Fine tune your MK3 Y axis  

Page 1 / 5
  RSS
RHDreambox
(@rhdreambox)
Honorable Member

This part is about the Y axis and how to fine tune it. It's part 2 of 3 planned posts.

Background, see Fine tune your MK3 Z axis

To check if there were any details on the printer that I could improve, I began to draw a Prusa Original MK3 in my CAD system (Inventor).
I have read the measurements in Prusas SCAD files.
With a 3D model of the printer, you can explore, twist and turn to every detail.

I then found that there are some improvements to make on the X, Y and Z axes.

Small vibration can be a source of uneven printing.
To eliminate these, the belt must run completely horizontally.
If the belt is operating at an angle to the drive wheel, the belt travels vertically when the bed is transported forwards and backwards.

Quick variations at elevation can cause the belt to vibrate.
And if the drive belt is angled, the movement will not exactly be what the g-code prescribes (Pythagorean theorem).

Attached Y-axis parts are modified so that the belt runs completely horizontally, and the screwhole dimension has been reduced to reduce the risk of misplaced components.
If the belt travels laterally, one of the two wheels is inclined.
Drive motor, eg. can be fixed with a cable tie around the motor housing and end plate.

Happy owner of a Prusa MK3...
Posted : 24/01/2018 10:16 am
Olef
 olef
(@olef)
Honorable Member

Looks like good work. Thank you!

Posted : 24/01/2018 10:20 am
Zaz
 zaz
(@zaz)
Estimable Member

One thing I'd like to see more of in these kinds of discussions are the calculated severity of the problem and expected impact of the fix.

Like how you say "Small vibration can be a source of uneven printing." - Sure, but how much? Are we talking artifacts on the scale of 1, 10, or 100 micrometers? Or even more? Will the fix remove the problem completely, or are we just mitigating it?

Because my guess is that there are probably hundreds of minor things contributing to the total "sum" of precision errors on my printer, and if this isn't in the top 10, heck top 3 even, it's probably not what I need to start fixing.

To be clear: I'm not saying this is a bad idea, or that nobody should do it - and I appreciate the work you put into it, RHDreambox. 😀

Posted : 24/01/2018 12:11 pm
RHDreambox
(@rhdreambox)
Honorable Member

Like how you say "Small vibration can be a source of uneven printing." - Sure, but how much? Are we talking artifacts on the scale of 1, 10, or 100 micrometers? Or even more? Will the fix remove the problem completely, or are we just mitigating it?

Because my guess is that there are probably hundreds of minor things contributing to the total "sum" of precision errors on my printer, and if this isn't in the top 10, heck top 3 even, it's probably not what I need to start fixing.

Zaz and alfred
I'm glad you appreciate my work for a better MK3.
3D printing is about precision and the difference between a bad or good end result may be due to very small details.

I therefore consider that if you can eliminate these sources of error, no matter how small they are, you should do that.
Then you can exclude these details if you encounter printing problems.
Happy printing 😀

Happy owner of a Prusa MK3...
Posted : 24/01/2018 12:25 pm
hiElvis
(@hielvis)
Eminent Member

Thanks for these detailed descriptions and model "remixes"!

But as Zaz already mentioned, some numbers as to the difference of certain dimensions of your parts compared to the original might help people determine if printing them makes sense. Talking about myself for example, I own a chinese i3 clone with ok-ish dimensional accuracy and if the error is below 0,1mm I won't bother printing replacement parts, cause the chances of the new part being better than the original are way too low 😉

Posted : 24/01/2018 12:50 pm
nathan0876
(@nathan0876)
Estimable Member


Thanks for these detailed descriptions and model "remixes"!

But as Zaz already mentioned, some numbers as to the difference of certain dimensions of your parts compared to the original might help people determine if printing them makes sense. Talking about myself for example, I own a chinese i3 clone with ok-ish dimensional accuracy and if the error is below 0,1mm I won't bother printing replacement parts, cause the chances of the new part being better than the original are way too low 😉

I suspect you are correct that for small improvement sub 1mm printing replacement parts will be a crapshoot as far as if they are exactly accurate to the intended dimensions.

When i get my printer, and its setup and running well on stock parts. Im going to print out all these improvements for all 3 of the axis and measure them very thoroughly with a good micrometer to ensure they are the correct size before installing them. I will likely print multiples of all the parts and install the best ones out of the bunch.

Posted : 25/01/2018 6:39 pm
RHDreambox
(@rhdreambox)
Honorable Member

Nathan0876, I hope you report the result!

Happy owner of a Prusa MK3...
Posted : 25/01/2018 7:22 pm
devdoc
(@devdoc)
Estimable Member

I'll give it a whirl. I wish there was a better answer than the zip tie for the motor, but it is what it is. I am looking forward to what you have on the X gantry as I think people are actually having numerous concerns with allignment of the belts an tolerances. Not non functional ones, but mostly nuisance related. I look forward to what you come up with. Any idea on when you might get a chance for part three. I am using an MK2 adjustable x at this point to eliminate the issues, but I look forward to your attention to detail on your solution. It might be nice if adjustment was a built in feature as opposed to taking off the back cover of the extruder mount. The dial o tension method I have now is great, but I have concerns that the geometry is not 100% the same as the MK2 for whichi it was designed.

...
Posted : 25/01/2018 7:59 pm
RHDreambox
(@rhdreambox)
Honorable Member

I hope you've already found part 3 about the X-axis 😉

https://shop.prusa3d.com/forum/original-prusa-i3-mk3-f30/fine-tune-your-mk3-x-axis-t13853.html

Happy owner of a Prusa MK3...
Posted : 25/01/2018 8:56 pm
nathan0876
(@nathan0876)
Estimable Member


Nathan0876, I hope you report the result!

I will bud, will be a few months before i get the printer, but when i decide to do something i dive in headfirst and 3d printing is something ive been interested in for years so im immensely excited to get started. Ill update these threads later with results for sure.

Posted : 25/01/2018 9:18 pm
themzlab
(@themzlab)
Estimable Member

I have an MK3, been printing for a few weeks now. Also, I have the CAD and I have not found a problem with the Y axis belt path.

The pulleys are offset in height so that the top span of the belt come out 5mm offset (vertical). The belt clamp has two slots in it offset by 5mm vertical.

This causes both spans of the function part of the belt to be parallel to the rails. The bottom span goes at an angle through the free air but this doesn't matter.

What exactly is the problem that has been found in the Y belt path? The main problem I know of is that the Y belt sometimes does not track correctly (left to right) as others have pointed out. I printed a new belt clamp with the holes shifted left 1.5mm (same affect as shifting the belt slot right 1.5mm). This moved the belt perhaps too much. However, this is not the same dimension being discussed here.

Thanks for your work on this - while results may vary I really like the idea this machine is open source for real and that we can all work together on potential improvements : )

Mark

Posted : 26/01/2018 12:18 pm
RHDreambox
(@rhdreambox)
Honorable Member

What exactly is the problem that has been found in the Y belt path?

The upper side of the drive belt slope in relation to the horizontal plane differently on the motor side and the runner side.
There are no big slopes, but they are there.
As I see it, there may be a risk for small vibrations in the belt when moving the Z unit fast back and forth. This also applies to the X drive belt.

And the X and Y units movement is not exactly what the g-code says (Pythagoras formula).

Happy owner of a Prusa MK3...
Posted : 26/01/2018 1:37 pm
john.g9
(@john-g9)
Eminent Member

Great work on all of these. I am printing them up but was thinking on the y axis motor mount is there a way to bed it up some? The factory specs just seem way to weak. It's thin, small and has enough force on it sure seems like over time it would give way. It would be nice to thicken it up if possible and specs allow? Thanks again for all the work let me know what you think about strengthening the mount.

...
Posted : 29/01/2018 7:36 am
RHDreambox
(@rhdreambox)
Honorable Member

Great work on all of these. I am printing them up but was thinking on the y axis motor mount is there a way to bed it up some? The factory specs just seem way to weak. It's thin, small and has enough force on it sure seems like over time it would give way. It would be nice to thicken it up if possible and specs allow? Thanks again for all the work let me know what you think about strengthening the mount.

Something like this? Feel free to make suggestions.
The important thing is that the Y engine is completely parallel to the rear end plate.
Therefore, a tip is to mount a cable tie around the frame and motor.

Happy owner of a Prusa MK3...
Posted : 29/01/2018 9:10 am
john.g9
(@john-g9)
Eminent Member

Okay so I was thinking about beefing up the Y-Axis motor mount as I had mentioned earlier. Any reason why this would not work? Or thought on how to improve it? Couple of sketches and file in the zip folder.

Basically I added some "meat" to the original and widened the back where it connects to the frame. This would use the stock fasteners still since the holes remain the same depth from the motor.

Just a thought. Also just added a fillet to the original file you had , I redrew it in fusion so hopefully the measurements are spot on?

...
Posted : 30/01/2018 1:16 am
john.g9
(@john-g9)
Eminent Member

Sorry I posted my feeble attempt before I saw yours lol. Yeah that looks nice- Is there anything you see stopping one from using a filet on the side facing the screw? I was looking at my machine and I do not see any reason why someone could not add the fillet and even thicken the mount all together a little on the opposite side of the motor.

...
Posted : 30/01/2018 1:31 am
john.g9
(@john-g9)
Eminent Member

Honestly I think the stock / original mount has some flex in it adding to the issues with eh MK3 (especially in PETG), the speed is faster but the bracket is so small it can't help but flex or add to the vibrations of the machine. It's actually the same size or smaller than the MK2 which I printed larger mounts for it as well before I built the kit. Also the bearings act as the end stops so there is really nothing interfering with the Y-axis motor mount being bigger or thicker from what I can see.

...
Posted : 30/01/2018 2:08 am
john.g9
(@john-g9)
Eminent Member

What do you think about this? I thickened it all and put the frame mounting screw actually in the center of the part, unlike the stock version offset. This should help alleviate the twisting tendency?

...
Posted : 30/01/2018 5:04 am
RHDreambox
(@rhdreambox)
Honorable Member

Yes, or why not go one step further?
Or is this "overkill" 🙂

Happy owner of a Prusa MK3...
Posted : 30/01/2018 9:13 am
john.g9
(@john-g9)
Eminent Member

Never Overkill can't be such a thing :). I think the weak spot in the original design is the fact that there's not enough connection and surface touching the frame and the screws or so small and so close to the motor itself that the mount acts sort of like a hinge it's very easy to pull and push because of its connection points. I'm going to have to look a little closer but I don't see any reason why you couldn't make a mount with a third connection point and just drill a new hole in the frame. That way it would have three points of contact / connection and more surface area with the frame. That's why on the one I was tinkering with I extended the thickness of the mount to 10mm where the motor mounts and then extended the top of the frame to put the screws in the center of the motor mount. Nothing wrong with what you've designed here at all I'm just trying to think of the weak spot in the system and it's definitely where (or now it is) where the motor mount meets the frame and the small connection surface and fasteners. Also don't see why a person couldn't just slide the Y motor forward and make a true Mount where the motor was set inside the mounting bracket and it had connection points on each side of the motor I just think that was cheap design on Prusa'$ part and they didn't spend a lot of time trying to make it robust. ('$ was intentional lol- or better yet (Pru$a'$). Good stuff !

...
Posted : 30/01/2018 3:03 pm
Page 1 / 5
Share:

Please Login or Register