[nateman_doo]

Well they are the ones made in China, and 3.5mm is noticeably larger then .125, so I can see why they didn't work. I would have to have a bunch of different sizes to experiment with. But what size would be suggested?

The ball nut itself does not move on the screw no matter how much I try to move it.

Also, how does one measure how much pressure is on a spring? I bought the belleville springs, but I have no idea (aside from full on torquing it down flat) where 200 lbs would be. It certainly is harder to turn the screw when I mash the wave springs flat.

There seems to be a "sweet spot" when I screw one screw closer to the other one. I wonder if there is such a thing as a sweet spot where the balls never shift around, and its still easy to move.

This is fun to experiment with while I don't have a table assembled, but as soon as I get everything I am going to want to have this all ironed out pronto.

[bru102]

The bellville's from Mcmaster have a "deflection at load" spec (others spec the force at 75% deflection.) So if it says 0.030", that's how far you compress it from "finger tight" to get the rated load. Not sure which one(s) you got, but whether it's 8 washers (in parallel) at 25lb each, or one at 200lb, you get 200lb in my example when you hit 0.030" compression. If you go 0.015", you get about 100lb, at 0.0075 it's 50lb, etc.

Remember that "torqueing it down flat" is the one place you can't predict the force accurately.

You should be able to come close on compression by measuring nut-to-nut with calipers "before and after", or estimating how many degrees you turned a cap screw to tighten down the ballnut. (a 1/4x20 capscrew moves 0.050" per turn, etc....)

The tightness you are feeling is preload torque. (See post #48 in this thread). You should expect to get about 20 oz-in of preload torque with a 200lb preload. You can easily measure what you are feeling...tape a light weight stick to your screw (radially, like a propeller)... then find something with a known weight like, say, 4 ounces. Hang the weight from the stick, and figure out how far out it needs to be to just begin to make the screw turn. A 4 oz weight, 5 inches out on the stick, gives you 20 oz-in of torque.

Ray B

[nateman_doo]

Ok, so using this part:

96445K284

Type
Belleville Disc Springs
Material
Steel
Steel Type
Grade 6150 Chrome-Vanadium Steel
Minimum Inside Diameter
16.3 mm
Maximum Outside Diameter
31.5 mm
Thickness
1.25 mm
Overall Height
2.15 mm
Load
1,913 N
Load Tolerance
+15%, -7.5%
Deflection at Load
.675 mm
Flat Load
2,359 N
Specifications Met
Deutsche Industrie Normen (DIN)
DIN Specifications (DIN)
DIN 2093
Notes
Springs are cold formed and fine blanked. IDs and ODs are machined.

Load is 1913N (430 lbs) @ .675mm (0.0266")
So roughly half of that is close to the 200 sought after lbs. 0.0133" is the space I should be looking for.

Since thats a pretty small amount to be looking for for (and they come 12 in a pack) I will add a few more, so I can have a more measurable deflection.

2 washers = 0.025 (0.133 x 2)
4 washers = 0.05 (0.133 x 4)
6 washers = 0.07 etc.

How does that sound?

[nateman_doo]

So I put 4 washers on like so:

NUT )()( NUT

works out perfectly (or at least close enough) where both flat sides of each ball nut are flat and on the same plane.

[wildwestpat]

In looking for dimensions of an MDL ball nut in my archive I found a paper by MDL which unfortunately I can't find the link but did save a PDF copy. The complete document is too large to post so here is sections five and six relating to rigidity / backlash. Please note that for all ball screws there is a minute pitch variation and it is only the variation under the nut that is encompassed by the ball circuits that is important from the view point of getting an acceptable torque with minimal backlash. Excessive torque is bad as it would shorten the life of the nut - brinnel the screw/balls/nut raceways if you over do it. Backing the nuts down hard against each other without measuring the pre-load is obviously wrong and using a spacer to get the pre-load correct using a bellvielle washer's compression as a means of measuring the applied load should work. Unfortunately there is little information on the pitch variation over a few turns expressed as a tolerance over the entire travel so some testing would be required. As I have tried to indicate the variations are very small even for rolled screws. The makers give the variation from an ideal pitch expressed as a deviation over a foot or 300mm but this only impacts on positional accuracy of the machine and only very indirectly on the backlash of the ballnut.

Hope this helps set the maximum pre-load safely at least for MDL nuts. Others may differ.

Regards - Pat

PS posting of the attachment failed and I have re-posted as post 517 on this thread.

[nateman_doo]

When you say shorten the life of the nut, I can understand why, but for a mill that is not in operation 40+ hours a week like a business would operate it, but a private guy in his basement- does that even apply? Lets say 10 hours a week TOPS?

I am not sure if mine are MDL nuts. I know they are made in china.

[nateman_doo]

So with the "200" lbs of load, I feel a "dead spot" in turning the screw by hand. Its very small, and more importantly the screw still moves. So there is no detectable backlash, but with this pre-load I feel that dead space where the screw turns very easily when changing directions, but back to some resistance when traveling.

I just used a 24V cordless drill on the screw, and it ate it alive. Tossed the screw around effortlessly. I am guessing that dead space is the small gap between the first and last ball bearing. You can only fit 17 in each race, and there is a small gap, so when its changing directions I guess that gap is the culprit? Still no backlash though.

The only reason I bring it up, is because when I was using lots of wave springs, a gentle pre-load didn't have that dead space.

[wildwestpat]

Here is the MDL Ball Screw information I failed to attach to my post earlier to day. I have had to scan the document from a paper copy as the electronic copy whilst displaying OK would not upload on to CNC Zone.

This is the simplest explanation of how to calculate the factors relating to ball screw selection and working out the life I have come across. If any one has a better then perhaps they could share it. There is also a section on how to pre-load these screws. Please note other manufacturers may have different limits and methods.

I have also looked back through my invoices for ball screws and the last four batches have been for C7 but the goods despatch notes are for C5 to I may have got lucky as the price charged is for C7. This may explain why I have never experienced difficulty with backlash elimination with dual nuts and no spring washers.


Nateman_doo the attached explain the life reduction which will be a problem if there is negative clearance in the internal load bearing surfaces within the nut see part 4 and part three covers pre loading.

Good luck - Reagards - Pat

[bru102]

Hi Nateman... Your arrangement of 4 washers is perfect... In "series" like that you get 4 times the displacement for the same load. It appears you've got that all figured out. If you stay in the 200lb range, you shouldn't see any wear problems from my research.

The dead spot is probably like the roughness I feel when I play with mine. I think it has something to do with the return tubes or paths. A lot of folks mention it, and its not clear it has any effect on operation... maybe it smoothes out with wear??

Are you gonna try measuring the preload torque? I learned alot about my machine by hanging weights to measure the torque.

Ray B.

[nateman_doo]

Explain? hanging weights?

[bru102]

Try this... from message #62 in the thread...

The tightness you are feeling is preload torque. (See post #48 in this thread). You should expect to get about 20 oz-in of preload torque with a 200lb preload. You can easily measure what you are feeling...tape a light weight stick to your screw (radially, like a propeller)... then find something with a known weight like, say, 4 ounces. Hang the weight from the stick, and figure out how far out it needs to be to just begin to make the screw turn. A 4 oz weight, 5 inches out on the stick, gives you 20 oz-in of torque.

Should work great with the nuts held in a vise like you showed. Just find a a small, light stick like 1/2 x 1/2 x 12", or a piece of aluminum like 1/8 x 1 x 12"...most any thing will work. Then fasten it to the end of the screw like a propeller. maybe hot melt glue? Maybe tape? even vice grips. Just get things roughly centered and horizontal so it isn't trying to turn the screw. Then hang a weight on it and find the spot where it has enough torque to turn the screw. The further out you hang the weight, the higher the leverage, and the easier it is to turn the screw. The preload torque is just the weight in ounces times the distance from center in inches.

I am fortunate to have an accurate postal scale... you'll need a small weight of known value. If you have a small piece of 6061 aluminum... it weighs 1.57 oz per cubic inch.

Once you've got the setup, you can easliy try it at zero preload, 200lb, 100lb, whatever...and get a good feel for what your drive will need to provide to overcome preload torque.

Ray B.

[nateman_doo]

ugh... Its all on HOLD. The guy I bought my bearings from screwed up.

I drew a child-like drawing I thought anyone could understand, that I wanted 24" to the outside of the blocks:


Apparently, it wasn't simple enough. This is what I got. Clearly too long.


Does my drawing look ambiguous is ANY way at all?? (chair)

[rwskinner]

At least it's an easy fix and they aren't too short!

[nateman_doo]

Maybe easy fix for you, I wouldn't dare throw those on my 7x10 lathe. He boogerd up an easy drawing to follow. He needs to fix the mistake.

[rwskinner]

Hey, I know it's disappointing, but too long is always better than too short.
Are these the ones from LMB?

[wildwestpat]

If you have an ER32 collet for the 7Xxx then it is possible to hold the screw and turn down the floating end. Use reduced speed a carbide tool with plenty of lubricant. Fix the nuts close to the back-end of the head stock with plenty of gaffer tape (Duct tape) to stop them flying off the screw. Pack out the screw where it exits the back end of the headstock with tape to help centre the screw in the spindle and reduce the tendency for it to wave about. The floating end is less critical than the fixed end when it comes to machining.

Your screw may be short enough to do with out the floating end support as the floating end support's main impact is in stopping the free end wagging about at the higher speeds. With a screw 16mm diameter and a length of 600mm you could expect to run the screw at 1,500 rpm and up 5,000 rpm if the floating end is supported.

Hope this helps if the seller will not help - Regards - Pat

[nateman_doo]

Hey, I know it's disappointing, but too long is always better than too short.
Are these the ones from LMB?

mmhumm. I paid quite a bit of money. I am baffled why he could not understand my child-like drawing. Please tell me if that drawing was too complicated and it is me.

[around]

If that was the only drawing then yes, your drawing left room for ambiguity. The measured dimension on the drawing shows a 24" dimension and then some additional bearing surfaces beyond. It looks like the screw as delivered has a 24" threaded section and bearing surfaces beyond.

What is not clear is what is happening in the bearing blocks. I assume you wanted the threaded section to stop at some point inside the bearing blocks (but this dimension is not given). As you did not draw what the screw alone was to look like, this became a guess for the machinist. I would, however, have expected a machine shop to ask (but maybe you had a verbal discussion about the job and he/she felt they understood the requirement).

Here is an example drawing for a ballscrew I had machined. The machinist reviewed the drawings and discussed any points that he was unclear about before even quoting on the job. Good practice on his part to make sure the customer gets what they think they are getting.

Cheers,

Adrian

[wildwestpat]

IMHO the way of specifying the screw would be by the active length i.e. inside the bearing blocks. He looked quickly at your drawing and took your dimension as being the inside measurement not as you wanted the outside dimension.

I have always specified the length of the ball screw track showing the end machining as an addition to the length quoting the standard dimensions for the way the end is to be machined pus the drive diameter and length of the drive section of the shaft if I want something different from the standard.

Some times I want the fixing centres to be the important dimension in which case I have annotated the critical parts in red adding note to the effect that the ends are standard but the finished centre to centre of the bearing blocks is critical along with the tolerance.

Afraid this does not help as you have the screw which is not to the length you wanted. Can I suggest you look at the drive RPM you need and see if the floating end bearing is really necessary - yes it looks nice to have blocks on both ends of the screw. Cut off the excess with an angel grinder wrapping the screw in a wet rag to prevent over heating. Then turning down is within the capability of your lathe if the screw will go through the spindle.

Regards - Pat

[nateman_doo]

I don't think there is any ambiguity. I clearly drew a line from the edge of each bearing block which I wanted to be 24". I even made it darker then the rest. I know its not any type of official drawing, but I think a child can follow where the lines go, and they go to the far edges of the bearing blocks.