long axis drive

[sinjin]

Hi guys, new to this forum area. Just curious as to what people are using for a drive in the long axis. If i was to make a bed size of 1300 x 2500. What are you using in the 2500 axis. Ie rack and pinion or syncro belt. If you are using rack how are you controlling the back lash in the gears.
Also from what i have read just about all are using stepper motor design. Is this because servo drive systems are to costly? And while i am at it is anyone designing for 3 axis to do 3D? Or more axis or does it hike the price to much.
Thanks

[rodm]

Hi sinjin,
R&P is the way to go with long axes and the rack and pinion is designed in a way to eliminate backlash.
Steppers are much cheaper for the home workshop and you do not need to be a rocket scientist to get them running.
All of our 3 axes machines are capable of 2.5D. Almost there but not not quite 3D. The best way to undertand 2.5D is to imagine cutting your head from the top down. It will shape everything on an outwards path but it cannot cut inwards in areas like the underside of your nose.
A true 3D is a five or six axes machine and just getting software and drawings costs heaps.
We sort of simulate 3D with our 2.5D designs.
See photo

[sinjin]

Hi Rod, and thanks for the help. I am actually designing a big system for work 12m x 9m x 2m.
I do the mechanical design side of things. The controllers cost in the order of 130k for a Siemens drive. This is a 5 axis system we are talking about.
I was really impressed by what people are doing in their back yard with simple tools and getting great results.
What really shocked me is how cheaply you can get ball screws and linear bearings.
So i thought i might have a go at a machine for home but a stepper drive system to reduce the costs substantially.
Most of the systems guys have made are using 1/4" collet routers, are there any who are using 1/2" collet machines. As for the steppers can you get fairly high torque systems.
Thanks Rod

[rodm]

Hi Sinjin,
That is one big machine you are designing.

You can get much larger motors than the Nema23 we use. I just had a look at a suppliers chart and a Nema42, 6 amp and 2830oz/in holding torque is available - sorry I really don't know if that is suitable for a machine but the numbers are much bigger than I am using.

Our machines are built for price and availalble space. I would love a 2400 by 1200 cutting envelope but it would take up a lot of space in my shed. A lot build a sort of prototype machine to get going and then improve on that or build another machine with the protoype. It is addictive for some (read me) and I have been building machines for three years now. Gantry routers mainly but I have converted a couple of mills and helped with a lathe. My first machines were built from second hand linear components through ebay but now I use new components from China at much the same prices.

Yes I have seen 3HP routers on machines. You need to find a router that can be mounted though as most have handles that get in the way. Spindles are now coming out of China at not much more than a good quality 3HP router. A 2.2kw spindle with a 13mm capacity collet chuck and 8k to 24k variable speed can be had for $800US. These can be hooked up to operate through the software and run considerably quieter than a 3HP router.

Another factor is that the software is now affordable as you can get 6 axis machining software for $159US now - Mach3. This is a one off cost with no maintenance charge
and the software is continually being developed.

[chrisb691]

Hi sinjin,
<snip>
The best way to undertand 2.5D is to imagine cutting your head from the top down. It will shape everything on an outwards path but it cannot cut inwards in areas like the underside of your nose.
A true 3D is a five or six axes machine and just getting software and drawings costs heaps.
We sort of simulate 3D with our 2.5D designs.
See photo
A great explanation of 2.5D, thanks Rod. :2tsup:

[sinjin]

Thanks for the tips again Rod. The machine i am designing will actually go into a Wind tunnel. The Z-axis weighs in at roughly 600kg on its own.

We also have a small gantry router we use at work which is capable of taking an 8' x 4' sheet. Its a servo drive unit and it can do 3D. But the z axis is only 100mm. And really the guys only use it and 2 1/2 d machine this machine is a 5Kw unit.
The machine for the long axis, 2440 is using synro timing belts about 100mm wide on both sides on the machine. The synro belts are only what i would call ok.
Its way to easy to tension one side up a fraction more than the other. run the machine for half an hour and find the tensions have changed.

When i was looking into techniques for eliminating backlash i found a few different ideas. Rack and pinion with 4 motors (2 per side) running a tiny amount out of sink to always have teeth engaged. Another was to use only one motor but the axis of the pinion gear fractionally off axis to the rack also to make the gears constant engaged.
One system have seen of recent times was to use a ball screw which is tensioned up a couple of thousand kg to avoid shaft whirl at high speed.
I haven't tried to price something like a 50mm shaft 2500mm long for that application but can only imagine cubic dollars.
There is another idea i have only heard about using a ball screw in the long axis was to have a dumb traveller supporting the shaft.
But for me if i have a go at designing one for home it will be rack and pinion for cost reasons. Once again thanks for the clues Rod

[twistedfuse]

Sinjin,

In regards to Rack and pinion, ive been doing a bit of research since i am building a large axis router and haven't used R&P before so it was something i needed to reasure myself on. In large scale routers (commercial) the higher accuracy grade cnc machines will use helical R&P these are designed much like the standard ones except it engages more teeth and like rod said these are designed for anti backlash. To remove more backlash in the even higher quality R&P cnc machines, they use what is called a split pinion (http://www.atlantadrives.com/splitpinion.jpg).
You can see that it is designed to engage one half against one side of the teeth, and the other is designed to push on the opposing side to remove the slop (if any).

I'd say dual R&P/Motors would prevent racking and eliminate almost all backlash and it should be suuffice, just imagine the price of having to pay for 4 drivers just to control your longest axis.

Daniel

[sinjin]

Thanks for the pic, i had seen on a couple of lathes with the idea of the split pinion before and i thought its a great idea.

Buy chance have you done any costing for the helical rack. I have spring loaded my drive pinion in the past also for engagement slightly below the pitch line.
On a spur rack the rack only imparts 2 forces. A seperation force depending on the spur gear curve ie 14deg or 19 degs. And then the radial force and so if you know the torque of the motor you can work out the spring force needed.
For the Helical there are 3 force, Seperation, radial,axial.
For a small drive system maybe its not much of an issue but if you are saying looking into large stuff it maybe. Then again if all your beds have been machines on a large Mill it won't be hard to set up.
For those in Melb i know of a Mill at Parken eng that has a bed length of 6m x 1.5m i think it is. Its and old manual machine still does a great job.
Thanks for the help