After reading the countless posts about belt jams and seeing the endless carnage from the picture I decided to investigate the current design. I have dealt with current sensing closed loops before with some of our assemblies at my work and it certainly sounds like we are having false current spikes due to unwanted resistance somewhere in the design.
The first place I looked was the spool because when I was assembling it i noticed i could not roll up the full length of timing belt without significant interference with the idler gear. You can see the the overlap in the image below.
There could be almost 2 belt thicknesses being squeezed in between there once assembled and driven via the electronic.
Next issue is the belt finding its way between gears and essentially killing itself. With most timing belt applications its highly suggested to use pully flanges to keep the belt from traveling outside the pulley track. The current design does a pretty good job of sandwiching the belt however those last few coils end up not being supported on the top and bottom. I think they might be getting loose and falling between the gears.
Unfortunately the proposed solution is not the easiest to modify, it will involve the changing of some major components, but lets get into it. All modified parts or surfaces below will be in RED.
First off lets make some more room for the coiled belt, to do this i will try and remove the hub of the idler gear, easier said then done, the gears are very hard. By doing this the more belt can coil without interference.
Too capture the belt better i propose we add a flange to the space between the gear and spool surface, its modeled as a single piece below but im working on a disc that can press over the gear profile and seat fully in the same position, that should be an easy addition that can be 3D printed with no major modification. This flange also captures the idler gear and prevents it from falling off the idler post.
Seems to have the perfect amount of clearance already also, but the flange addon can be thinned if needed.
Now comes the major modification section, still thinking of a way to make an add on, so it doesnât have to be completely remade. In order to capture the other side of the belt we need to increase the diameter of the arm body some. Gluing an extension on might not be robust enough but could work, i will probably 3D print it at work to try it out.
From there some material will need to be removed from this feature to clear the other arm.
As of this moment, last step is to modify the belt guard for any interference and i really think the belt guard needs to create a single channel for the belt to travel through and not allow the belt to spread out/snake about within the arm assembly, i will take a crack at this lastly.
More to comeâŚ
