Ultimaker Extruder Drive Upgrade

by Bertho, published

Ultimaker Extruder Drive Upgrade by Bertho Jul 1, 2012


Liked By

View All

Give a Shout Out

If you print this Thing and display it in public proudly give attribution by printing and displaying this tag.

Print Thing Tag

Thing Statistics

44596Views 3615Downloads Found in 3D Printer Extruders


Ball bearing & spring loaded pressure plate for uniform feed regardless of filament thickness


Included in the zip file are:
.par =Solid Edge
For detailed instructions see:

The bearing can be found here for $7.41:
It is a Dynaroll SMR128ZZ A5

The spring OD is 7.72mm, has 8 turns of 1.14mm wire, and a free length of 20.9mm. Sorry, I do not have any part number since it came from my big junk box...

Since I needed to measure the spring used for the Ultimaker Upgrade I documented the steps both for a high accuracy measurement procedure and a simplified one.


Note, the spring for the upgrade is not critical. There is a tremendous adjustment range.

I added a section on Extruder Drive Force Measurements:

The long screw is 50mm.
The shaft screw needs to be 20mm since the yoke is 18mm.

More from 3D Printer Extruders

view more

All Apps

Upgrade this Thing with Thingiverse Apps

Tools and Utilities

Repair, slice, or enhance this Thing

Auto-magically prepare your 3D models for 3D printing. A cloud based 3D models Preparing and Healing solution for 3D Printing, MakePrintable provides features for model repairing, wall thickness...

App Info Launch App

Thanks a bunch for this! finally upgrading from the simple plate after 4 years

Thanks for the great design. The one I had been using previously was very fragile, but this one works great. The only modification I had to do was to lengthen the yoke by 10%.

You are welcome Cabonbased. I am still using the same design. I have switched from PLA to ABS. So much nicer to work with and easy to use with additional machining when required.

The original from Ultimaker has taken a different approach to this design and holds the bearing within the main frame.
Bertho's instead holds the bearing in the spring loaded clamp.
This difference allows Bertho's setup to put more pressure on the filament than Ultimaker's design.
I tried both setups and I ended up printing one of the pieces of Bertho's design and modifying the acrylic parts of the Ultimaker kit.
I found this solution being the most effective so far :)

There is no space for the locking nut inside the main plate, so be prepared to drill pits for it or file down the protruding edges of the nut.

I do not know what is different or wrong with your setup since the original works fine and so does the production version. Make sure the white locking lever is properly closed or it will not work.

i have the official parts from Ultimaker shop.
Doesnt work with my Ultimaker...

absolutely ZERO improvement on my side...
not a single one of my filaments can be extruded with this "upgrade"... not even by hand.

any ideas?
the filament has too much play sideways.. also i think the slits for the pressure-movement are too short. the bearing is not realy pressed against the filament but against the ends of its slits.

I bought a pack of 150 hobby springs and none of them were strong enough. I happened to pass a motorcycle parts shop and wandered in on the off chance they would have something. He had a box of springs on the counter and I managed to find 3 candidates, one of which worked. As I was walking home I spotted a washer on the ground that worked for me too! All in all it was a good day. My prints are much more consistent now. Thank you!

I spent a few minutes looking on McMaster for the cheapest springs that would fit the bill. This is what I think will work: http://www.mcmaster.com/#cadinlnord/9434k164/=nble81http://www.mcmaster.com/#cadin... - they are about $5 for 5, and can generate 7.5kg of force at full compression.

I got my spring from a hole puncher.

For those sourcing out a suitable spring, just want to share that I found one from an unlikely source -- a bicycle shop.

Bike derailers usually contain a spring or two with force and overall dimensions that's suitable for this extruder.

See pics below:


Direct link to derailer pic:

Compare to Ultimaker shop image:

The bearing is another story though as I'm still unable to get the best fit. I found a 12mm OD bearing but with a bore of 4mm, which is too thick to fit in the filament canal.

M3 50mm screws are hard to find! Where did you get yours?

 I did not buy it, I have a BIG box with various screws.  The 50mm dimension is not critical.  40mm is fine with my spring and with a little stiffer, shorter spring even a shorter screw can be used.

Hi, What size is the ball bearing, i try to find it at mcmaster. but that site is a joke for searching, it has locked me out for 24hrs for searching: Dynaroll, SMR128ZZ, 128Z.

Id like the measurements in metric please :D

From McMaster Carr, looking it up with the dimensions you gave, I get 7804K115 stock number. $7.41 plus SH.

http://www.modelfixings.co.uk/Heli%20cylindrical.htmhttp://www.modelfixings.co.uk/... has lots of suitable bearings, 4mm width should be fine. I am waiting for a shipment of 2, one with 3mm ID, for a perfect fit for the M3 axle, and 8 or 10mm OD, and another 4mm wide bearing with 8 or 10MM ID and 14-16mm OD, basically a bearing in a bearing setup, since 2 standard bearings are still cheaper than one odd sized bearing. and it has a deeper reach to the bolt.

Thanks for listing the supplier.

You need at least 12mm OD bearing or the bearing will not reach the filament. Of course it is possible to grind away some material from the original plywood sides to make it work.

Similarly, a 4mm wide bearing will be a tight fit between the plywood sides. You might have to add a thin spacer between the plywood layers to slightly increase the spacing or grind away some material.

yes, 4mm is tight, but only between the two large bearings. the lower hinge part could be 4.6=4.8mm thick, to prevent the bearings from touching the sidewalls, and the parts that reach into the plywood could be 3.8mm, so it flips nicely up and down. 4mm bearings are also cheaper and easier to get than 3.5, and I think the extra reach of 14mm OD will help a bit

Lower cost is always good!

The larger 14mm diameter is also good since 12mm is marginal. It will give more clearance for the shaft and the yoke.

You need to add spacers so the two plywood sides will be a slight amount more apart or you have to grind off some wood. At least my UM only has 3.8mm spacing between the two plywood outside parts.

I did a lot of swearing last night trying to assemble this, and will be reprinting several parts I broke. The bit I missed in the instructions was "file off two sides of the 50mm screw head." :-)

I was trying to somehow assemble it with the full width of the screw head-- which emphatically does not work. That was after I realized it was the head and not the nut that fit into the slot in the delrin part. headdesk

Clearly I shouldn't start assembly at 10pm on a Saturday after a pleasant
dinner with a bit of wine.

Instead of filing the screw head down to size, you can also make a bit room in the plastic side panels with a knife or drill bit.

I intentionally did not provide any clearance in the side plates for the screw head. The screw head is filed down on the sides to form a
“T”- head so the screw cannot rotate. It works well.

I am happy you got it resolved. Too little sleep will also create similar problems.

The good news is that soon you can print without worrying about fine adjusting the spring force because of changing filament diameter.

Currently it's not working, but not I think for any reason related to this part, though the symptom is lots of grinding and slipping. There's a lot of resistance on the filament when I try to advance by hand; I'm not sure quite why. Just took the head apart looking for clogs, but there wasn't one (the new PEEK insulator with the threaded bowden tube is working! yay!) so I'm not sure what it is.

Ah, well. I have other things that also need to happen today. I'll get back to it this evening.

You probably already have tried it:

Open the drive and feed by hand to get a feeling for what is happening.

It should slide easily back and forth through the Bowden until it hits the extruder. If not, resolve and fix.

That leaves 4 options:
Plugged or constricted Bowden in that area
Too low temperature
Blocked nozzle
Something else :-)

Good luck troubleshooting

Yup. There is a fair amount of friction in the Bowden, but not so much I can't hand advance. Once it hits the nozzle it gives intermittent resistance. It extrudes smoothly for a bit, then I need extra pressure as if to get past some sort of blockage, and then it feeds smoothly again for a while.

I thought it was a filament thickness problem, which is why I built your extruder-- this was happening last week. My second guess was the filament had taken on too much moisture, but the batch I'm using today spent the past 24 hrs in a dehydrator, so I don't think that's the case either.

is no sign of a blockage or leakage at either the nozzle or the Bowden junction, so I'm kind of out of ideas at the moment. Thermal fluctuations, maybe? My heating element and thermocouple have been used pretty hard with all the times the head's been apart and back together.

shrug I'm sure I'll
sort it. Deadlining on something else at the moment, though, so I have to let it stew.

My extruder design will avoid the problems of the fixed pressure setting that drastically change vs. filament diameter. It slips when too thin or deforms the filament if thick. With the spring-loaded ball bearing the diameter does not matter for the drive mechanism as long as it fits in the Bowden.

I have seen increase in required extrusion force after sitting hot for a little bit. Presumably the plastic melts further up and keep moving upwards and when it cools, it gets hard to push it into the extruder.

Found the source of friction. The Bowden tube where it's screwed into my new PEEK insulator appears to have expanded. I can't push filament through it smoothly when it's off the machine. I have some 4x6mm teflon tubing so I may rebuild with that to give it more expansion room.

One other possibility is that the threaded section is forcing it to a smaller inside diameter. Maybe rethread it and do not tighten too much.

Very nice upgrade. I am getting more even extrusion and less problems with a filament plug just above the hot end brass tube. Note: If a different size bearing is used, you will need to adust the length of the yoke forks to account for the difference. I tried using a small diameter bearing without ajustin the yoke and the preformance was worse than with the original Delrin piece.

Thanks ChemWhite,

Yes, bearing size is critical. Too small diameter and it will not reach, too wide and it will not fit between the plywood sides and too large it will not fit in the cutout area.

Happy printing!

Great work, thanks. Printed one. I just added two spring caps because i use a different spring (shorter and wider) that looked ugly...

It looks good with the spring caps.

Happy Printing!

I've started to print this, but wonder how do you deal with the fact that the yoke is 18mm wide, whereas the default assembly is only 16mm? Because right now in the default assembly, my gear is almost touching the heads of the screws already.

the bushing seems a bit too thick. I had a better fit printing it 3.5mm instead

I wonder how you do 'tighten' the shaft screw, since there is only 2mm left for the nylock nut. Or did you use a longer screw for it?

I see the same on my extruder. I would have to add a washer to my large gear. Otherwise, the yoke screw will definitely interfere with the large gear. Maybe if I use a countersunk head screw on that side of the yoke it might just clear it.

The yoke is 1mm wider on each side than the plywood. The screw heads are 1.5mm so the gear must have more than that in clearance. To line up the big gear with the little gear I also have an extra washer on the big gear inside.

The Delrin part is 3.7mm and the bushing should be the same for a sliding fit. that assures that the 3.5mm bearing will not rub against the sides.

I used a longer screw for my final assembly.

If anyone has ordered all the parts somewhere, please share the information :)

I'm searching for the spring, the screw and shaft screw.. I don't have any junk boxes here to salvage those parts.

A list of parts that can be ordered online would help a lot.
(I didn't find the bearing from mcmaster site, but I found it by googling the name).

Here is another adress, where for example the german ones could buy the bearings. I
´ve ordered some today for 1,51€/piece.


The bearing is readily available at low cost from Ebay. Either search for SMR128ZZ to get the exact same thing, or search for any ball bearing of dimension 8X12X3.5 mm

Are you sure you're not overlooking the obvious - your local hardware store


I have a combined "paint supplies and hardware" store in the tiny town I live in and they have all I could want of nuts and bolts. Anything else you can find easily on ebay (like the bearings)

I could be... but this far I haven't found a shop that would have bearings this small. Probably there is a shop somewhere, well hidden ;)

I ordered the bearing online and it came in just two days, ready to pick up already today. So it's ok now, I can finally start building tomorrow :)

The bearing is 12mm OD, 8mm ID 3.5mm Wide.

Search on McMaster page for: Metric, 12mm OD, 8mm ID
amp; 3.5 W.

The part # is 7804K115

Thank you for the quick reply :)

I tried to order few things from mcmaster site but it didn't work out. They sent me mail that currently they don't ship outside United States.

I'll try to find similar stuff from some European webshop.

http://rs-online.comrs-online.com most likely will have the parts, but not sure if the price will be good.

Thanks Daid, they certainly seem to have them.

I suppose this would be ok for the ball bearing (the info is in English just below the product image): http://fi.rsdelivers.com/product/nmb/ddl1280zzmtha5p24ly121/min-plain-bearing-8-idx12-odx35wmm/0747800.aspxhttp://fi.rsdelivers.com/produ...

I also can find the other stuff like 50mm M6 screw and nylock nuts which I don't have
but they always come in boxes of 100 pcs.. This gets quickly rather expensive :'(

Yes, that looks like the correct bearing, the one without the flange.

some form of link to where to find springs might be helpful, the rest is trivial... maybe you could consider a design option to use (grooved) standard skate bearings (22x8x7mm)

Sorry, the roller blade bearings are much too wide. The bearing must fit in-between the plywood sides.

I really like this upgrade. I need more information though. What size is the ball bearing? Where did you get it? What length is the long 3mm screw? How long is the screw on the yoke piece going through the bearing? How long is the spring uncompressed. Where did you get it? An idea of how stiff the spring is?

from what I remember, the force exerted by a spring is given by F=kX where k is the spring constant and x is the deflection. What that means is if you can find a spring that fits the diameter of the yoke, and is roughly the right length, you should be able to change the deflection until you get the right force. So if you're worrying about finding the right spring, don't - just find whatever you can, and experiment.

I'm going to print one of these over the next week, and will post back with my progress on finding a spring that works, and how it all turns out.


Yes, your assumption and suggestions are correct.

The sample spring I used will give three times the force if I fully compress it.

Did you see my data and testing?


that's very thorough - good work! Am printing the yoke right now,will be scavenging some springs this morning and will update with progress. Not sure, but I reckon an old spring from a pen might well do the trick...

I added the info in the instructions. I will update with the spring force data.