Has anyone done a dual leadscrew on the long axis with a timing belt connecting them to the stepper motor?

Why would you want to tackle that problem??

The gantry wants to pivot around the current leadscrew. I have not been able to take it out with bearing/pipe adjustment. I figure two screws near each edge of the will correct that.

You might be trading one problem for another. Explore the causes of the problem first. Is the lead screw possibly a shosh off center? Could you increase resistance by increasing the bearing wheelbase?

If you mean that it is rotating around the axis of the screw, you need smoother-running nuts. Back off on the anti-backlash, and if that helps change the design. If you stay with draggy nuts, and ask the stepper to torque twice as many, that's another new source of grief.

Hi rweatherly

Plenty of people have done exactly that for increased rigidity of an axis. When you are using el-cheapo home brew linear guides there isn't a great deal of resistance to 'twisting' of the table/gantry.

As a test of how your guides perform twist the table/gantry with your hands and see how much it moves. I'm tipping its a couple of millimeters and if so there are two ways to fix the problem. One is to buy proper linear guides to dramatically reduce the potential for twist (in homebrew terms it is negligible) or as you have discovered use two screws spread as far apart as is practical on your machine. As far as driving the two screws you need to compare costs of new guides vs. two screws and the stuff required to drive them. There is always the option of slaving two motors instead of making a synchonous belt system.

Just for interest sake take a look at the Mori Seiki japanese website (its ok there is an english version lol) and look at DCG theory of machine building.

http://www.moriseiki.co.jp/english/index.html

then click on the blue bit at the right of the screen which says
"Driven at the centre of gravity". There is a cool presentation on DCG.

Its way over the top for homebrew CNC but still interesting reading.

Cheers
derekj308

Dual screw design, I am working on building my machine right now- it will have dual 1/2-10 acme, belt driven. I have some mock-up pics of the drive layout in my build log. I do forsee some issue with the screw alignment, I have designed in enough adjustment so that i should be able to get it set up well.

Remember that the motor will be seeing about twice as much drag in the system as opposed to a single screw- spec the motor- i am using a 640oz/in @ 50v.

you can use a system like this to keep the gantry from twisting.

Cheap, practical, functional. I may just use that idea on my homebrew machine. Thanks joe2000che.

Dual screw design, I am working on building my machine right now- it will have dual 1/2-10 acme, belt driven. I have some mock-up pics of the drive layout in my build log. I do forsee some issue with the screw alignment, I have designed in enough adjustment so that i should be able to get it set up well.

Remember that the motor will be seeing about twice as much drag in the system as opposed to a single screw- spec the motor- i am using a 640oz/in @ 50v.

with that type of arrangement will you end up with a backlash problem or at least some "slop " in the system.

I am contemplating building this type of system .With the nut drive situated under the bed a pretty hefty moment results at the gantry, which is then transferred to the guide rails, essentially the cutting tool load times the distance from the cutter to the Xaxis drive nut.The smaller the moment arm between the nut and the cutter [coplaner]the smaller the moment. I am thinking of designing the machine with two Xaxis drives one either side of the bed and sitting just above it. this should minimize the moment eliminating the torque on the black pipe guide rails.

My main concern is how to drive the two acme screws from a single motor, belts and pullies or a second motor slaved to the first .!!!

wouldn't all moment calc's be based from the rails to the cutter head? the machine could operate in a mode without lead screws, to demonstrate what my opinion is... in addition the majority of machines us 2 rails for the x-axis, i think Joes2006 machine using 4 will lessen the importance of screw placement... all this is my observation- feel free to agree/disagree... _jon

Hi Guys
Ideally the place which to apply the force to move the gantry/table is in the same plane as the linear guides. Plenty of people don't do it that way and have the screw in a less than ideal position and still achieve good speed and accuracy. In the case of Joes 2006 machine with the 4 pipes the ideal place to drive would be at a plane centrally located between the two levels of pipe. As Joe has pointed out in his diagram above you don't need to have the screw located centrally in the mechanism. By using cables to apply the force in the desired plane you can put the screw anywhere you like along the driven component of the gantry/table and it wont 'rack' since the forces on each side are the same because of the cables. The position it is in in the diagram above is ideal since it is as close as is practical to the cables line of action of force.

I would like to point out another consideration that can be made during a machine design is the position of the centre of gravity of the gantry/table being driven. During acceleration and deceleration a moment is generated about the linear guide due to the position of the centre of gravity (COG) relative to the linear guide. You can minimise this force by keeping the COG of the driven member as close as is practical to the linear guide. You could even add weight (he didn't say add did he, he's MAD!!!) to shift the centre of gravity to be in the same plane as the linear guides. If your system could cope with the additional torque and weight (deflection) it would allow much higher acceleration and deceleration of that axis. Of course the COG changes as other axis shift position but you can still design out the majority of the moment applied to the linear guides during accel and decel due to a COG not in the same plane with the linear guides.

Cheers
derekj308

hi Derek

I have been piddling with some gantry loads as I see it . Again I have not fooled around with this stuff for a very long time . so feel free to correct me .

I have tried to minimize the gantry moment loads on the black pipe/skate bearing trucks and have included a hand sketch of how I think the loads are distributed .

I do not think it is a good idea notching the side rails so that the bed crossbeams carry through it virtually halves the beam depth and as all loads end up in the side rails it will severly reduce their bending strength [varies with the depth cubed if I recall. The loads from the gantry result in outward bending load on the gantry side arms which could be strengthened if necessary with and outer web .

The vertical loads on the truck should result in a differential loads on the front and rear bearings respectively ,the sidewall will experience a reactionary load to the gantry side load as an inward force on the sidewall. The black pipe will try and deflect downward at one end of the truck and upward the other trying to force into a shallow "s" form locally. This possibility could be reduced effectively in my opinion by nesting the pipe between top and bottom angle irons bolted into the sde rails along the length of the pipe and if required each tack welded at intervals along its length [belt and braces if you get my drift.

So in short I think if the moment is minimized by making the drive axis coplaner with the cutter axis, or as near as can be [obviously the cutter plane varies with the X axis position .The deflections in the bed and tracks I feel could be improved by the suggested mods .

I mean this in a constructive way ,and heavens knows I have been wrong many times ,this thread and Joes efforts have been a great inspiration to me and I do not in anyway want to his efforts .....mjh

Mike,

Looks good, couple of comments. increasing L2 does a lot for stability and effectively reduces F2. Spreading L2 means the forces at the bearings will distort the pipe less due to load of the gantry.

Moving the cutter side bearing past the cutter will reduce the same moment diagram for forces generated by the cutter in use.

Increased L2 will reduce the racking moment.

It's worth doing the same analysis for the gantry in the other plane, Most home built gantry's are sufficiently braced and so can go out of square in both the horizontal and vertical.

Andrew

I just got my dual X leadscrew machine going. I went with two motors instead of a belt. It seemed easier to me and about the same cost with more power. I bumped up my HobbyCNC to control 4 motors and I then told mach2 to slave them and it works like a charm. With hand pressure, I can not rack it or stop it. Either way, it seems like a good solution to reduce gantry racking.

Steve

I just got my dual X leadscrew machine going. I went with two motors instead of a belt. It seemed easier to me and about the same cost with more power. I bumped up my HobbyCNC to control 4 motors and I then told mach2 to slave them and it works like a charm. With hand pressure, I can not rack it or stop it. Either way, it seems like a good solution to reduce gantry racking.

Steve

I imagine you created an identical setup on the rt. side of your machine.
Also You may want to place a simple box enclosure maybe with a hinging top around the entire length of the leadscrews to avoid dirt and material from gettings all over the leadscrews and also would help protect the rails from falling debre.

Joe

Joe, point well taken. Leadscrews on the sides of a machine need to be protected from swarf and other junk no matter the drive type.

Steve

The first machine I built, had a 1.5 m gantry driven from one side. The second machine I built is driven by one servo with a drive shaft the goes along the gantry and drives the gantry from both sides. You could do dual motor drive, but for some reason it doesnt feel as right in my mind, as a drive shaft connecting to both sides. Many have sworn that dual drives are good though.

Having had both machine to compare to, the dual side driven gantry is rock solid and makes a world of difference to the machine. Bracing the gantry etc will not help as much and will just put bad loads on your linear drives. A linear drive is not meant to stop racking, its sole purpose is to guide the gantry in a straight line.

The very first animation I ever posted on this forum was the one below. While it was done out of naievity by a wet behind the ears CNCer, many years ago. It may give you some ideas. (Possibly not, I just had to blow the dust off it.) (I should do an updated one)
http://www.cnczone.com/gallery/data/500/2793routmove.gif

Hi, as you can probably tell by the dates of the prior posts on this thread, I have been doing a lot of searching around for information.

BTW - this is for a home build router setup. It is not intended to be commercial duty, but I like fast cutting and want decent results on a 4 x 8 sheet - something like +/-0.010 in or less if possible / affordable.

The animation from ynneb above is exactly what I am thinking of doing, but of course with a few twists:

a) Long Axis rotation
- Imagine that the X axis shown in the animation is my Y axis
- Instead of being horizontal, the motion is vertical
- Yes, there really is a method / need to the madness
- This axis movement is approx 6 ft / 2 meters long

b) Y axis rotation / change
- Imagine that the Y motion in this animation is now the Z axis instead
- The direction is the same, just the router is facing sideways instead of the normal "down"
- This axis movement is approx 18 inches / 1/2 of a meter

c) The X axis is a moving gantry
- Details still in progress
- Setup sort of like a plywood cutting setup at a lumber yard.

Perhaps more simple - just rotate the picture 90 degrees clockwise and imagine the router going into the material when the motion goes left and right.

Some design assumptions

1) Belt selection
- Brecoflex AT 5 by 50mm belt (2 each as shown)
- Zero clearance timing pulleys

2) Gantry mass - approx 30 Kg / 70 Lbs
- The force due to gravity will be largely reduced by using constant force spring coils.

3) Desired speed
- Cutting - approx 12 Inches per second (300mm / sec)

4) Acceleration - I don't know - perhaps 1 G ?

5) Timing pulleys - approx 100mm dia / 4 in dia.

Questions:

1) What shaft should I use to mount the pulleys ?
- I am assuming at least 1/2 in / 12 mm for the short axis, and perhaps 1 in / 25mm for the longer one
- I have seen shaft in McMaster which has keyways cut all the way through it - good idea ?
- How do you keep the shaft from moving around laterally ?

2) What bearings should I use to mount the shaft ?
- Is a bronze bearing ok ?
- What about the plastic bearings from Igus ? http://www.igus.com/igcat/ig_cat1.asp?p=ig_23_4.gif
- I have mostly hand tools, so life needs to be simple here on bearing mounting.


3) How do you mount the bearings in a way to achieve reasonable alignment ?
- It seems like this will require adjustment capability to really work. How is this done professionally, and perhaps more importantly, by DIY'rs ?

Thanks

Harry