back driving ball screws

[jeffroot]

Is there a certain pitch that a ballscrew can no longer be back driven?

I am looking at using a 2or 4mm pitch ball screw and nut. In the application the screw will be turned by hand only 1/4-1/2 of a revolution, but I want the screw to be returned to its original position using a compression spring acting axially along the screw, backdriving the nut.

 

[MikeHalloran]

It depends on the helix angle, not the pitch, but ballscrews are generally easy to backdrive.



Mike Halloran
Pembroke Pines, FL, USA

 

[SincoTC]

As Mike said, it's the helix angle that matters. However, I was surprised to find that a 6mm dia 1mm pitch single start ballscrew in a THK linear actuator that we use will backdrive due to the very low internal friction of the recirculating balls. It does require a fair pressure but I just didn't expect it to happen with such a small helix angle!

If you are going to rely on backdriving (to a stop?) for positioning, you should be aware that some uncaged recirculating ball mechanisms that I have seen, exhibit a tendency at slow speed and small movements for the balls to srgue a little in the return channel because they are not being constrained and rolled by the preload, but just being jostled and pushed by the following balls, so little micro-jams can occur giving a stick-slip effect.

As you require such a small rotation, wouldn't it be easier to incorporate a torsion spring in your manual handle and keep the mechanical advantage on your side?

Trevor Clarke. (R & D) Scientific Instruments.Somerset. UK

SW2007x64 SP3.0 Pentium P4 3.6Ghz, 4Gb Ram ATI FireGL V7100 Driver: 8.323.0.0
SW2009x32 SP1.0 Pentium P4 3.6Ghz, 2Gb Ram NVIDIA Quadro FX 500 Driver: 6.14.11.7751

 

[tbuelna]

jeffroot,

Remember to keep your fingers out of the way when you let go of that spring-loaded handle!

Ball screws and roller screws are pretty efficient devices, so they backdrive quite easily. I would guess that a lubricated ball screw with a lead angle even as low as 0.50 degree would still be capable of backdriving. As long as the rest of the system frictions are low.

If you only need .040 to .080 inch linear travel in your device, you might consider a crank & slider mechanism instead?

Good luck.

 

[SincoTC]

With the high efficiency of ballscrews, and the mechanical advantage of the screw driving the nut, very little spring rate would be needed from a torsion spring in a handle that is only being turned half a rev maximum. Whereas, with the lower mechanical advantage of a low helix angle nut backdriving the screw, the axial compression spring mentioned by jeffroot may require a considerable force to overcome both the lower MA and the ball jam phenomena that I have seen at slow speed. Bear in mind that this force will also have to overcome by the operator through the handle and will probably be higher than a torsion spring in the handle.

My point is that ballscrews are almost universally used to provide a finely controlled linear movement from a low torque rotary input. In the much rarer cases where backdriving is required to give a rotary output from a linear movement, then multi-start, high helix screws are used to keep the required force reasonable.

I do agree with tbuelna in that with so little movement required, some other mechanism may be better, maybe a roller cam or even a simple pushrod pulled back onto an anvil made from a micrometer head or similar.

Trevor Clarke. (R & D) Scientific Instruments.Somerset. UK

SW2007x64 SP3.0 Pentium P4 3.6Ghz, 4Gb Ram ATI FireGL V7100 Driver: 8.323.0.0
SW2009x32 SP1.0 Pentium P4 3.6Ghz, 2Gb Ram NVIDIA Quadro FX 500 Driver: 6.14.11.7751

 

[jeffroot]

Thanks for all of the info!