[dudz]

I have used my home built machine for years now with good results. I have 1605 Chinese rolled ballscrews (c7) on my X Y and Z axis. ( sliding bed )
( Chinese 2.2kw water cooled spindle)
Now have play in all the ballnuts, and the 20mm supported round rails on the Z and Y axis have a little play in the bearings and so does the 25mm X axis rail bearings.
So my plan is to now go square linear rail on all axis and C5 ground Ballscrews / nuts on all axis.
I am looking at used ballscrews from ebay that are ground C5

I would like to know would it be overkill to go up to a 20mm ballscrew on the Z axis (approx. 140mm of usable travel) and maybe 20mm on the other axis ? X being 400mm of travel and Y being 300mm of travel

Not sure on the size of linear square rail to go ?

I use this machine for Aluminium and want good finish and reparability without replacing everything. Hopefully I can still use the same motors



3 axis CNC/router / Alu profile frame....25mm Alu Cutting bed X=400mm Y=300mm Z=140mm.....Supported 25mm X rails (BED) ....Supported 20mm Y rails....Supported 20mm Z rails.....2.2kw Chinese WC spindle..... CSLAB CSMIO/IP-M 4-axis Ethernet Motion Controller....M542 Drivers..SY60STH86-3008BF Motors...running....Mach3 / Cambam / Emachineshop.

[awerby]

The only downside to fatter ball screws is the greater degree of rotary momentum they develop. I don't see this as a big issue with the Z axis, since it never has a chance to get going very fast, but on the longer X and Y screws, you might find yourself losing steps with a sudden change of direction.

[joeavaerage]

Hi,
aside from cost, the big penalty about oversize ballscrews is rotary momentum.

I bought some second hand 32mm diameter, 700mm long BNFN double nut C5 THK ballscrews for my project, without really considering rotational momentum.
Despite the linear axis weighing 150kg the rotational momentum of the ballscrew is 80% of the total momentum with the servo armature at 12.5% and the linear axis at only 7.5%.

Yes....you read that correctly a 150kg iron and steel axis moving at 25m/min is still a small fraction of the momentum of a 32mm ballscrew rotating at 5000rpm. The point is that the axis
despite its weight is moving slowly whereas the ballscrew is rotating flat-stick. It surprised me when I did the calculation. I have subsequently done the same calculation for many
on the forum and the result is consistent, in most machines the ballscrew dominates the momentum equation.

Before you commit to any given purchase post the details of the ballscrew, servo, axis etc and we can do the calculation. You could include friction and ballscrew efficiency and so on
for an even more accurate assessment but the thumbnail calculation I do will determine if you are in the ballpark.

I used 750W Delta servos so the extra rotational momentum was not an issue....but you need to be aware of just how much momentum your ballscrews has as that will determine
the required torque of your servos to achieve your desired acceleration.

The rotational momentum of a ballscrew increases as the fourth power of the diameter. Thus a 32mm screw has 2.7 times the momentum of a 25mm screw and 6.5 times the momentum
of a 20mm screw. Larger screws are more rigid and tolerate high speeds before whipping, but you may need to increase the torque of your stepper/servos to accommodate the extra momentum.

I would not describe 20mm as too big. In fact I would not describe 32mm as too big either, although 25mm diameter would have been a better match for my machine.

Craig

[dudz]

Well most of that is helpful thankyou but over my head. The details of my motors are in my signature here. The Z axis prob weighs around 20kgs with the spindle.

The link to the eBay z ball screw is here :https://www.ebay.co.uk/itm/202979986...&segname=11051

[joeavaerage]

Hi,
that ballscrew is 5mm pitch and therefore very little torque is required to offset the weight of the Z axis/spindle.

For a 5mm pitch screw the mechanical advantage is 125kg (weight or 1250N) for every 1Nm of applied torque. So if your spindle/Z axis weighs 20kg you'll only need
20/125=0.16Nm from your motor to oppose gravity....bugger all. You want much more torque to accelerate the axis rather than hold it up.

I bought all my linear rails/cars and ballscrews from that same company in Korea. Good supplier, recommended.


Craig

[dudz]

Hi,
that ballscrew is 5mm pitch and therefore very little torque is required to offset the weight of the Z axis/spindle.

For a 5mm pitch screw the mechanical advantage is 125kg (weight or 1250N) for every 1Nm of applied torque. So if your spindle/Z axis weighs 20kg you'll only need
20/125=0.16Nm from your motor to oppose gravity....bugger all. You want much more torque to accelerate the axis rather than hold it up.

I bought all my linear rails/cars and ballscrews from that same company in Korea. Good supplier, recommended.


Craig

Thanks so much !
What size square linear rails would you recommend using 20/05 ballscrews ? . I'm guessing I go big just for extra stability ?
I have been advised in the past to go double ballnut where I can, to limit backlash etc . But with C5 ground screws and nuts I would guess it would be fine eh ?

[joeavaerage]

Hi,
in my machine (pictured above) I used new (old stock) THK 20mm rails and cars, because that is what I could get from DY-Global at the time. I would have preferred 25mm, but I got what I could.
There is no disadvantage in getting bigger aside from cost. I think 20mm would be fine, but even 15mm at a pinch. 25mm or 30mm would also be fine especially if you can find good ones cheap.

Double nut ballscrews allows the manufacturer to easily and precisely set the preload. The downside is the extra nut adding to the cost. Most genuine ground C5's are fine even without double nut. There again if you can find
double nut screws at a good price go ahead. Remember that the overall length of the nut is somewhat longer with double nut which in turn will reduce your travel a little for a given length screw.

Just beware there are some Chinese suppliers selling what they claim is C5 and C3 but at C7 rolled screw prices.....I call BS. Good, genuine ground C5's and C3's from the leading manufacturers are ten times the
price of rolled C7's. If you are of the opinion someone can produce something as good as they claim but at a tenth of the price of THK, NSK, Steinmeyer etc then go for your life, they won't get a cent of my money!!!

Craig

[dudz]

Ah i was referring to this Korean ebay supplier with the C5 grades.

What confuses me a little, is the total height of the rail and bearing blocks never seem to match the height of the ballscrew nut brackets. In other words, you either have to shim the ballnut bracket or the rail bearings on the axis. Or am I missing something ?

[joeavaerage]

Hi,

What confuses me a little, is the total height of the rail and bearing blocks never seem to match the height of the ballscrew nut brackets. In other words, you either have to shim the ballnut bracket or the rail bearings on the axis.

Sad fact of life. You're not the only person to have come to the conclusion that you need blocks (poor choice for rigidity) OR a gutter milled into the axis bed.

I my case I had the axis beds cast and so the 70mm height difference between the rails/cars and the top of the ballnut is no drama....but with other techniques it is a substantial hurdle.

Craig

[dudz]

Ah yes. I thought about that other option but thought there might be another explanation. I will just mill out the back of the Z plate for the higher ball screw nut bracket. (using a thicker Z plate)

Thanks