Throw away your short belts - direct drive.

[albertdesigns 0]

Hi all,

I also made a direct drive, but did this by drilling (turning) a hole in one x and one y axis, and glueing the motor shaft in.

So what you get then, looks like a stepper motor with an enourmous shaft.

This motor + axis can then be mounted to the UM, without the coupling, and without the extra added length.

Watch out: Because especially the motor shaft will never by exactly straigt, so rigidly mounting the motor to the UM-frame is no good idea. Therefore you will need what we call a torsion arm to transfer the driving torque from motor to frame.The torsion arm should be stiff in rotational direction, but flexible in the other directions, in order to allow the wobbling of the motor. I made two torsion arms from 1.5mm thick aluminum plate, and connected both with 4 countersunk screws to the stepper motor. I put both motors on the left/back side. Will add a photo later.

I had to reverse the direction of x and y motor, otherwise the printing of multiple object went wrong. I did this by reversing the wires of one pole of each stepper motor. (in the connectors there is a little catch for every wire, you have to press it in, so you can swap wires without any soldering)

I did not measure the before/after accuracy, but the belts are definitely the most important compliance (1/stiffness), so it will improve accuracy. Because friction in the xy-mechanism is quite high, accuracy will mainly be determined by the so called virtual play: friction/stiffness. (Newton / Newton/mm = mm)

Main reason for this "hack" for me was that especially the short belts lost pretension very often, but the extra accuracy was another reason of course.

Regards,

Albert

 

[jacobs 0]

Hey everyone,

I've been wanting to do this mod for quite a while, but am having a bit of trouble finding a supplier for the shafts. I was looking at the ones snowygrouch suggested from conrad.de, but they charge 40euros to ship to australia, which is a bit steep considering the shafts are only 10.

The only local ones i have found that meet the specs are a bit overkill, costing around 50dollars each. Or maybe I just don't know where to look.

Someone suggested one of the rod sets from ultimachine, maybe this one:

https://ultimachine.com/content/prusa-mendel-smooth-rod-set

I could get this while ordering some other things, but are they suitable?

So, If anyone could point me in the right direction, that would be fantastic...

Cheers,

Jacob

 

[snowygrouch 1]

Personally I refuse to buy anything that doesnt have a diameter tolerance given.

 

So instead get these:

http://australia.rs-online.com/web/p/linear-shafts-rails/3411560/

400mm long

60+ HRc

8mm dia nominal to h6 tolerance

In fact these are probably better than the ones I have on my own machine...

 

[owen 19]

An ebay seller from Victoria is selling rod

[jacobs 0]

Thanks for that, I just thought I could get away with something a bit cheaper. I didn't see any with high enough precision and specs on ebay, so if nothing else I'll get the ones snowygrouch suggested, but 140dollars ouch....

Otherwise the ones I was looking at, were these ones:

http://www.smallparts.com.au/store/item/sh08000x037500c1060/shaftinggroundprecisionchromesteel/

They are a bit cheaper, but still the same specs, I think.

-Diameter 8.000mm

-Length 375.0mm

-Straightness 0.0003 mm/mm

-Surface finish 0.25-0.4 um

-Hardness 60-63HRC

-Diameter tolerance -0.000/-0.009 (which i think is h6)

I was thinking if I get these I might not need to cut them down, as I lack proper equipment for this... or is that a bit of a gamble?

 

[snowygrouch 1]

Those also look fine, h6 is as you correctly state 0/-0.009

 

I didnt see the price of those....thats absurd. I guess because they are SKF shafts, so are probably

about the best quality you can buy. But still....really exessive !

To cut them down will be a real pain at that level of hardness, a steel file is 60 to 65 HRc

and a typical cheap hacksaw blade is below even that.

However you can get really good hacksaw blades that are up to 68HRc.

Even cheap blades will probably "work", but you might just end up going through about 3 blades

to make 2 cuts. If you have a vice to hold the shafts in (with tape around the shafts) it should

present no problem. Just make sure you do a really nice job rounding off the ends or you will never

get them through the bearings in the UMaker case.

 

 

 

[owen 19]

This was the ebay seller I was talking about JacobS

http://stores.ebay.com.au/cncandcupcakeworld

Looks like he only comes down to 10mm

It may be worth asking him about 8mm and even give him the exact length you want

Those other 2 suppliers have a lot of stuff but are quite expensive

 

[jacobs 0]

Owen, yeah that could be an idea. I did spot his site but didn't look further as he didn't have 8mm. However, I think I found the same rods from china in 8mm. But these say tolerance g6, is that the asian standard equivalent to h6 tolerance, or something different?

Well, if anything, I am learning more about steel rods than I ever thought I would need to know haha

 

[owen 19]

But these say tolerance g6, is that the asian standard equivalent to h6 tolerance, or something different?

 

I don't know. I'm still learning myself. Do you have a link?

 

[jacobs 0]

http://www.ebay.com.au/itm/2pcs-OD-8mm-x-400mm-Cylinder-Liner-Rail-Linear-Shaft-Optical-Axis-/290765610622?pt=LH_DefaultDomain_15&hash=item43b2fb1e7e&_uhb=1

that is the chinese supplier I found. The australian supplier wrote the following regarding precision tolerance:

Asia tolerance standard g6(um): -5-14

Europe tolerance standard h6(um): 0-9

which made me think they might mean the same thing.

 

[snowygrouch 1]

8nominal to g6 is -5/-14 micron

8nominal to h6 is 0/-9 micron

 

I think "Asian equivalent", is the internet sales code for "not the same but if you pay we dont care".

g6 would probably work "ok", it might just be fractionally sloppier than you would like. Its hard to say

exactly because I dont have tolerance data for the ultimaker sleeves.

Here is all the info you might need on that stuff...

http://us.misumi-ec.com/pdf/press/us_12e_pr1263.pdf

http://us.misumi-ec.com/pdf/press/us_12e_pr1261.pdf

 

[jacobs 0]

I suspected that might be the case snowygrouch...

Well, I went ahead and ordered the chinese rods. I thought I might as well try them in the event that they do work, before spending 140dollars on good ones.

Thanks for the help.

 

[snowygrouch 1]

http://www.smallparts.com.au/store/item/sh06337x033020c1060/shaftinggroundprecisionchromesteel/

If they dont work out you can try this place....sorry I just remembered about it this morning

 

[nick-foley 5]

Ordered the goods to make this happen (along with a GT2 upgrade) today. Using couplers and pulleys from Adafruit, shafts from McMaster, belts from SDP-SI.

Hopefully the upgrade combo will give some better XY repeatability. Currently, even with tight belts, i'm noticing a very slight shift whenever the print direction (ie clockwise vs. ccw) of the skin changes. Totally negligible on large parts, but problematic on small parts with critical fits. Will post a part quality before-and-after photo as soon as the upgrades are implemented.

 

[albertdesigns 0]

Instead of buying new shafts, you can also (ask someone to) drill a hole in your current shafts, (on a turning lathe for accuracy) and glue in the shafts of the steppers with Loctite.

Like I said earlier, a torsion arm should be mounted to the motors, to transmit the torque to the UM frame. A used 2mm aluminum plate for this. These arms will allow a small wobbling movement of the motor when rotating, caused by the fact that the motor shaft will never be perfectly straight. The same is true for the hole in the xy-shafts.

Regards,

Albert

 

[nick-foley 5]

Finished the upgrade, happy with the print results and the simplicity of the machine. As I worried, however, the machine is definitely louder now. The steppers have a high-pitch whine which they didn't used to have. The motion system itself is very smooth and easy to move, and the motors are mounted to the frame just as rigidly as they were previously, so I suspect the noise is coming from the steppers resonating through the 8mm axes.

Has anyone experienced this?

 

[ayahoo 0]

I know it is an old thread, actually i did it few month ago, but forgot to upload a photo.

My simple solution for mounting the motors.

using 4 hexagonal spacers found on dx.com or local hardware store. (couldn't print motor mounting part since my short belt was torn)

 

[aviphysics 9]

I am thinking of trying this version, which uses the original rod.

http://www.thingiverse.com/thing:189742

 

[aviphysics 9]

Also wanted to add that Heli-Cal makes good helical couplings for the U.S. market. You have to call for a quote, but they are inexpensive, will take small orders, and ship fast. Always been really friendly too.

The other thing I like about them is that they seem to have coupling for virtually any combination of diameters.

 

[nick-foley 5]

I am thinking of trying this version, which uses the original rod.

http://www.thingiverse.com/thing:189742

 

I like the idea of that mod... makes it even easier to upgrade! If you have access to McMaster, getting new rod is quite cheap too, though. You can order this rod for $23, cut it in half, and use it for both axes: http://www.mcmaster.com/#6112k15/=ptmuak

Anyway I designed a direct drive motor mounting bracket that doesn't require any modifications to the UM frame, holds both motors with one piece, and is somewhat easier to assemble than other brackets. It should work with this mod, too. I'll post the files shortly.

 

[aviphysics 9]

I like the idea of that mod... makes it even easier to upgrade! If you have access to McMaster, getting new rod is quite cheap too, though. You can order this rod for $23, cut it in half, and use it for both axes: http://www.mcmaster.com/#6112k15/=ptmuak

Anyway I designed a direct drive motor mounting bracket that doesn't require any modifications to the UM frame, holds both motors with one piece, and is somewhat easier to assemble than other brackets. It should work with this mod, too. I'll post the files shortly.

 

Funny that they don't list it as h6, but the tolerance seems to match.

 

[nick-foley 5]

Here's the bracket I designed to simplify the process of converting to direct drive. It could be used in conjunction with the solution posted above to enable re-use of the original shafts.

http://www.thingiverse.com/thing:208343

 

[aviphysics 9]

Just printed the version 9 motor bracket, and I am wondering what settings it was intended to be printed with.

The part is functional, but the printer had a tough time with the countersunk screw holes on the part that mounts to the UM Frame. Then end result is pretty rough, but I am guessing it will work.

Will also look at your mod Nick, though I think I will just switch the wires so I don't have to customize the firmware.

 

[aviphysics 9]

Finally implemented this mod. Haven't printed anything with it yet, but the machine just sounds more precise and the axis seem to move much more smoothly now. Also replaced the outer frame, which had some sort of defect or warping, so that might have had an impact.

I went ahead and used the rod I had bought from McMaster, but want to order some higher qaulity rods when I get a chance. The McMaster rods don't seem to meet the stated specs. They were nice enough about shipping replacements, but the replacements were also not entirely up to spec. That said, they were only slightly worse than two of the four 8mm rods from the Ultimaker kit. I also found that once installed in the machine and everything was under tension, the slight banana shape of the rods was not observable when they rotated. I am guessing that the limited degrees of freedom somehow keep the rods strait.

 

[Dim3nsioneer 557]

@aviphysics: You won't regret it... :-P