[rabidhamster]

Greetings all. I'm building a ~ 4"x3" 8020 machine, and using ~32mm ballscrews and THK precision linear rails. I have 2 precision ballscrews outta a decommissioned mill that are matched in every way but length. My impulse is to just dual-drive the x axis to avoid any chance of racking, but my gantry will be, it seems, very rigid, so perhaps centerdriving would be the way to go. What would you do if it were your machine?

Heres a sort of mockup photo to help with your visualization. Linear bearings not pictured, the steel rail wont be used there

[crane550]

After seeing it both ways I have one honest recommendation.... DUAL! I cannot stress this more. At very minimum, run dual screws and drive them both with a belt, like the Solsyva's do. I switched mine to dual motors and the X and could not believe the difference. Dual screws over a single offer so much more rigidity and accuracy. Looking into the laws of force and torque the math certainly back this claim up.

That being said, it is possible to run 1 screw. People have done, and some with good success. However, (and this is big) your potential for racking and problems goes thru the roof. Any looseness in your system will show up there on the ends of your gantry.

All my designs use dual drive. Your free to do your own, but this is my recommendation since you asked.

[louieatienza]

It depends on how you're ultimately going to have your system set up. but normally, at even your shortest width 3', you'd probably have to use dual screws unless you space your LM blocks far apart, and then you'd lose travel.

I'vev asked before if there was an optimal ratio between bearing rail spread and distance between blocks (end to end) that would give good ridgidity without excessively reducing travel. Your system as it sits would have a mechanical disadvantage if the leadscrew is mounted under the table, since it would be at a maximim distance from the bearing blocks IIRC.

Barring commercial machines, it seems the most successful single leadscrew designs for DIY have the rails 24" or less, and larger widths with two leadscrews or even r&p.

[rabidhamster]

Thanks guys.

Dual drive seems like it'll still be the way to go then. And really, given how much I hate redoing things, it was foolish to consider not dual driving to begin with.

[mmomike]

I do not have a cnc router yet.

What happens if one of the motors, in a dual X configuration, stops working in the middle of a run?

Since only one side is working, will it tweak major parts of the machine and require serious repair?

[JerryBurks]

I do not have a cnc router yet.

What happens if one of the motors, in a dual X configuration, stops working in the middle of a run?

Since only one side is working, will it tweak major parts of the machine and require serious repair?

Probably depends, if the machine or the motors are stronger.

I had the same concern when I built my machine and I force synchronized both steppers with a belt. This is kind of unusual from what I have seen here but it works very well. In normal operation, the belt is just idling along without load and does not contribute to inaccuracies, but if a motor drops out or worse if it stalls hard, the belt prevents the other motor from twisting the gantry. Another convenience is I can move the x by hand, tugging on the belt or turning the pulleys without the gantry losing perpendicular adjustment.

[TSJ]

There have been many posts with multiple arrangements about using cables to keep the gantry from racking. When I get to building the gantry I will put a cable system in the same as the "Paralex" fence on my Powermatic 66 table saw. I have had it for almost 30 years and only had to adjust it once, I removed the saw top to replace the bearings on the arbor.

[msimpson99]

Dual all the way. The only reason to go single is if you are moving the table on a small CNC router or Mill.

[kevincnc]

I was worried about potential damage from 2 motors if one stopped working, and am doing mine like this-

[msimpson99]

The other should just stall if one quits. On my KRMx02 machine if one stops the other can drive both axis.

[louieatienza]

I attempted to stall one of the steppers intentionally as a test of my new dual stepper setup. No permanent damage done, though the gantry had to be resquared, which I actually do now by running it into hard stops (slowly of course) and letting them stall, until I get the homing switches wired up.

I believe Kevin's design calls for servos, which might cause damage?

[kevincnc]

I believe Kevin's design calls for servos, which might cause damage?

Mine will use a single servo motor to drive both ballscrews. What do you see as the potential for damage?

[ger21]

Mine will use a single servo motor to drive both ballscrews. What do you see as the potential for damage?

He means if you were using 2 servos.

[bebob1]

Most lead or ball screws have very low backdriving efficiency, which is also dependent on the lead angle or pitch of the screw itself. Some low lead angle screws are even negative backdrive efficiency so a lot of force is needed to overcome it (essentially the nut is fixed for all practical purposes). Not to mention if one drive fails during operation you immediately have huge moments generated to cause lots of friction on the slide. That is also assuming once the drive fails, the servo or stepper acts like an open circuit instead of a close circuit which will otherwise make the screw even harder to turn.

[Devastator]

I use dual leadscrews driven by one motor. No problems so far. If I did use two motors, I'd tie them together with a belt like JerryBurks did.

[yellow68gto]

I have a similar question regarding single vs dual screws. I am also using decomissioned cnc mill parts to build a very heavy duty fixed gantry router. My parts list include

2 identical 40mm x 6mm pitch ball screws from a haas vf5 with 30inches travel.

45mm thk profile rails the are 65" long

And a table from a vf3 that is 48x18x2.5 inches thick.

I was hoping to sell one ballscrew( only 1100 hours on it) and put that money towards my spindle.

I want to run the table lengthwise to help keep the machine more square. I am going for x50" y30" and z16" travels.

If I run the bearings under my table at 24 inches apart and a single ball screw down the middle, do you think the table is going to rack? I was hoping with the beef of the rails and bearings it would keep things stable. Also I could run 3 bearing per rail if that will help.

I am not looking at doing much steel but lots of aluminum.I am looking at the fire cnc bt30 spindle and 6krpm capabilities.

My parts I currently produce are approx 20x12x2.0 aluminum plate. Speed isn't as much as a concern as accuracy. I haven't even started looking at electronics yet. All I know is They will be big.

Any thoughts?

[JerryBurks]

.........Any thoughts?..........

If you have a massive moving table with such heavy duty bearings spaced apart 24" in x-direction I would think racking should not be an issue (but that is just my somewhat uneducated guess).
Racking is mostly a problem with moving gantries that can twist or bend and the closely spaced bearings (usually not more than 6-10") will move along x before the increasing gantry deformation under load puts up enough resistance. In that case the dual ballscrews will cure the racking.

[wizard]

First thing first, I have no problem at all with a single motor driving dual leadscrews via a timing belt or other drive arrangement. It can work very reliably if properly engineered.

I have a similar question regarding single vs dual screws. I am also using decomissioned cnc mill parts to build a very heavy duty fixed gantry router. My parts list include

2 identical 40mm x 6mm pitch ball screws from a haas vf5 with 30inches travel.

45mm thk profile rails the are 65" long

And a table from a vf3 that is 48x18x2.5 inches thick.

This is sounding very nice even before we see pictures.

You should have no trouble pushing that table around with a single lead screw assuming suitable use of linear bearings. The question then becomes can you mount the bearings in a way that is suitable for the table to keep it free of racking and still have suitable travel. More info on where you are going with your machine design is needed.

I was hoping to sell one ballscrew( only 1100 hours on it) and put that money towards my spindle.

I want to run the table lengthwise to help keep the machine more square. I am going for x50" y30" and z16" travels.

That would be a good approach. However ultimately it depends upon the final arraignment of components and you may find that your rails are a bit short. Your linear bearings can't be more than 15" apart to get that travel, that in a machine that mounts the rails on the bed.

If I run the bearings under my table at 24 inches apart and a single ball screw down the middle, do you think the table is going to rack? I was hoping with the beef of the rails and bearings it would keep things stable. Also I could run 3 bearing per rail if that will help.

That would work but your rails are only 65" long. You wouldn't get your desired X travel. Racking is pretty easy to understand, the closer your two bearings for each rail are the harder it is for them to resist twisting. Your table is only 18" wide which means the lead screw at worst is only 9" from a rail. In your case I believe the real challenge is properly supporting the table in the long direction, bearings 24" apart imply lots of overhang or no support, which could be an issue depending upon the tables stiffness. My gut feeling considering this is suppose to be a router is that you will do OK.

I am not looking at doing much steel but lots of aluminum.I am looking at the fire cnc bt30 spindle and 6krpm capabilities.

My parts I currently produce are approx 20x12x2.0 aluminum plate. Speed isn't as much as a concern as accuracy. I haven't even started looking at electronics yet. All I know is They will be big.

Any thoughts?

Sounds like a big involved build, should be interesting to see what you come up with.

[hanermo]

I would not use slaved axis on anything important, expensive, rigid, or "industrial".

Where you use steppers and low power stuff, it wont matter, but any proper resolution/gearing/torque will break expensive stuff immediately.

I like the idea of slaved+timing belt .. and think its ok for less-rigid applications.

For my vertical z axis bridge, on a bridge mill/VMC, I will be using 2 vertical ballscrews, spaced 2 m apart.
I dont trust any of the slaving arrangements in corner cases, based on lots of experience.

Drive is servo, and force == 1500 kgf, = 15000 N.
More than approx 0.2 mm error and the bridge will rack, breaking some of the linear guides, maybe.
Even though the linears are very strong (50 metric tons rated) (6 rails, 12 trucks, 35 mm profile, 4000 kg rated load each).

Plan is to use large belts, and one servo.
Cost is not an issue - reliability is very much an issue.
Large belts, HTD8/30, cost approx 400€ for the whole arrangement, and this is ok, for me, for this use.
Approx 48 teeth pulleys (== 200 mm in diameter, several kg per pulley) at each end.
About 100 kg belt tension.

Bridge, going up/down, is == 800 kg in mass, fwiw.

For comparison, even tiny steppers (nema 23) do 500 kgf force, and 500 kgf up vs 500 kgf down = 1000 kgf differenential load.

This of it this way:
Can you put a full size family car on your bridge or gantry, at one end, without it bending/breaking, something expensive, permnanetly ?

(I can put a 3000 kg truck on end end. I´m unsure about putting on 2 trucks with racking stress at end end only..)

[yellow68gto]

Wizard thank you for the reply but I don't think I explained very well. My gantry X is going to have my 50 inch travel. Table Y is only going to have 30 inch travel.

Rotate the table 90 degrees in your thinking.my machine will be fairly square.

This is why I am asking about racking. I will be able to mill 25 inches from my Y ball screw center and 13 inches past each rail.

I am going to look into mounting the rails farther apart.I was just hoping to utilize the pick mounting already in my vf3 table.

Hope this helps

Ken