Belt Options for new Machine

[webgeek]

Hi all; I'm making a new machine that I want to be a combination of laser cutter, 3d printer, and router. I had plans to build three separate machines but I realized the overlap for the laser cutter and router is pretty substantial. The machine is going to be like 4' x 4' though depending on the price, I might just jump to 4x8 and be done with it. The laser tube will be on the gantry though removable when I'm routing.

The primary use of the machine will be either laser cutter or router so I'm focused on those. I'm trying very hard not to compromise either to get what I want. I'm using HiWin linear rails and DC servos from Keiling.

I was planning on using belts when it was just a laser cutter. Now that I'm adding a router to the mix it gets a bit more interesting. A laser cutter requires high speeds and a good bit of repeatability. The router requires a bit less speed but adds the requirement for torque. That's why I want to use servos so I can get the best of both worlds. I'm going to be stepping them down via a belt already.

My question is this: what tooth style and belt width is likely going to work OK here while ensuring the least amount of spring and stretch? I looked into ServoBelt and while it's beautiful, it's also a bit out of my budget for the size of this project. Worst case scenario I could machine something similar to ServoBelt for this one-off machine. Anyone done something like that? How is the lower belt bonded to the extrusion?

I'm very interested in people's opinions on using belt for something like this and any learnings they've had. I like the CNC Router Parts rack and pinion but I don't like the lack of bearings or limited precision it offers - several thousandths at best is what's documented.

Thanks much for any input!

-Mike

[robe_uk]

Well I think r&p will be as good if not more accurate than belts. Loads of people are having great success with r&p, also with belts for than matter too. You haven't said what you will be cutting, if wood a few thou is more than accurate. If you want mega precession you will just have to man up and buy top quality ground ball screws.
Personally I think people get caught-up looking for 0.00000000001 accuracy when its not really needed.
Search these forums as its all been discussed before and ultimately comes down to budget.

Good luck with whatever path you follow.

[louieatienza]

If you're using servos with r&p you can gear them down far more than with steppers. The CNCRP I believe is 2:1 or 3:1 depending on motor size. With servos, you might be at 7:1 or more... Even with just a 500 line encoder you should be able to get good resolution.

I don't know if putting the laser tube or collimator on the gantry is a good idea. Usually it's hidden away in safety. If you're doing raster laser cutting you'll probably need high acceleration as well as high speed. Maybe not so high if vector cutting. Usually this would mean making the gantry relatively light, which may not be good for routing. You'll have a lot of fun keeping the optics clean as well; those mirrors and lenses can scratch up just looking at them!

[ger21]

Worst case scenario I could machine something similar to ServoBelt for this one-off machine. Anyone done something like that?

I've built a prototype stepper powered Servobelt with 4:1 reduction, but haven't tried it yet. There are pics in the ServoBelt thread.

How is the lower belt bonded to the extrusion?

3M VHB tape.

[harryn]

The trick with belt drives is that most people (including me ) often start by design around the tooth strength rating rather than the belt stretch rating, and it gives entirely different calculated results.

If you start with an assumption, that 100 Lbs of force are required to really hold a router where you want it to be, add in significant acceleration, etc, it starts to be a large total force required.

Since belt stretch is (stretch per meter x meters of belt), then the longer the belt, the more stretch under a given load. For what I was trying to do, which is wood routing, it was starting to pencil in at 100 mm wide AT10 belt on each side of the table for a 8 ft axis, which is far more belt than anyone would normally think is needed.

At the same time, the common complaint / criticism of belt drive systems is "bounce", which is exactly what you get from too narrow of belt.

In both belt drive and rack + pinion designs, the motor must have high holding torque, as the total force is directly on the motor. In screw drive approaches, the screw is providing a considerable mechanical advantage, reducing the holding force needs on the motor.