Quieting linear bearings?

Skip to first unread message

Michael Cook

Nov 22, 2012, 7:25:15 PM11/22/12
to make...@googlegroups.com
I've been doing some upgrades on my Thing-o-Matic recently, and I switched to using linear bearings instead of the bronze bushings. I've never used linear bearings before, and now the movement is incredibly smooth.

What I didn't expect is that the machine is quite a bit nosier now. It sounds like it's from the balls rattling inside the bearings as the X and Y axises shake. Is there anything I could do that would cut down on the noise? I haven't really done anything to lubricate them (I just applied some light oil to the rods as I always have), would putting some kind of lubricant inside help?

Thanks for any tips.


Nov 22, 2012, 10:56:18 PM11/22/12
to make...@googlegroups.com
I think the bearings want a light grease, not oil.  Grease will almost certainly quiet the bearings down some.


Nov 22, 2012, 11:11:58 PM11/22/12
to MakerBot Operators
The problem is the rods. Linear bearings require 2 things:
Hardened rod (surface hardening) since you are rolling hard steel
acros it and will eventually show serious wear.
Extremely smooth rods, typcially hard chrome plated.

In a bushing setup, we can get away with non-hardened rods as we use
soft bushing material, but anytime you have a harder metal, even in
rolling contact, it will destroy the surface in short order.
Bascially, what happens is microwelds whwre the ball picks up the
metal from the soft surface of the rod.

Trust me on, this, I saw with in a Cupcake with Aaron Double's X-Y
rider with 623 hard surface bearings rolling on the stock Cupcake rods
and within 2 days the rods had greves in them even with clear silicone
grease from day 1.


Nov 22, 2012, 11:17:04 PM11/22/12
to MakerBot Operators
Sorry for the typos, it appears that turkey and the Tryptophan are
getting to me . Happy Thanksgiving!
> > Thanks for any tips.- Hide quoted text -
> - Show quoted text -

Michael Cook

Nov 23, 2012, 12:46:49 PM11/23/12
to make...@googlegroups.com
Thanks. Any tips on how to apply it? I'm assuming I don't want to over do it.


Nov 23, 2012, 10:24:28 PM11/23/12
to make...@googlegroups.com
You must remove the shafts from the bearings since they have seals,put grease inside, then slide the rod back in to keepthe grease inside between the front and back seal. Putting stuff on the shaft does nothing,the seals keep the diret out and the grease in!
Personally, I use this as it's clear and silicone based wich is much cleaner in the long run http://www.super-lube.com/synthetic-multipurpose-grease-ezp-49.html 

Michael Cook

Nov 24, 2012, 9:30:30 AM11/24/12
to make...@googlegroups.com
I figured I'd have to take them off the rods to do it. Thanks.


Nov 24, 2012, 12:11:02 PM11/24/12
to make...@googlegroups.com
I'm not sure what linear bearings you guys are using but mine do not appear to have a seal.  However, they are kind of noisy.  

I've got these ones and even though it shows the steel retainer one in the photo I've got the resin retainer version.

I guess what I'm asking is since my bearings don't seem to have seals could I just lightly grease the rods for the same effect without disassembling anything?


Nov 24, 2012, 12:29:55 PM11/24/12
to make...@googlegroups.com
Also what type of grease is recommended for linear bearings?


Nov 24, 2012, 1:15:56 PM11/24/12
to MakerBot Operators
Personally, I use this as it's clear and silicone based wich is much
cleaner in the long run

g. wygonik

Nov 24, 2012, 1:27:58 PM11/24/12
to make...@googlegroups.com
I got 3 of these for my X-carriage: http://www.vxb.com/page/bearings/PROD/3-8inchLinearMotionSystems/Kit7960 -- which are the same dimensionally as the ones you linked, but with a steel retainer (the ones you linked say in the description they have a resin retainer, though they do seem to use the same photo).

They were a little noisy at first, but I put a few drops of 3-in-1 oil on the rods and now they are practically silent; so much so that I ordered a few more to do my Y-carriage too :-)

I'm going out today and will be getting a synthetic grease, a-la the one Jetguy linked to; or, really, whatever the store has that isn't just oil.


Michael Cook

Nov 24, 2012, 2:13:43 PM11/24/12
to make...@googlegroups.com
MacGyver, that's the exact set I've got. I'll report back on my results.


Nov 24, 2012, 4:57:55 PM11/24/12
to make...@googlegroups.com
I went around to various hardware stores today and couldn't find any PTFE grease.  I have been using some Tri-Flow Superior Lubricant though.  It makes the rods super slick but doesn't do much for the chattering.  I have some Lithium grease that I used to use on my garage door when I had one but it doesn't have PTFE in it but it does say on the can that it can be used for bearings so I may give it a squirt if I can't find the other stuff at a reasonable price.

Shane Graber

Nov 24, 2012, 5:05:37 PM11/24/12
to make...@googlegroups.com
Personally I use Marvel Mystery Oil: http://www.marvelmysteryoil.com/  Found it at my local automotive store. Worked *really* well on my LM8UU's...


You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
To view this discussion on the web visit https://groups.google.com/d/msg/makerbot/-/DFX4ZeLE_rcJ.

To post to this group, send email to make...@googlegroups.com.
To unsubscribe from this group, send email to makerbot+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/makerbot?hl=en.



Nov 24, 2012, 5:17:43 PM11/24/12
to make...@googlegroups.com

Marvel probably isn't viscous enough.  You want a grease so the lubricant stays put when the bearing stops moving.  Oils run off.

Michael Cook

Nov 24, 2012, 8:47:59 PM11/24/12
to make...@googlegroups.com
Lithium grease is what I've got too (couldn't find PTFE locally). I plan to tear things apart tomorrow, apply it, and see what happens.


Nov 24, 2012, 10:03:02 PM11/24/12
to MakerBot Operators
Lithium is really for a bushing type system not a ballbearing rolling
system that we have with the linear bearings. It goes back to basic
sliding friction 101. You know, the stuff they covered in basic
physics? For a sliding system, we use a soft material such as brass,
zinc, or even aluminum, and then a hard metal for the other part. The
hard metal and the soft metal results in a low friction joint. It's
complex, but the facts are always true, two hard metals or two soft
metals do not work in a sliding situation. Ball bearings are unique in
that we roll two hard metals against each other. They both must be
hard or else, the one will erode the other from micro welds. To that
end, why the lithium grease isn't for ball bearings ever. The balls
must roll, not ever skate (slide rather than roll) or else they get
flat spots and the bearing gets noisy and fails rapidly. Lithium is a
soft metal like zinc. We use that as an additive in bushing type
joints where we have the soft and hard metal interface. Again, we do
not ever use lithium in a roller or ball bearing as it will cause the
balls or rollers to skate, and make the thing actually fail.

This type of basic info should be taught to everyone as it has wide
implications. For example, when you put a new engine together, you use
assembly lube high in zinc content because engines use bushings almost
everywhere. The zinc prevents direct metal to metal wear during the
initial break in period. But what you don't do on a car is go and pump
your wheel roller bearings full of lithium grease. If you wouldn't do
it on a car, you certinaly shouldn't do it to precision ball bearing
linear slides.

And this goes back to my original post. If you replace a brass bushing
with a linear ball bearing, you had better make sure the shaft is
hardened before you ever put it together. If not, the soft shaft will
erode, but also, the balls may skate and develop flat spots. This then
shows up as noisy bearings. Putting the wrong grease or aoil can make
it even worse and eventually, you have ruined both the shaft and the
bearings and have to replace them both. I know the original shafts in
the T-O-M are not hardened at all, thus if you are using the original
shafts, they are part of the cause and may have damaged the bearings
already, hence the noise.


Nov 25, 2012, 10:54:34 AM11/25/12
to make...@googlegroups.com
That's good info Jetguy, thanks. I don't remember that from any of the physics classes I took. I silenced my linear ball bearings by switching back to brass bushings! Those linear bearings are neither necessary nor ideal for this application. That along with about a dozen other mods finished a couple of days ago has my tom running quieter and better than ever, to where I can't hear it over my music... a significant improvement over being able to hear it down the hall from my apartment! Other mods include roller bearing idlers for the x and z axes; decoupling the steppers from the mounting surface by routing out the mounting holes to make room for nylon bushings and using 3 types of damping gaskets (just to be sure, but also to definitely prevent the raised locator around the axle from contacting the surface): rubber, paperboard, and cork; running the hulking mk6 stepper on it's own 24v power supply (until I can get one with enough current to run all of them); swapping the idler pulleys with roller bearings and isolating the axles similar to the steppers; and putting heatshrink on the metal tab of the x endstop. It took at least 5 full days to do that and a few other things, but the end result was absolutely worth it.

Michael Cook

Nov 25, 2012, 11:20:57 AM11/25/12
to make...@googlegroups.com
dreamer.redeemer, do you have links for any of the mods you've made? I'd love to know more. Noise is an issue for me in my small apartment.
Message has been deleted


Nov 25, 2012, 4:30:23 PM11/25/12
to make...@googlegroups.com
Aw man, my big, somewhat detailed post got deleted somehow... the interbuttz sometimes, eh? Oh well.

I know your pain, I live in a studio. I don't have links to specific how-to guides, but I can give you what informed and inspired me.

The 24v psu came from this discussion (and others around) mainly between Jetguy and Ed: http://wiki.makerbot.com/forum/t-358724#post-1154987 (thanks Jetguy!) I did this mostly in hopes of stopping my extruder from skipping steps, which has been a huge problem for a long time now, but I have read that steppers run quieter and smoother on a higher voltage and I do plan to drive all of them with 24v when I get a psu with enough current.

Ed's post here http://softsolder.com/2011/05/05/thing-o-matic-mbi-stepper-motor-analysis/ is referenced in that thread, he concludes that the kysan motors are underpowered, which is true, but with all the overconstraint gone and Jetty's fw my bot is able to run around half or less of what I've heard other people running; still fast enough to complicate getting quality prints.

I got ideas for isolating the steppers from here: http://www.bbastrodesigns.com/steppermotornoise.html
I made gaskets from an old bicycle tube and paperboard stock from rescued packaging, also found on the back of top bound paper like legal pads, but you probably want at least 2mm thick to keep the locator step from contacting the mounting surface. I'm still using metal bolts, but also put paperboard washers under the caps for more damping.

My idlers were sticking and groaning, so they were due for an upgrade. I isolated the axles similar to what was done for the steppers, since the belt is steel cored and probably transmits vibration well.

The z rider is similar to this http://www.thingiverse.com/thing:9434 but with bearings only fore and aft so as to not overconstrain z as per the discussion here: http://goo.gl/U2XJN
I would've used that design, but I only had 2 bearings that size (somewhat common in computer fans that actually have ball bearings, also the same that are used on the x idler/rider/follower/whatever) and wanted at least 3 points of contact to add additional constraint for rotation due to the weight of the extruder (and ended up not designing very well in that regard). I put the rider on the side I had discovered was noisiest based on putting pressure on the z platform just behind and perpendicular to each rod while the machine was running.

I also moved the emergency stop off the z platform and isolated it to keep the relay from rattling, shimmed the brass bushings for a snug fit, and elongated the hole for the bottom y idler bearing to remove rotational play about the y axis.

My guess is that isolating the steppers and reducing the load on the psu made the biggest difference, it did all add up though; after everything else was done, I went back and did the e-stop and z rider because I wasn't quite satisfied and those were obvious problems with straightforward solutions. ymmv.


Nov 25, 2012, 8:26:56 PM11/25/12
to MakerBot Operators
Or, you could just upgrade to a better frame , hence my latest which
really need to give most of the credit to Andrew from Type A machines
on whihc the design is based. Basically, get the build are you always
wanted, keep your gen4 electronics, and your MK6 and print just as
good as the new replicator.


On Nov 25, 4:30 pm, "dreamer.redeemer" <dreamer.redee...@gmail.com>
> Aw man, my big, somewhat detailed post got deleted somehow... the
> interbuttz sometimes, eh? Oh well.
> I know your pain, I live in a studio. I don't have links to specific how-to
> guides, but I can give you what informed and inspired me.
> The 24v psu came from this discussion (and others around) mainly between
> Jetguy and Ed:http://wiki.makerbot.com/forum/t-358724#post-1154987(thanks
> Jetguy!) I did this mostly in hopes of stopping my extruder from skipping
> steps, which has been a huge problem for a long time now, but I have read
> that steppers run quieter and smoother on a higher voltage and I do plan to
> drive all of them with 24v when I get a psu with enough current.
> Ed's post
> herehttp://softsolder.com/2011/05/05/thing-o-matic-mbi-stepper-motor-anal...
> > love to know more. Noise is an issue for me in my small apartment.- Hide quoted text -


Nov 26, 2012, 2:53:37 AM11/26/12
to make...@googlegroups.com
As long as the motors are coupled to the frame without damping, there will be excess noise. I took a good long look at your machine when it was first uploaded, but had my eye on other things, so I don't know if you've done anything to handle that. I'm curious, it looks like all the axes are overconstrained, am I missing something?


Nov 26, 2012, 5:57:07 AM11/26/12
to MakerBot Operators
They are over constrained but with linear bearings and the way the
rods are mounted, self aligning In other words, the 3/8ths in version
is using the original T-O-M bushings which bind with any misalignment,
but because the rods are located with caps that attach to the side
plates, you simply slide the carriage to each side and then tighten
the cap. The T-O-M locates the rods in the side plated with the laser
cut holes, then the caps just hold the rod in.

Over constraint is not a bad thing, we just deemed it as such in the T-
O-M. You must take things into context. Overconstraint means we know
100% that axis must maintain it's exact travel and not deviate. It can
cause friction when not set up right and in the case of the T-O-M, we
were moving a high mass stage with small motors and no acceleration
firmware. The problem with the T-O-M design is that there is no room
to upgrade the X motor without cutting a slot in the middle acrylic
plate and removing the power supply. Thus, since we couldn't upgrade
the motor, or reduce the mass, and didn't have acceleration firmware,
the concentration was on reducing all friction, and overcontraint
concerns along with the way the rods wer mounted mattered.

Again, you must take things in context, the new design addresses
alignment and assembly issues, uses baller bearing linear assemblyes
which are much, much smoother and lower friction, and the frame size
lets us use larger motors with double the torque of anything you could
fit into the T-O-M. Some of those things aren't obvious from a
pictures or just a causal look at the design, but make a huge
difference and the entire reason why I chose it as the basis to help
others. One key aspect is dual Y belts tied together, By pulling both
ends of the long Y axis, we ensure the force moving the carriage
maintains alignment. This is the key change from a T-O-M, and of
course, the same design the Replicators use (along with most other new

Let me spin it this way. I spent serious money trying to improve the T-
O-M. I tried the Gecko servo drive and huge motors with the middle
plate mod just to make it fit. The hole setup was a pain with multiple
power supplies, all kind of little mods to mount the linear sensors
and tapes. It printed, it was a lot quieter, but I never got the
impression that it was so much better, it should be a mainstream mod.
A lot of work, a lot of cost, and not a lot of benefit. There simply
are so many problems with the T-O-M design, it's not a quick and easy
fix. Everything from the rod mounting, to the motor mounts, to the
belt path has issues.

Again, here is my short list of reasons why I like the new design.
Space for larger motors, even 3 stack motors liek the MK6 for ultra
high torque
Larger build area than anything else out (9x9x9 in).
Space for multiple extruders and any mod or tool attachment you can
think of.
Dual belts for Y to prevent cocking of the long axis.

The proof is in how well it works, and sorry, I have done a poor job
of demonstrating that to others. Maybe it takes me selling a couple of
frames to key folks to try out for next to nothing. For example, If I
cut a frame for a limited time of $50, you cover shipping too, and
send these out for some people to try out, would that be of interest?
You reuse most parts of the T-O-M. You only need to buy some smoooth
rod (`$30), about 150 M3 x 14 screws and nuts ($12 from Amazon), and I
highly suggest you get the motors from Ultimachine for another $40
($19 each).
Pick up probably 3 extra 40mm cooling fans from Ultimachine while your
at it. You also need belts and pulleys from SDP-SI, but they aren't
that bad. Those who had a Cupcake before have the required pulleys.
Or, for only a few $ more, you could get 8mm rods for the same price,
get 8 linear bearings at $1.75 each making it just a viable to stick
with the original design ( in other words, $14 more for the linear
bearings). I just tried to give the option since it's supposed to be a
T-O-M upgrade. And, the side benefit is that due to the open, dino-
less design, and that attached 40mm cooling fan, The MK6 has never
printed this well with PLA. You could take some of the ideas and apply
them to the T-O-M, but, it's still the same build area, still the same
rattling noises of the design. I don't have anything like that
happening with the new design and I've been slamming it around day and
night at 120mm/s pushing MK6 1.1lb motor. BTW, dual MK7/8s weigh the
same as a single MK6, but just like the T-O-M, the Replicator cannot
use larger 5.5kg/cm motors as they just won't fit, even with mods.
This is where this design shines. It can do everything a Replicator
can do, doesn't have the crappy platform sag from plastic arms, and
most important, doesn't cost you $2,200!!!!!!! I believe it keeps in
the spirit of the original MakerBots, in that we had some very good
parts (gen3, gen4, Mk6), but the frame was our limit. Look at the mods
for the Cupcake and what it could be transformed into. I love my
Cupcake with the 5D shield, the new motors, Aaron's amazing lowrider
and Z rider upgrades transformed it. Many tried to do it with the T-O-
M but none are the magic fix. Sailfish techncially is the magic fix,
but because it helps to mask the physical faults, we are at the limits
of what the T-O-M can be without a new frame. But the new design is
about besting the Replicator series, while helping T-O-M owners who
feel like they are left behind. Further, the Guy with a Cupcake, MK6,
and I 5 D shield is really in the best shape for this upgrade. They
have the pulleys and at least 2 of the belts!! Again, it's thinks like
that that make me feel this might be the best upgrade in a while for
those who invested in the older machines.

On Nov 26, 2:53 am, "dreamer.redeemer" <dreamer.redee...@gmail.com>
Reply all
Reply to author
0 new messages