They normally prevent the grease to deteriorate under huge pressures. No such conditions on a printer. EP's don't hurt per se, although they can be aggressive and compatibility with the rod material would need to be checked. Typical chrome steel and stainless steels are OK.
In the end, the ideal is a Semi-fluid, NLGI 0 or lower, Lithium grease with low viscosity base oil.
SuperLube grease (NLGI 2-3 synthetic with PTFE particles) and SuperLube oil (synthetic ester high-viscosity oil with PTFE) will work, but according to my best understanding of the matter, will have increased risk of sliding and scoring the rods and the balls. PTFE powder in the grease may make cheap low-precision LM8UU's roll smoother, though.
I'm myself using Zeller+Gmelin Divinol Lithogrease 000/150, specifically developed for linear guides. 2000+ hours on the printer with no wear at all and absolutely smooth movement on Misumi LMU8's with Misumi rods.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
The Misumi LMU8's are specified for any axle with 8mm outer diameter with a g6 tolerance - that means diameters from 7.996mm to 7.986mm.
I just measured all of my rods, they are all between 7,988 and 7,996, so it looks like the meet the g6 tolerance.
These are the raw values:
right Y-axis: 7,9946 - 7,9956
left Y-axis: 7,9944 - 7,9950
right z-axis: 7,9900 - 7,9904
left z-axis: 7,9878 -7,9881
lower x-axis: 7,9944 - 7,9949
upper x-axis: 7,9928 - 7,9931
Each rod was measured 3 times on slightly different locations a few mm away from each other, always on one side of the rod. The lowest and largest value was recorded. However, the last digit shouldn't be trusted, as I measured it by hand in a non climate-controlled room. Relative accruracy should be high enough, as I usually get the same value +- 100 nm if I measure the same spot twice.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Those look like good rods indeed.
RE: Comparison of original bearings, Misumi LMU8 and Vesconite polymer bushings
Thickness of NLGI 0 and lower is appropriate for LMU8's because higher numbers (most greases are 2-4) will make the balls not move well in the recirculating raceways, increasing a risk of sliding instead of rolling
This makes a lot of sense. Is there any way to find out if the balls are rolling or sliding? Due to the total absence of noise while moving the bearing after greasing it, is suspect that they are sliding. If I increase the pressure, I can feel some mechanical motion inside the bearings, so I suspect that they start to roll under a little bit of pressure.