I’m thinking of using an idler to bring the belt down to the motor. This will allow me to use one belt length tensioned in the xy block.
Is there anything I’m missing here that means this is a bad™ idea?
I had never thought of doing it that way. I can’t think of any particular reason why it would be a “bad” idea. Let us know how it works out. Should save money on an extra belt and pulley.
Only issue I see is the cantilever on the idler. I haven’t ever been a big fan of it because over time they always seem to migrate with the force causing them to ride to the edge of the idler and rub.
Only a few teeth will be engage the pulley in the corner of the picture. You will tell us whether this is a problem or not.
Yet another choice could be to set the motor in the middle of the axis instead of at one corner and use to two idlers, same belt distance, one more idler, more teeth in both toothed pulleys.
I quite like that idea of certainly a possible variation if the number of teeth causes problems. I will be moving the idler higher up to be more in line with “original” line of the belt in the next mockup.
A hand drill and a piece of sandpaper.
@Richard_Mitchell That’s what I was proposing above but too lazy to draw (but I suggested to put motor mid-way in the axis, with your way it can be kept in the corner).
Still I have the doubt about having two different belt lengths on each side and how that will affect getting them similar tension.
@Miguel_Sanchez
The length is actually pretty similar x and z. The main issue with putting multiple idlers is going to be clearance of the xy block I think. Yet to play with it and work out a block design.
@Dale_Dunn
May play with a crowned idler instead of guide washers for the fun of it,
I am a big fan of idlers, I think they give you a lot of options for power-train design.
In the same vein as above, this is the same thing, but packaged differently. The idler can be as simple as a piece of pipe sheathing an M5 screw - it’s just to change direction, which makes for a lot of options when you want to fit into a small space.
Same bits, different packaging:
The first one looks a bit of a tight radius for the belt although I doubt if that’s actually an issue.
I had thought of the second idea but it’s quite a bit more belt and I’m trying to put 4 motors, PSU, RAMPS and raspberry pi underneath the floor panel already.
Be interesting to see how it works in practice.
For reference found some documentation on GT2 belts. In short I think 624ZZ or bigger is ok for the backside(snigger) idler and tooth grip should be OK(ish).
http://www.sdp-si.com/D265/PDF/D265T056.pdf
@Dale_Dunn “Backside idlers should be flat and
uncrowned.”
@Miguel_Sanchez “The
designer should make sure that at least 6 belt teeth are in mesh on load-carrying pulleys…In order to minimize the potential for belt ratcheting[tooth jumping], each loaded pulley in the system
should also have a wrap angle of at least 60°.”
@Mike_Thornbury “Flat backside idlers should be
used with diameters at least 30% larger than the minimum recommended inside pulley size.” [OD 12.7mm for 2mm GT2]
@Richard_Mitchell , in this application, it looks as though you could make this into a front-side idler by adding a twist to the belt.
@Dale_Dunn If did that I’d probably be back to needing an pulley as the minimum OD for a flat inside idler is 25.4mm which is much larger than 12.7mm for a backside flat idler.
If your drive pulley only needs 60deg of contact, packaging becomes even easier. My M5 idler is 10mm, not far off the recommendation in your reference. I also find it coincidental that the sizes given are 1", 1/2", 1/4" equivalents… how much R&D went into deciding that 12.7mm was correct and 11mm wasn’t? I didn’t read the doc, is it from Gates?
From Gates’ GT2 Engineering design guide: “[Idler] Diameters should not be smaller than 1.3 times the smallest recommended inside sprocket size.”
That’s similar to the docs I found but they differentiate between inside and outside idlers. They say 30% bigger too but for smooth idlers they have to be bigger for inside idlers because of the tooth profiles. I’m only going to have 4.5 teeth in contact on my 90° turn but I don’t think we’re anywhere near the torque limit.
Not even close. The 5mm 2MR belt can take horsepower 
Look for ‘Gates GT2 Design Guide’ - about 150+ pages of tech application details.