One ring to rule them all?

[irving2008]

Listen to Jazz on motor drivers.. nothing you can develop with an Arduino and an off the shelf motor shield will be useful. The typical Arduino motor shields are toys, good enough for a NEMA17 stepper for a 3D printer or a small robot, even the so-called high-current ones are limited in reality because they don't have big enough heat-sinks and are restricted to 24v. For the sort of machine you're contemplating to perform at any useful rate, those screws will require big NEMA23 or even 34 size steppers, and the driver will need to deliver phase currents of 4 - 6A @ 60-80V
.
Plus an off-the-shelf stepper driver has complex algorithms for anti-stall, anti-resonance, fault-alarms, etc. You won't get close to anything useable developing your own software or using open-source stuff for several months if not years of development...

[Rufe0]

Bah Fred can only do C5 ballscrews upto 1m long. Says he can do C7 longer but I think its probably worth paying abit extra.

[dodgygeeza]

Bah Fred can only do C5 ballscrews upto 1m long. Says he can do C7 longer but I think its probably worth paying abit extra.

You can pick up C5 and C3 screws up in the UK.
I know that Zapp have C5 screws.

[JAZZCNC]

Bah Fred can only do C5 ballscrews upto 1m long. Says he can do C7 longer but I think its probably worth paying abit extra.

What machine are these going on.! . . Router or Mill.?

[Rufe0]

What machine are these going on.! . . Router or Mill.?

Both really

[Jonathan]

Bah Fred can only do C5 ballscrews upto 1m long. Says he can do C7 longer but I think its probably worth paying abit extra.

If you compare the errors due to the screw grade to other errors in the system (e.g. due to the ballscrew and associated stiffness's), your conclusion will likely change.

[JAZZCNC]

If you compare the errors due to the screw grade to other errors in the system (e.g. due to the ballscrew and associated stiffness's), your conclusion will likely change.

The Boy cum-eth Man just said what I was eluding too.! . . . . .Screw grade will be the least of your troubles.
Also let me just say forget having a Router and milling machine in one they just don't play well together, very very difficult to achieve and keep your sanity. Even then the result will still fall short of average milling machine or Router.!!

Keep it simple do one or the other and save your self some dissapointment.

[Rufe0]

No choice in the matter, I need 3 machines but only have space for 1.

Not sure what errors your talking about, with the raising bed I will be able to lower the Z travel to almost nothing 99% of the time.

I will be using servos all around or steppers with encoders, so hope fully I can keep the bed level, if not i've seen linear encodes relatively cheap on ebay.

I would like abit of advice on the bed though. Do you think just 4 rails and 4 screws will do it or would it be worth adding say 2 in the middle? I can't imagine the bed itself would deflect much, not when it's made from 100x100 box.

[routercnc]

Came across this recently - it has 1m Z travel. I can't quite make out the gantry / Z axis arrangement, probably double beam gantry with box Z inside?

But it is also 58 tons so not going in that transit!

http://verticalmachiningcentres.blog...ng-centre.html

What Jonathan means is that total stiffness as seen by the cutting tool is everything between the end of the tool and the bed - the long way round! By this I mean back up the spindle, any offsets, gantry, ballnut stiffness, ballscrew axial stiffness, ballscrew AC bearing stiffness, frame stiffness, bed stiffness, clamping arrangement stiffness, and the stiffness of the part being cut. It all adds up.
Are you going to lock off the bed? If it is just held on the 4 ballnuts then there is 50 microns vertical freeplay right there. Or is it going to be very, very heavy?

[Rufe0]

Ha yeah thats a monster.

I decided to make it 2000x1900mm I'll just have to get it trailered away if I ever need to move it. 1900 will just squeeze through the garage door lol. And I think i'll probably make it only 500mm Z travel. The raising bed will make up for it I think. One thing I can envision needing todo in the future is deep pockets, where the length of the cutting tool and the width of the spindle would be the limiting factor, maybe make a super thin spindle if it becomes a problem.

I see what your saying about the accuracy but I may aswell do everything I can to make it better. If I get the poor quality screws thats just another inaccuracy to add on. I can't do anything about those other inaccuracies but I can about this and it wont cost much more so why not?

Not sure about locking it, maybe once i've got the machine built I can use it to make some locking devices. I've no way to make them especially with any accuracy at the moment. Not sure it will be a big issue to be honest, the weight of the bed will take up the ballnut backlash. I guess the main benefit would be reduced power consumption you'd get from turning off the motors. I could also use double ballnuts which would probably be a much simpler solution, the ballnuts I should be able to get cheap from fred.

I had a question about the gantry design though, need to get it designed so you lot can see, I've been trying to learn how to use fusion 360,solidworks and sketchup but I just can't get along with any of them. Give me AutoCAD 2000 any day of the week over these, I don't know how they can take a program thats so simple and works so well and thoroughly ruin it.