belts vs leadscrew

[DrStein99]

Does anyone have any experience with building linear rails using belts and lead-screws?

Because I am having real trouble working with lead-screws, in the cost - and also alignment, and not owning a lathe, I can't tool-out the lead-screws to fit bearings etc...

It appears to me that if I provide enough tension on an XL belt, it seems strong enough to hold- in addition to being a whole-LOT faster.

I was wondering if I could get some words of advice, and experience please. I think I can save myself alot of work and money if I just used belts instead.

[keyne]

Hi,
there are a number of belt drive users here, ynneb for example. Axxus also built a machine with belt drive (see ynneb's gallery). Also look at Tim Koene's video of his commercial belt driven p&p machine.
Belts will be fast but for direct drive i think you will need big servos. (i am planning such a machine, 500W servos, 25mm HTD5M belt).
Rack and pinion is another alternative... i may go for this on the Z axis (200W servo).
Belt tension stresses the motor bearings..
Max. torque is limited on belt drives (depends on pinion size, belt tension, belt width, belt style,....) but should be good enough for a wood router.
No hands-on experience from me YET but maybe this helps you further anyway.

[andy_ck87028]

Hi

I have been designing around belts on my X axis for a long time now.

I think the critical factor is around cutting strength. Threaded rod will take more punishment than belt and so if wood and light aluminium is your aim then belt should be fine.

Precision is also a factor with belt and therefore gearing down is mandatory. I am planning on 5:1.

In order to give myself maximum cutting strength I am going for the widest belts I can afford. I suspect (but don't know) that kevlar, steel, fibre glass belts are still prone to tooth shearing which is why I would a) go for widest belt and b) maximise number of teeth engaged to pulley.

A benefit of belt is that there will be no whipping. Therefore, if you plan to build a machine 500 metres, you just need a belt 500 metres long! (exaggerated for example ony)

Another benefit is transfer speed as there is no whipping and your limit is then the torque curve on your motor.

I have been vacillating between belt and threaded rod for a long time now but ultimately I think I have to be clear to myself about the intended use of my machine. It is easy to try and create machine that is both a racehorse and a donkey but unrealistic in the end. A lot of my thought processes have been wasted because I keep delaying on the decision re. the racehorse vs donkey.

I am sticking with belt for this machine (I think ... )

Andy

[keyne]

Alot of commercial belt drives list their accuracy as 0.1mm (not including accuracy of the motor). Therefore i am not sure if you gain much by having motor resolution much better than that. I am planning for direct drive with a servo and 0.075mm per step, but i hope with a better encoder i can get half that.
Was at a huge industrial exhibition today and saw lots of belt drives on FAST robots. Several were direct drive.
My machine will have only about 100N average cutting force, but it is a fast wood router for light cuts.

[DrStein99]

I was trying to resolve the headache of dealing with alignment, backlash, fabricating end-support structures, motor coupling, and turning-down the leadscrew shaft (since I dont own a lathe and am at the mercy of finding help w/ that).

I felt that pullies would be a more simpler setup; if I can manage to tighten them enough, and provide enough motor power - they pretty appear to me like a self-centering, flexible rack and pinion.

I suppose I could/would be able to get it to work; as it seems now to me the factor that determines success isnt the equipment, rather the amount of time and effort I devote to the construction.