Welcome to the SeeMeCNC Community

Since I don't know anything about pulleys, timming belts and stuff (rostock is my first project of this kind) I want to have better control over belt tension, So I managed to made my own belt tensioners which uses orginal not used holes of rostock. This gives me two things:
1. more accurate tensioning
2. leveling the pulley which helps keepping belt in the center. This helps to smooth moves, it changes when belt hits the wall.
Here are the pictures. As you can see I also stick teflon tape to reduce friction if belt try to hit the wall. By using tensioners i can keep belts in the center on the top pulleys. The position in bottom ones and on the shaft is another story. There is more factors (position of the belt clamps, angles of two bottom pulleys, angle of the shaft ).

I hope this can help somebody. (STILL DON'T KNOW HOW MUCH TENSION BELTS SHOULD HAVE )

I might give that a try today. I wish I could help with how much tension to set, but I'm not too sure either. I have it tight because i'm afraid of slipping.

I've been using timing belts on CNC equipment and printers for awhile. I get my belts at sdp-si.com and they have a great reference library. Belt Tensioning is item #8. The MAX has GT2 2mm belts 6mm wide. You can get the details on pages 51 and 52. These are registration drive belts for positioning (bottom of page 51). That means you do want them on the tighter side to minimize backlash. Even the experts say "Tension values for these applications should bedetermined experimentally to confirm that desired performance characteristics have been achieved." But here is some guidance using this information:

Since the belt is confined in the extrusion slot, you won't be able to do a traditional deflection test by pushing on the belt. So I home the carriages and locate the mid point of the belt. Use something to reach in under the belt and pull it out - I use a big paper clip. Then I pull on the belt. Now, you have to use a little judgement. If you can pull the belt out 1/2" or more using a fair amount of force (roughly 6 pounds), it is too loose. If you can barely move it 1/8" it is way too tight. This is a loooong span so you should expect some deflection.

I've been thinking about this same thing, since I've been having issues getting things right regarding tension and centring. Looks good.

can you tell me what you used to make these? parts numbers?

mhackney wrote:I've been using timing belts on CNC equipment and printers for awhile. I get my belts at sdp-si.com and they have a great reference library. Belt Tensioning is item #8. The MAX has GT2 2mm belts 6mm wide. You can get the details on pages 51 and 52. These are registration drive belts for positioning (bottom of page 51). That means you do want them on the tighter side to minimize backlash. Even the experts say "Tension values for these applications should bedetermined experimentally to confirm that desired performance characteristics have been achieved." But here is some guidance using this information:

Since the belt is confined in the extrusion slot, you won't be able to do a traditional deflection test by pushing on the belt. So I home the carriages and locate the mid point of the belt. Use something to reach in under the belt and pull it out - I use a big paper clip. Then I pull on the belt. Now, you have to use a little judgement. If you can pull the belt out 1/2" or more using a fair amount of force (roughly 6 pounds), it is too loose. If you can barely move it 1/8" it is way too tight. This is a loooong span so you should expect some deflection.

Do you have the part number for the belts on SDP-SI? I searched this board using Google like you taught me and could not find the part number. SDP-SI.com is the worlds worst web site for finding something if you don't have the
the exact part number you are screwed. Search for GT2 - nope. Search for 6mm belt - nope. Silly me, I actually thought a could find it on that site. Thanks for reading my rants.

Carl

https://sdp-si.com/eStore/PartDetail.as ... roupID=343

You need about 6M for all 3 belts.
The secret on SDP-SI's site is to select Metric before searching for anything even vaguely metric

Thanks for the part number.

Carl

Hey guys, they changed that part number on the belts, will this one work?

A 6Z16MC060 at
https://sdp-si.com/eStore/Catalog

I would buy this assembly (I'm not a machinist) if there was a 3rd party source. I just adjusted mine this morning. First time in about 50 prints.

mhackney wrote: Since the belt is confined in the extrusion slot, you won't be able to do a traditional deflection test by pushing on the belt. So I home the carriages and locate the mid point of the belt. Use something to reach in under the belt and pull it out - I use a big paper clip. Then I pull on the belt. Now, you have to use a little judgement. If you can pull the belt out 1/2" or more using a fair amount of force (roughly 6 pounds), it is too loose. If you can barely move it 1/8" it is way too tight. This is a loooong span so you should expect some deflection.

I was trying to set the Z-height on my MAX last night (just got to that point in the manual) and after setting things in the EEPROM, when I homed again and then brought down to Z0, the head hit the bed. I had to take another .4mm off the height. So my thinking was that perhaps there's too much play in the belts (having never adjusted a belt before I wasn't sure how tight they should be. Searching found this thread. I followed the instruction above and realized that I can *easily* pull out 1/2" and without much force can get to 1 1/2". So obviously my belts are too loose

Does anyone agree that that would cause my Z Max Height calculation to change, due to slop/backlash?

mhackney wrote:If you can pull the belt out 1/2" or more using a fair amount of force (roughly 6 pounds), it is too loose. If you can barely move it 1/8" it is way too tight. This is a loooong span so you should expect some deflection.

I usually get one belt exactly where I want it, and then I pluck it like a guitar string and just tension the other 2 belts until they play the same note

Found this on Thingiverse.
http://www.thingiverse.com/thing:146815
I haven't tried it yet (though I plan to), but it looks like it would need a whole lot less hardware.

That's me

I modified barnett's design, found in this older thread here: http://forum.seemecnc.com/viewtopic.php?f=54&t=1860

After taking some slack out of my belts by pulling them through a bit more, these really helped get them tight - and centered.