What's new
What's new

Sanding Belt not tracking properly

Marco Schwan.

Plastic
Joined
Jan 18, 2020
Heyho!

(Click the link on the bottom for high-res pictures)

I'm building an oscillating edge belt sander right now and I have some major problems getting the belt to track.

Its a 3 roller design, with two fixed, stationary rollers and one roller which is responsible for everything else.

The two stationary rollers are made on the metal lathe, so I know they are perfectly flat, round, parallel and so on.

The Last Roller has a crown of about 2mm on it.


From the first video you can see that without the tracking roller the belt acutally tracks decently well. By applying a bit of tension on one end I can get it to wander quite nicely.

But despite the fact that flatbelts track towards the highest point, aka the highest tension, the belt wanders to the exact opposite. It wanders away from the point with the highest tension. I really don't get it. Its pretty frustrating to me at the moment,


Any Ideas about what is wrong with my setup?...

The belt is 200x2000mm (~8"x79") The big roller is 250mm diameter (10") and the small ones both 8cm (4")
IMG_9237.jpg
IMG_9236.jpg
LINK PICTURES: Belt sander - Album on Imgur
 
You are going to find that no two belts will track exactly the same. All my belt sanders have adjustment on the roller to accommodate this. You need adjusters on your fixed rollers. At least one should have them. Nice job on the sander!!!
 
Yeah, one of the rollers has to be able to be angled. Probably the small end one. Maybe turn those screws to run in slots, although that might be tedious. Changing it to be a pivot might make it a more reliable adjustment
 
I really don’t wanna make them adjustable since the whole point of the fixed rollers is to be always perfectly 90degrees to the table.
I’ve also seen this design commercially before so I know it works.
I can’t even remotely get it to track so I have bigger problems now
 
I really don’t wanna make them adjustable since the whole point of the fixed rollers is to be always perfectly 90degrees to the table.
I’ve also seen this design commercially before so I know it works.
I can’t even remotely get it to track so I have bigger problems now

One of the two end rollers will have to be crowned and adjustable or you'll never get it to track right. Being
in the middle of the belt and approaching it at such a shallow angle the third roller does nothing but tension
the belt--it's really superfluous to your needs. And I think you're also going to find that, even if you get it
tracking right, you're going to be adjusting it a lot because that wood frame isn't going to be rigid enough...
 
I really don’t wanna make them adjustable since the whole point of the fixed rollers is to be always perfectly 90degrees to the table.
I’ve also seen this design commercially before so I know it works.
I can’t even remotely get it to track so I have bigger problems now

When everyone says it needs an adjuster they are talking in thousands of an inch adjustments. You will never notice it on the table. Don't ever believe that those new belts are within a thousanth of an inch when they were glued together.
 
I really don’t wanna make them adjustable since the whole point of the fixed rollers is to be always perfectly 90degrees to the table.
I’ve also seen this design commercially before so I know it works.
I can’t even remotely get it to track so I have bigger problems now

Umm, it has to be adjustable, this really is not an option.


Adjustment does not need to mean sloppy
 
Get rid of the middle roller. Make the small roller at the end with a crown in the middle, .080 diameter larger in the center than on each side. Put the tensioner on this one as well as the tracking adjustment. then it will run true.
 
You can see if crowning one roller will work by applying a couple of wraps of tape to one roller. I have a worn-out 6x48 sander that works well with the tape on the drive roll. Recrowning the roll has become a some-day job since reapplying tape is so fast.
 
Last edited:
I've wondered if, instead of an adjustment on one end of the roller, you put in a compliant roller support. That way it would move in if the belt was too tight on the compliant end and would move out if the belt was too loose on that end.
 
The reason for the frustration is that the belt doesn't track towards the highest tension!

Ignore the third roller and imagine that you are tilting the left hand roller. If the roller is tilted so the bottom is tighter and the top is looser, the belt will track downwards, just as you expect. But the reason the belt tracks downward is that the belt travels around the circumference of the roller. When it leaves the roller, it is pointing in the same direction as the roller is, so it goes down.

When you apply pressure to the bottom of the belt in your experiment, you are lifting the belt, because of the position of the pivot, but not changing its direction very much, because the belt does not go very far around the roller. the belt goes aound the left hand roller in a straight line, but higher up, and the belt gradually rises.

It isn't an error, it's just not what you expect!


PS:- That is going to be a nice belt sander. :cheers:
 
I think most are missing the point that you want the belt to oscillate (or wander).
After watching your video I would try making the tension roller hollow instead of crowned, that way it may move up and down on its own. ( I think).
I have a 16" thickness sander and to make the belt oscillate they use air pressure to move the top roller forward and back from a centre pin which causes the belt to move back and forth on the rollers. This needs to be protected with safety switches to shut off the machine if anything goes wrong with the air system pressure. The belt will move into the micro switch paddle and turn off the machine if it oscillates too far.

You may have to make the tension roller taller to make it work properly but just guessing on that, some trial and error needed with this design. If it works it will be a nice simple solution. It may just find the centre and stay there unless it has some runout to begin with and is always trying to compensate.

If the above doesn't work then you would have to go with a mechanical movement of the tension roller to cause the belt oscillation. If you go to that trouble then the crowned roller will work just fine. A small motor slowly rotating with an eccentric wheel pushing on the tension roller might do the trick. Basically recreating what you were doing with the stick in the video.

Have you looked at the commercially available machines to see how it is done by those manufacturers?
 








 
Back
Top