Belts, tensions and improvements
Somehow I managed to damage one of my Y belts, so got a roll of steel core GT2 9mm replacement and read up on belt tensioning techniques. During this, and whilst it may be only a minor hassle, the process of creeping the belt one tooth more or less in the clamps, holding it all still whilst tightening the lock screw bugged me.
I saw various others had been similarly bugged, so the task was born… Adjustable belt clamps
After exploring and sharing various ideas, and enjoying the banter and humour that accompanied it, I settled on a design. I worked up the rotational geometry first, then designed the parts in CC (and copied them into VC too).
The requirements were:
- Avoid modifying the 3XL where at all possible
- First ‘threading’ of the belt should be fuss-free
- Screw adjustment of the final belt tension
- Be unaffected by vibration and debris during operation
- Be made on the 3XL
- Ideally not turn a whole block of Aluminium into chips - so cutting then bending sheet stock
- Cater for 9mm and 15mm belts if possible
- The same part for left or right Y fitment, and for X left or right too
But first, I had to take the plunge and cut some aluminium and find out feeds, speeds and clamping that works for me - with the standard Z and Dewalt spindle. Tests run, chips made, noises understood. Now I know I won’t wreck things.
Bought some sheets of 6061 Aluminium, 2mm thick and set about cutting out the profiles (CC files attached, and Vectrics too).
Reading up on Aluminium grades, after snapping a piece whilst test bending, it turns out that 3000 grade might have been easier to work with - however, annealing is also effective on 6061. Using the ‘candle soot’ temperature test (fascinating topic of its own) to hit annealing temperature, then folding the parts using a vice-mounded vee-bender (bought from Warco Tool Supplies in the UK). Happy with the results, given the bend radius achievable wasn’t quite as tight as I ideally wanted. Stainless Steel 1.5mm could have been much tighter, but baby steps…
The small ‘arms’ are the adjustment mechanism - more on that later.
The initial threading and clamping of the belt is acheived with an eccentric cam above ‘teeth’ (2mm spaced, 1.2mm/60-deg vee carved grooves 1mm deep in the clamp body) this profile quite closely matches the profile of the GT2 teeth and was simple to cut on the 3XL. The cams were cut from a sheet of HDPE, two layers bonded together with Locitite Ionic Activated Cyanoacrylate adhesive (£6 on Amazon), to achieve 20mm thickness.
The arms are assembled by press-fitting two 4mm ground steel pins, one being the actuator that the adjustment screw pushes against, the other forming an eccentric diameter as the arms rotate, pulling more of the belt through the slot from the 3XL and thus tensioning it. I added a spare pin position just in case I wanted to adjust things later on.
The clamps are simply assembled with 2 cap-head screws to hold the adjuster and eccentric cam in place. The pics show overly long screws, but will fit correct length screws before final mounding on the 3XL. A little adjustment of the cam profile was necessary as hand bending the aluminium didn’t give a precise wall ‘height’ which the cam is dependent upon.
The 3XL frame has to be slotted to allow the belt to pass through, so I drilled and filed a 4mm x 16mm slot taking care not to damage the extrusion surface. My arms look like Popeye’s now, but the result is worth the effort (and sweat). This is the only mod necessary to the 3XL and won’t affect rigidity at all.
Mounting the clamp now only requires removal of one plate/extrusion screw, and the original tensioner screw, then re-using the plate screw to hold this new clamp in place. Pass the tail of the belt through the slot, over the top of the new adjuster arm, and under the eccentric clamp. Pull the belt tight by hand and close the eccentric ‘snail-cam’ clamp. With a longer screw passing through the original adjuster screw hole in the 3XL chassis, tension the belt with it as it pushes the adjuster arm open. Use your preferred method (pencil, ping or sound test) to achieve the right tension and ensure the adjuster screw lock-nut is tight.
I know the holding force of the snail cam is good, I couldn’t move the belts when pulling on a scrap piece with pliers - the GT2 belts are designed to hold by tooth to groove lateral contact, not from compression force. The concern over vibration loosening the adjuster arm was overcome in the final design by using a locked-off screw to move the arm. I did for the final assembly of the clamps add a HDPE sleeve over the arm pivot screw to avoid thread contact damage.
Just finishing off some Christmas promises on the 3XL (inlays and trivets from Bamboo - must have caught @Julien’s bug for these…), then I will finally fit and show these clamps in situ. I can’t wait as overcoming the original ‘fiddly clamp bug’ has resulted in these quite-nifty parts.
What a fantastic machine. And a fantastic forum community too - the initial ideas were critiqued and from that I could make improvements. If you fancy making some, would be happy to share the trials and tribulations as I experienced them - either way the geometry and action works however you choose to actually fabricate the parts. And who knows, someone at Carbide 3D might be watching with interest too for a future product enhancement Would be my gift if adopted.
Files for Carbide Create and Vectrics VCarve attached and on Cut-Rocket.
Belt Tensioner Clamp Flat Profile Rev-2.c2d (238.8 KB)
Belt Tensioner Snail-Cam Rev-2.c2d (16.6 KB)
Belt Tensioner Snail-Cam Rev-2.c2d.zip (3.1 KB)
Belt Tensioner Clamp Flat Profile Rev-2.crv.zip (357.2 KB)