Questions after UMO upgrade to GT2 belts and pulleys

[skewcrap 3]

Hi everyone

I have an Ultimaker Original and 2 days ago I performed the upgrade from XML to GT2 which means:

- I changed from the original XY Blocks to Lars ones (https://www.youmagine.com/designs/lars-ultralight-ultimaker-xy-blocks). This includes also new 12mm bushings from robotdigg.com

- I changed all belts and pulleys from XML to GT2 (ordered as well from robotdigg.com).

Basically, everything worked out in terms of assembling it. I changed the step values in the Ulticontroller, and it prints.

- Pullyes are no longer wobbling.

- The belts are no longer loose, but tight.

But I'm quite worried, because:

- Print quality did not improve, at least I could not find any improvements in the parts I printed so far.

- The print head is quite unmovable by hand. There is a huge force required to move it by hand. I have to hold the printer with one hand, otherwise I move the printer and not the print head when applying force to the print head. I did the axis alignment with the axis suqare tool (https://www.youmagine.com/designs/ultimaker-xy-axis-square), the axis are well aligned. When I loosen the screw which holds the XY bushing, the situation is improved, same for the two screws which hold the 6mm rod. If I move one XY block instead of the print head, less force is required, but it's still way to much.

Do you have any reccomendations? I mean, I have to thighten the screws right? Mechanically, the reason for the problem is quite simple. But I just don't know how to practically improve the situation?

At long sight, it is not a good idea to have a print head which can't be moved by hand?

Thanks for your help.

[jhertzberg 18]

- The belts are no longer loose, but tight.

How tight? IIRC, there was a video on the old instruction wiki where the user strummed the belts to test the tightness by musical pitch (I think it was an F or F#). If your belts are playing a higher pitch, then they are too tight.

[neotko 1,362]

I did the mod on 3 umo+. Two move smooth. One moves tight no matter what I do. Weirdly enough the tight one prints a bit better than the other two. Check if the belts of the pulley motors move on the middle or if they are getting friction from touching the pulley sides.

[skewcrap 3]

@ jhertzberg

I don't quite understand what you said, the length of the belts should be ok, I'm not the first one who tried the ones from robotdigg.com. The old probably were in use several 1000 hours, so I would assume it is normal that the new ones are a bit tighter than the old ones?

@ neotko

I will check that. But I can hardly imagine that this is the single reason having to apply that much force to move the print head.

[neotko 1,362]

Oh also. The tension to hold the bushings shouldn't be too tight. They move over oiled rods, so they just need 'enough' tension, but don't overtight or the rods will suffer tension and will get scratchs.

[jhertzberg 18]

@ jhertzberg

I don't quite understand what you said, the length of the belts should be ok, I'm not the first one who tried the ones from robotdigg.com.

 

Belts are sold in many lengths, and if you bought 302 tooth belts instead of 303 tooth, you would have belts that you could stretch to fit, but would be too tight.

[skewcrap 3]

I have GT2 Belts, 303 teeth, 606mm long, 6mm wide, closed-loop. Those belts should be fine?

[jhertzberg 18]

I have GT2 Belts, 303 teeth, 606mm long, 6mm wide, closed-loop. Those belts should be fine?

 

They should be. I assume you are not using belt tensioners.

[skewcrap 3]

no, I'm not using belt tensioners at all...

[neotko 1,362]

Check if the belt/rods are aligned

[skewcrap 3]

They are. I mean, the position of the belts is basically determined by the XY blocks. You have to position the pulleys the way, that the two pulleys and the xy block are in one line right? I don't have a toold, I aligned everything by eye.

[skewcrap 3]

Update: After >100h printing time, situation improved a lot. Basically I did not change anything, but the print head moves easy now.

[lars86 36]

I'm glad things improved!

The two likely suspects are:

-Bushing deformation

-Cross-shaft squaring issue

On the bushings, this is why I made the bushing clamp an independent piece of hardware. The bushing should slide into the XY block very easily. If not, you should sand the ID of the block. From there, you only want enough tension on the clamp screw to make sure the bushing doesn't slide out. This is VERY little force. Push the bushing back and forth within the XY block as you slowly tighten. The unintuitive part: you should still be able to move the bushing in the block without much effort! It just needs to take more force than the bushing sliding over the rod, which is very little. As an extra factor of safety, you can add a drop of superglue between the bushing and block once it's all together.

On the rod squareness, from my description on the XY block page:

"I actually discovered that my printed print-head was holding the 6mm shafts slightly out of square, so forcing them square to the 8mm shafts actually induced bind! You can simply loosen the belt clamps on two opposing XY blocks, move the print head back and forth by hand and reclamp."

I would recommend using the squares to set the relative position of only one rod. Then use the above technique to lock the other rod in position. Ideally, both being square, is also the lowest friction position. But I think allowing a very slight amount of out-of-square, to lower bind substantially, is a worthy compromise.

[lars86 36]

I might try to implement a new print head design which decouples the two bearings, allowing you to square both shafts, then lock the print head in position last.