View Full Version : Rolled or Ground ball screws

12-20-2004, 06:54 PM
How can you tell by looking at them if it is a Rolled or a Ground ball screw? :rolleyes:

Also why does some have two ball tubes and others one tube?

No name or part number on this so I can't identify.


12-20-2004, 08:35 PM
A ground ballscrew will be extremely smooth in the balltrack. I would say in your closeup picture, that this one is rolled, because I can see lines, unless it is just grease lines.

As for the number of ball return tubes, I don't know if there is a hard and fast rule that the manufacturers follow, but I don't believe that it matters how many there are. Of course, a multi-start screw will need a seperate tube for each track.

12-20-2004, 09:27 PM
I would have said it was ground, the rolled ball screw by Nook and a some others have a give-away ridge on the very top (non bearing area) of the thread, which has the appearance of a non-machined look.

12-20-2004, 09:40 PM
Yes there are grease lines where I wiped it off. And all surfaces have a fine grinding pattern finish. I have seen one bearing that I suspect was rolled and it had a turned up lip on each side of the ridge liked it was pushed up, during manufacturing, this one doesnt have it. So I'm thinking it's ground.

Are there advantages to a multi-start screw?

12-21-2004, 09:33 AM
I agree dose not look like the rolled screws I have seen.
Multistart Screws are often found on screws with a high pitch.
Picture two equal dia. screws one with 4 TPI the other 1 TPI due to
the geometry the 4 TPI would have much more contact surface then
the 1 TPI with equally long nuts reducing the load rating significantly.
There would also be much wider spaces between the threads on the 1 TPI.
To fix this 3 more Threads could be added inbetween the 1TPI to gain both
surface area and load rating.
You end up with a 1 TPI 4 start screw.

12-21-2004, 09:45 AM
I get it Torsten, looking closer I realize that it is a multistart screw and therfore has a higher pitch. That's why there are two ball tubes.

I will be using this on my Z axis so lifting a 1 3/4 hp Porter Cable router won't be a problem. But the height accuracy will suffer for some unknown amount.

12-21-2004, 10:11 AM

I thought about it more and wrapped a string around the screw and there is only one start, because the string didn't skip any turns.

Measured the travel of bearing for one turn, and it moved 0.200" or 5 threads per inch. That's half of my old acme 10, so it's not too bad.

So, the pitch is finer than a two start screw, which is good. Plus it uses two sets of balls for more weight handeling capabilities. And meaby less backlash???


12-21-2004, 10:44 AM
I don't think two or more return tubes will give less backlash, because the nut itself would (most likely) have been ground right through (nonstop), which would make the entire track symmetric. They would actually have to displace the ball race (axially) in the zone of the second tube, in order to attain some kind of preload factor.

12-21-2004, 10:52 AM
That is possible some have the ballraces ground slightly offset for
backlash free operation.
The Ballnut I have came in 2 configurations a regular shorter version with
one return tube and a longer version with 2 return tubes.
The longer version had a much higher load rating on the same screw.
I think the 2 return tubes are meant to prevent the balls from crowding up
on a longer tube.

12-21-2004, 11:35 AM
This ball screw has a 6 layer clutch on the end of the shaft that is adjustable for load. Why would someone want to put a drag clutch on a ball screw?

Is there any advantage in leaving it on and using it on a CNC machine. Havent heard any discussion on clutches.


12-21-2004, 01:10 PM
Is that the drive end? Maybe its a safety overload so you can "safely" run the table right into a hard stop, and really screw up the position information :D

12-21-2004, 05:38 PM
No, the drive pulley is on the other end.

When the two ends are bolted down the clutch/brake puts a certain amount of drag on the screw for some reason. Don't know why.


12-21-2004, 06:30 PM
I wonder if they were originally used with stepper motors, There used to be a practice of damping the response of steppers by various means, to reduce the 'settling' time or ringing when a step was taken, for e.g. SloSyn used a viscous damper to achieve this.
There were other means used, including electronic and mechanical, to provide a resistance to the output.

12-21-2004, 06:40 PM
That sounds reasonable. If there is space think I'll just leave it on and try it with and without damping, to see if there is any noticable difference.



12-21-2004, 07:42 PM
Hagar, If you have a constant apreciable load, you probally will not need it, it was used I believe on steppers that came to rest with no load attached. I have an old SloSyn manual that shows the effects in depth, with typical waveforms etc of before and after use of dampers. It is beneficial if you don't need it as a damper causes some loss of torque and speed.

12-21-2004, 10:11 PM
For sure the Z Axis will have a constant and appreciable load. So that piecs is history.

Thanks for all the input. :cheers: