What's the Best Belt Tension?

[kbob]

We do not know what the best tension on the Rostock MAX's drive belts is. Some people have belts so tight that they are wearing out or stripping teeth. Gene's Official Authoritative Manual says, "Play with it a bit and go with your instinct on it."

Let's see if we can brainstorm a real answer here.

I'll start with two approaches.

This document has a theoretical analysis of dynamic tension starting on page 5. I am probably oversimplifying it, but it basically says that the static tension needs to be just enough that the dynamic tension on the slack end of the belt never drops below zero.

See the paper for a detailed formula to calculate dynamic tension (equation 16 in particular) from the carriage's static mass, max acceleration, and belt friction.

An alternate way to get there is to start from the motor. The Wantai 42BYGHW811 has a max torque (holding torque) of 4800 g·cm. A 20 tooth GT2 pulley has a radius of 6.4 mm (is this right?), so the motor can create T / r = 7500 g = 16 lb. That implies we need 16 lb. tension on the belt. (Yes, I just confused mass and force. Wantai started it, so there.)

16 lb. is much higher than the tension on my belts.

And I'll throw this one out too. Starting on page 139, there is an algorithm to calculate desired tension and a method to measure tension by deflecting the belt. I haven't studied it. http://www.cad.sun.ac.za/catalogs/Machi ... l17195.PDF

[ApacheXMD]

Calculating belt tension from stepper holding torque like you're doing doesn't seem right.

Belt tension in our case is acting on both sides of the pulley so they effectively cancel each other out. The tension isn't the load on the steppers. The cumulative mass of the carriages, arms, effector, and hotend is the load.

I say just tighten the belts until they're no longer floppy, then tighten is a little bit more. Very unscientific, I know, but you're probably overthinking it.

[kbob]

Yes, I'm probably overthinking it. The primary reason I care is so that the Rostock MAX experience can be more repeatable. It bugs me that people are stripping belts. It bugs me that overtight belts might have caused cope413's motors to skip steps.

If we had some kind of guideline -- "put N lb. of force on the belt, then tighten the screws", there would be more happy MAX campers and fewer threads in the troubleshooting forum.

That's why I brought it up. I'm pretty sure my own belts are fine.

[cope413]

I was thinking the exact same thing. There really needs to be a clear, definitive answer to the question of how tight the belts should be.

I will also start thinking of ways to measure/test belts. I've got a few ideas...

[Polygonhell]

FWIW the usual way to measure belt tension is to measure deflection at the center for a given force.
Another way you could quantify it is the force used to raise up the idler pulley.
One of the reasons it's hard to quantify with the Rostock is the long belt runs, there will still be a lot of deflection even when they are sufficiently tensioned.

FWIW when I was messing with mine there is a pretty well defined point where there is obviously more resistance when moving the carriages. The reason for this is the bearings in the steppers are going to be rated for a maximum side load and if you exceed that they will bind and eventually fail.

[ApacheXMD]

I agree that there should be a baseline

I tried measuring the note a tight belt makes using one of those guitar tuning apps for Android, but couldnt get repeatable results.

I then gave up and just tried to evenly "tight enough".

Measuring force on the top idler might not work because at least on my machine the idler fits pretty tight between the two melamine pieces on the top.

Maybe measure deflection while pulling the belt outwards with a mini scale to measure force.

[foshon]

I use the plier roll method. I grab the end of the belt as it comes out of the clamp. Then I roll a pair of pliers (the stripper/crimper type w/ bolt cutters) as far as I can. Then I tighten the clamp. I have done it this way on every printer I have ever had, always worked for me. My tensioner/idlers have never left dead bottom.

[Eaglezsoar]

I use the plier roll method. I grab the end of the belt as it comes out of the clamp. Then I roll a pair of pliers (the stripper/crimper type w/ bolt cutters) as far as I can. Then I tighten the clamp. I have done it this way on every printer I have ever had, always worked for me. My tensioner/idlers have never left dead bottom.

That is a good idea Neal. I'll have to remember that one.

[inventabuild]

I'm also wondering if anyone has a device printable or bought that makes it easy for one person to tighten the belts?

[Cleveralias]

I'm also wondering if anyone has a device printable or bought that makes it easy for one person to tighten the belts?

Barnett's design is working well for me: http://forum.seemecnc.com/viewtopic.php ... ler#p12448. I used a 5/16" x 1.5" aluminum standoff for the upper support though.