Yet another "Y" axis design

I have thought about a design exactly like this.  The only issue I can think of becoming a problem would be backlash from stretching of such a long belt.  Last I looked they have belts that are very resistant to stretching of course.

Pretty soon after another part or two, probably a new board, I'll have everything short of motors for an entirely new printer.  I'm going to experiment like crazy on whatever design I can think of or acquire from other sources.  My favorite thing to do is think of ways to improve on the design and efficiency of a machine.  I'd love to get paid to do that.  If you don't beat me to it, I'll be trying this out and I'll post back.

Hi Pirvan ,    Bill D again    I drilled a 5/16 inch hole in each end of the Y axis pulleys and insert 8mm x 5mm x 3mm bearings also from Bocabearings.com (8 for $10) and also did the same for the X axis idler pulley. This stopped the wobble and squeaking noises and needing to adjust belts often . due to wear of the aluminum pulleys against steel bolts.

Single motor drive, probably in the back left corner.  Everything else is a idler pulley.  The 3 in the corners are technically sprockets as they have teeth to engage the belt, the inner ones where the back of the belt rides on them are smooth. 

As I mentioned in my original post, I would also add a spring loaded tension idler to take up any slack that may occur due to slippage or stretch.  Effectively creating a self adjusting/tightening system.

Mounting the motor vertically shouldn't pose any problems, but then again, these ideas I'm throwing out there are more conceptual than anything.  I wouldn't go about gutting out a working SoliDoodle to do this.  But if I ever get going on a brand new, scratch build, I would definitely consider something like this.

Yeah, I'm familiar with those designs.  In fact I spent about 3-4 hours staring at the "H" bot design to figure out how in the heck it works.  It's absolutely brilliant.

But the problem with those designs, especially the "H" bot, is that they will require a re-write of the code to get the motors synchronized.  Whereas a standard design like the Solidoodle (and even my idea drives the motors independently, the "H" bot design drives the motors so their output is differential.  I don't know if that is the proper term for it, but I don't have a better description than that. 

For example, with standard drive, when you move the carriage 100mm in the "Y" direction, you only tell the "Y" motor to turn "X" number of steps.  The "X" motor is not doing anything.  With the "H-Bot , to move do the same thing, you need to drive both motors simultaneously, but at different rates.  It's weird, but it works.

Suffice it to say, I don't have the knowledge on how to write that code, and as far as I remember, the guy that posted the "H" bot design never published the software for public consumption

As for the "Y" axis, it still travels on rails, it's just that it's driven by a single continuous belt.

Just uncomment the line in config

//#define COREXY

Yeah! No, i like how you did that. I wonder why it's not used often. I am not sure, perhaps that problem with slightly lopsided forces due to the different lengths of line. Maybe warrants searching the reprap forum. Anytime I think of something, I usually find some obscure post from 2004-2006 with a fanatic reprapper having already tried it....It's really impressive.

If you're worried about stretch in the belt being a problem, which is really the only one I can see, what about using spectra line as per the regular spectra conversions? That stuff just does not stretch... Also not expensive per m.

Fair call, that hadn't hit me as a disadvantage yet; I was hearing good things from other reports re accuracy, but may have just been in the melee of disappearing backlash!
Will spring-loading idlers cancel the backlash of the belt stretching, or add a new source of backlash? In one direction, the drive pulley will be pulling via the tension wheel before it gets to the first carriage, and in the other direction will be pulling the first carriage directly.
Also re tension, there are effectively two belts in there between the fixed clamping points on the carriage ends, so you'll have to keep those both tensioned and fairly even I think. See image.
If it was me, I'd perhaps keep the pulleys all rigid and tension the belts the normal way, it's still pretty effective.
Thoughts?

I dont know where people get the idea that the belt stretches. It does not. Belts are reinforced. What they DO have is slack. They dont want to be perfectly straight between pulleys. But if you pull them too tight, you could have problems with with the pulleys. Thats where the backlash is.

BTW, that is a bad place for your tension pulley. You add tension by pushing it into the other belt, which is teeth-to-teeth. You would be better off mounting one pulley directly on the motor shaft, and have a motor mount with oval/long mounting holes so you can slide the motor in order to tension the belt.

I like the idea of keeping a constant tension on the belt!

Here is something I have been pondering for a while; it uses twin lead screws and one driving belt.  My thought is that it would increase the accuracy due to the screws requiring MANY more steps per mm of travel.  Problem is I would have to put a screw on the carriage as well so both axes have the same precision... Your thoughts?

Got it...maybe a smaller idler with teeth deployed on the interior of the loop biased in one direction...

I was looking at your design, and you might want to reconsider your design.

If you're thinking of using a leadscrew with a tight pitch in order to increase the umber of steps per mm of travel, you'll find that the speed will dramatically affected to the point of being way to slow.  Leads screws are OK for the Z travel, but for the X/Y motion, you have to be careful what you pick.  There's a fine balance between speed and accuracy.

I had a similar design I put together a while back using ballscrews. 
http://www.soliforum.com/topic/2834/pre … m-scratch/
The high precision screws I have move 5mm for every 360° of rotation.  Our belt system moves about 36mm for 360° of rotation.  That's 7 times faster.  So to achieve the same performance with the ballscrew, I would have to drive the motors at 7 times the speed.

You would be in a similar predicament.

There is one way to get around this, create a gearing system that increase you gearing ratio on the output (I'm not sure if that;s the right terminology).  Anyway, you could put a larger gear on the motor shaft, driving a smaller pinion gear on the lead screw.  Your gear ratio would have to be such that it compensates for the difference between the belt and the lead screw. 

The drawback to this is that your torque may not be enough to drive the setup, requiring an upgraded motor.