PDA

View Full Version : Ballscrew Thickness!



AdCNC
15-08-2009, 10:54 PM
Hi guys im in the middle of building my machine but i would like to know your thoughts on what thickness ballscrew i should be looking at to span over 1829mm ( 6ft ) without none to much "whip" and another thing is the pitch as on zapps website they do a 5 or 20mm pitch in the 20mmOD screw, the machine im building is for precision work mainly but with it being a 6x4 bed i was thinking that a 5mm pitch would take all day get to the other end? :confused:

tribbles
16-08-2009, 12:05 AM
I used 25mm OD ballscrew with a 25mm pitch for mine, and with the motors I've got, I get 18m/min, which for about a 1.4m run gives ~4.7 seconds for total travel. However, you will need to worry about whip.

Gary (from Zapp) gave me a website which allows you to calculate the whip speed based on the diameter and so forth - unfortunately, I've only got the link at work, but Gary may come on here before I get the chance on Monday to get it.

AdCNC
16-08-2009, 12:40 AM
thanks tribbles for that info, 25mm OD is some fair chunkey stuff!

Gary
16-08-2009, 01:15 AM
Use this link.

http://www.nookindustries.com/ball/BallCalculators.cfm

use the root diameter that is normally about 2-3 mm lower than the outside diameter and select the mounting type you are using, and it will calculate the max speed before it will whip.
If you have any question, or need help give me a ring on monday.

Gary


I used 25mm OD ballscrew with a 25mm pitch for mine, and with the motors I've got, I get 18m/min, which for about a 1.4m run gives ~4.7 seconds for total travel. However, you will need to worry about whip.

Gary (from Zapp) gave me a website which allows you to calculate the whip speed based on the diameter and so forth - unfortunately, I've only got the link at work, but Gary may come on here before I get the chance on Monday to get it.

AdCNC
16-08-2009, 01:41 AM
Thanks Gary much appreciated ;)

Ross77
16-08-2009, 01:50 AM
Gary (from Zapp) gave me a website which allows you to calculate the whip speed based on the diameter and so forth


As Tribbles said the diameter required to reduce whip is fairly easy to work out. The pitch is not so. You mentioned speed and accuraccy..... I dont think the two are easily achived without lots of money.......

You havent said what you are cutting or what accuraccy you require. But if you are using steppers when 200 steps per rev at 5mm pitch gives 0.025mm resolution. With a 25mm pitch you get 0.125mm resolution.

Also given that for your 1829mm travel you will probably need 20mm + dia. then motor torque will be an issue, lots of mass and inetia to control.
I may be corrected by others but you are probably in Servo Motor terratory for the distances and speeds you require.

See the attached pdf that I got from Cnc-zone.( hope its ok to repost it. cant remember who its from. {Lee any probs just remove it}

Goes in to detail about designing motors, screws and feeds.

Ross77
16-08-2009, 02:01 AM
Sorry. Gary and your self posted whilst I was writting...
30mins to post. Must be time to go to bed....:heehee:

tribbles
16-08-2009, 11:16 AM
thanks tribbles for that info, 25mm OD is some fair chunkey stuff!

I was looking at 40mm before realising the motors wouldn't go as fast as the speed I was anticipating, so Gary suggested using the 25mm pitch stuff instead - and it's worked out great (thanks!)

AdCNC
17-08-2009, 12:02 AM
You havent said what you are cutting or what accuraccy you require.

Well i was planning on cutting woods and none ferrous metals!

Ive been looking at the 25mmOD ballscrew and im leaning towards that! as for the motors being not power full enough i can gear them to something like a 5:1 or more and get a 20-25 pitch ball screw!

What do you guys think!

Ross77
17-08-2009, 12:36 AM
Well i was planning on cutting woods and none ferrous metals!

Ive been looking at the 25mmOD ballscrew and im leaning towards that! as for the motors being not power full enough i can gear them to something like a 5:1 or more and get a 20-25 pitch ball screw!

What do you guys think!

Did you mean none or some? Gearing will definatley help with the motor power and accuracy (200 steps at 5:1 is 0.025 resolution) but will also slow the travel time down. Max stepper RPM is 350 ish I think

If your using belts then you might also want to check the 5:1 ratio, sounds a bit high. the small pulley will not have enough teeth in mesh if the bigger pulley is to large. Ive got a calculator at work so if its any help I'll post it tommorow.

Hope you dont think I'm being negative but I had to make the same decisions for my build. I chose Slow and accurate because Im only using it to build the fast machine :heehee:

Have you concidered rack and pinon? its cheaper and very fast.

Ross77
17-08-2009, 01:01 AM
Ok I've had a rethink, Have you already got your steppers and drivers yet?

If not then the better drivers have micro stepping. if you get the correct size motor then u can direct drive it and use the micro stepping to get the accuracy/ resolution you need.

Give gary a ring. Im sure he will help you and is much better qualified then me. Im just trying to cover all possibilities to help you make the right decision. :smile:

tribbles
17-08-2009, 01:22 AM
I'd agree with Ross - the drivers I got (from Gary) will do from 400 steps/revolution to 10,000. With a 25mm pitch, it equates to 0.0025mm accuracy.

Assuming I've divided it right :smile: (it's late...)

John S
17-08-2009, 01:32 AM
Remember though, every time you increase microsteps you decrease torque.

.

AdCNC
17-08-2009, 01:58 AM
hmmm ive already got the drives and motors from kelling in the states, the motors and drives are the 425 oz-in motor and KL-4030 Bipolar Driver the drive features are ..

* Voltage up to 24 to 40VDC, current output up to 3.0 A
* Selectable resolutions, 1,1/2, 1/4, 1/8, 1/16, 1/32, 1/64
* Suitable for 4,6,8 lead motors
* DIP switch current setting:
0.9A,1.2A,1.5A,1.8A,2.1A,2.4A,2.7A,3.0A
* Over voltage, heat and short-circuit protection

If i did end up gearing the ballscrew with some kind of belt and belt grip was a issue i could devise some sort of roller bearings in between the two pulley wheels so it would wrap the belt more so around the small drive pulley! just a thought, i suppose that all depends on what hight the teeth were on the pulleys and belt.

John S
17-08-2009, 10:07 AM
I think you need to come to accept realistic terms with what you have.
Running large 25mm screws at high speed with these motors and drives isn't going to happen.

You have a 6 volt motor that to accelerate up to speed is expecting to see 6 x 20 volts or around 100 to 120 volts and you only have a 40 volt max supply that is really only going to be happy on about 32 to 35 volts to stay inside a decent safety factor.

Gearing up, then micro stepping will only bleed off available torque making the situation worse.

.

Gary
17-08-2009, 10:18 AM
Those motors and drivers will be way too small.
If you are going to spin a 2500mm x 25mm screw you need a large 34, and a good supply of volts.
Under the speed you need those motors will most likly be only giving about 1Nm, also the inertia mismatch is way to large, even if you gear it.




hmmm ive already got the drives and motors from kelling in the states, the motors and drives are the 425 oz-in motor and KL-4030 Bipolar Driver the drive features are ..

* Voltage up to 24 to 40VDC, current output up to 3.0 A
* Selectable resolutions, 1,1/2, 1/4, 1/8, 1/16, 1/32, 1/64
* Suitable for 4,6,8 lead motors
* DIP switch current setting:
0.9A,1.2A,1.5A,1.8A,2.1A,2.4A,2.7A,3.0A
* Over voltage, heat and short-circuit protection

If i did end up gearing the ballscrew with some kind of belt and belt grip was a issue i could devise some sort of roller bearings in between the two pulley wheels so it would wrap the belt more so around the small drive pulley! just a thought, i suppose that all depends on what hight the teeth were on the pulleys and belt.

AdCNC
17-08-2009, 01:05 PM
hmmm, ok guys what motors would you recommend to spin a 1900mm of 25mm ballscrew? would you go stepper or servo? and what drives would you suggest using?

tribbles
17-08-2009, 01:49 PM
It depends on how fast you want it to spin (and therefore how deep your pockets are).

I've used the PM2278 stepper controller, and the SY85STH156-4208 motor, which is about the most powerful combination (my philosophy is that it's easier to slow something down than to speed it up).

My steppers are running at 720RPM, which is about the fastest they can do. For more speed than this, then I'd need to go to servo motors instead.

However, my pockets aren't that deep - the cost of the three steppers and drivers was about the same ballpark figure as for a single servo and servo driver (so you're looking at three times the cost).

If I need the extra speed, then I'll upgrade, but for the moment it's fast enough.

Gary
17-08-2009, 01:57 PM
What kind of speeds do you want?
If you want speeds like Tribbles has, you are on the border line of servo /stepper.
The setup he has is the SY86STH156 and the PM2278 run with 110V AC. (If i remember right)
With this setup you are looking at about £250 per axis, and you will most likely need two ball screws on X or it can crab, even with profiled rail.
A servo system will be about twice this, but will be far more accurate and much more responsive.
Our servo systems have 2500 line encoders that are quadrature so you get 10000 pulses per rev, and these are actual encoder positions rather than extrapolated positions that you get with an open loop stepper system.
With a servo system then the actual backlash of the mechanical system becomes the weak spot rather than the accuracy of the motor.

Donít forget with a servo system, you will need to gear it, and also consider the max frequency of Mach/parallel port, as this will be a factor in your max speed.



hmmm, ok guys what motors would you recommend to spin a 1900mm of 25mm ballscrew? would you go stepper or servo? and what drives would you suggest using?

AdCNC
17-08-2009, 05:20 PM
ok, so if i get my self 3x PM2278 Drivers 2x 1810 oz-in NEMA 34 steppers for the X axis and 1x for the Y and use the NEMA23 for say the Z for the time being i should be able so work at a half decent pace and it will get the 6x4 table in motion? ive got no chance of going for servos at this stage!

tribbles
17-08-2009, 05:30 PM
What height did you want to achieve from your Z-axis?

I personally think that a PM2278 driver would be a bit of an overkill for the Z-axis if you're using a NEMA23 for the stepper.

AdCNC
17-08-2009, 05:51 PM
Roughly 7-8" the drive for the Z axis would remain the same as i have now! so looks like ill be selling 3 NEMA 23 425 oz-in steppers with 3 KL-4030 Drives rated at 3A if anyones intrested! Brand new!

Another thought was i could even go for a single NEMA 42 for the X axis and drive both of the ballscrews with a belt coupled together! what are your thoughts on this guys? the NEMA42 is rated at 2830 oz-in and can work at a max of 210v do you think this would work. ( im just trying save money as funds are abit low at the moment )

Gary
17-08-2009, 06:04 PM
Yes the Z axis does not need to be as big, and the motor you currently have would most likely be ok, as you have it allready use it to see if you can get what you need, if not, then upgrade it.
does the Z need to be as quick? if not you could most likely get away with a 1605 ball screw.


ok, so if i get my self 3x PM2278 Drivers 2x 1810 oz-in NEMA 34 steppers for the X axis and 1x for the Y and use the NEMA23 for say the Z for the time being i should be able so work at a half decent pace and it will get the 6x4 table in motion? ive got no chance of going for servos at this stage!

AdCNC
17-08-2009, 06:09 PM
hi gary, no the z can plod along at a nice steady speed really! its the x and y thats the main concern!

AdCNC
17-08-2009, 06:09 PM
taking into consideration that the belt coupling could be 1:1 effectively each ballsecrew would see 1415 oz-in of torque thats not taking into consideration drive losses! thats if i used a monster 2830 oz-in stepper!

Gary
17-08-2009, 06:28 PM
So are you going to use one motor for the X and two ball screws?
This can work well, but you need to take note that if the belt is quite long it can get a bit springy and you will see a bit of hysteresis.
I would personally use two motors and drivers.
A belt system can mean a loss of 10%.


taking into consideration that the belt coupling could be 1:1 effectively each ballsecrew would see 1415 oz-in of torque thats not taking into consideration drive losses! thats if i used a monster 2830 oz-in stepper!

Gary
17-08-2009, 06:32 PM
Those motors have very high inertia and high inductance, so they dont spin as fast as the 34's, also the PM2278 goes up to 220V, so you will need to get a transforma regardless because it is not advised to put 230V single phase straight in.



Roughly 7-8" the drive for the Z axis would remain the same as i have now! so looks like ill be selling 3 NEMA 23 425 oz-in steppers with 3 KL-4030 Drives rated at 3A if anyones intrested! Brand new!

Another thought was i could even go for a single NEMA 42 for the X axis and drive both of the ballscrews with a belt coupled together! what are your thoughts on this guys? the NEMA42 is rated at 2830 oz-in and can work at a max of 210v do you think this would work. ( im just trying save money as funds are abit low at the moment )

Ross77
17-08-2009, 07:22 PM
Another Idea. :whistling: Have you thought about fixing the ballscrew and spinning the ballnut instead?

You would reduce the amount of torque needed as the mass is reduced. I have one like this that I got from ebay.

Finally why do you need so much speed? Unless it has to return home after along cut, wont you be limited by the cutting speed anyway?

Gary
17-08-2009, 07:35 PM
I have seen this done before, but this normally involves more money and because the motor/gearing that is associated to this type is moving with the ball nut, the moving mass is higher.


Another Idea. :whistling: Have you thought about fixing the ballscrew and spinning the ballnut instead?

You would reduce the amount of torque needed as the mass is reduced. I have one like this that I got from ebay.

Finally why do you need so much speed? Unless it has to return home after along cut, wont you be limited by the cutting speed anyway?

AdCNC
17-08-2009, 07:40 PM
options options options, think im going be scratching my head till ive build my machine and ready for the motion! and thats not to far away!

Ross77
17-08-2009, 07:56 PM
I have seen this done before, but this normally involves more money and because the motor/gearing that is associated to this type is moving with the ball nut, the moving mass is higher.

Didnt think of that :eek:. Ist the main problem in this case the actual torque needed to spin the screw? Although the extra weight wouldnt be ideal surley it is a lesser evil ? wouldnt it be possilble to also use a smaller screw which would make it cheaper?

Sorry to hijack your tread :naughty:



options options options, think im going be scratching my head till ive build my machine and ready for the motion! and thats not to far away!


Isnt that all part of the fun.....:heehee:

AdCNC
17-08-2009, 08:04 PM
haha yeh but its not fun anymore my wallets burnt out!

Ross77
17-08-2009, 10:19 PM
Tell me about it, this definetly isnt a cheap hobby....:nope:

I would love to throw 5K at a scratch built machine, with everything designed for its purpose....as opposed to buying S/H parts and making the machine fit them :whistling:

Thank god for Ebay tho.

irving2008
18-08-2009, 12:29 AM
a 16mm dia screw with support both ends would work fine over that distance (~1.8m) up to 700rpm approx, you don't need a 25mm screw...

tribbles
18-08-2009, 12:37 AM
Well, according to the calculator:

- 14mm root diameter (subtract 2mm due to thread size)
- 1800mm length between bearings
- B-style end fixing (one fixed, one floating)

Maximum speed is 522RPM.

Even with a 16mm root diameter, it's 597RPM.

(But I know engineers like to put some extra for safety, so it probably would work up to 700RPM).

Or do you have another way of calculating it?

irving2008
18-08-2009, 12:42 AM
Well, according to the calculator:

- 14mm root diameter (subtract 2mm due to thread size)
- 1800mm length between bearings
- B-style end fixing (one fixed, one floating)

Maximum speed is 522RPM.

Even with a 16mm root diameter, it's 597RPM.

(But I know engineers like to put some extra for safety, so it probably would work up to 700RPM).

Or do you have another way of calculating it?14mm root, C style fixing, 1850 long, 727rpm

AdCNC
18-08-2009, 01:17 AM
ok. goin for a thiner screw would drastically reduce starting and stopping inertia and rotational mass! meaning that i dont have to go overboard with the motors! do you think a small 430oz-in NEMA23 gear say to 5:1 would move a Y axis of a 6ft (5ft effective) long table with say 16mm OD screw? this got me wondering because having studied joe's 4x4 CNC videos and pics, it look to me like he's using NEMA23 motors with a long screw, more friction than ball screw and his machine has a inch lead, and its surprising to see thats it flys!

AdCNC
18-08-2009, 01:20 AM
oh btw i would be using 2 motors and 2 ball screws one either side of the x axis

Gary
18-08-2009, 08:31 AM
If you are set on a smaller screw, and by the looks of it, you could get away with it, then it may work, but you also need to take into account the mass you are moving.
the heavier the Y axis the more mass the X has to move, and if you want it to move quick, that is a lot of inertia it has to overcome and that requires power.
You allready have the motor, so you can allway try them, but i dont think you will get the speed you want, especially with the current drivers.
The voltage of the drivers is the killer, so i suspect you will need to change them to a higher voltage driver.

What are you making the frame out of? if it is quite large it needs to be as ridgid as possible, or any small amount of flex will be seen.

We are starting to develope a Y and Z axis combined but it is made from steel, so is quite heavy.
The attached pictures shows a large Z axis (400mm stroke and a 34 frame motor) and this weighs about 15 KG, so i suspect our Y axis would be about 50KG.

tribbles
18-08-2009, 09:49 AM
14mm root, C style fixing, 1850 long, 727rpm

Yeah, having two fixed ends does improve things - although on my machine it's one fixed (BK), and one floating (BF), and I would guess that's what the standard machining on Gary's ballscrews would provide.

I suppose you could use BK on both ends, but you'd need to make the ends with a lot of precision (or have a number of different size shims to pad out the end with the screw-thread).

Gary
18-08-2009, 09:54 AM
You would need to have the fixed end machining done at both ends and the fixed end machining is were all the money is so it would be more expensive.
I dont do the machining, this is done at a local engineering company so i would need to get a quote.


Yeah, having two fixed ends does improve things - although on my machine it's one fixed (BK), and one floating (BF), and I would guess that's what the standard machining on Gary's ballscrews would provide.

I suppose you could use BK on both ends, but you'd need to make the ends with a lot of precision (or have a number of different size shims to pad out the end with the screw-thread).

Gary
18-08-2009, 10:05 AM
I have customers who uses this method to get faster motion, but normally would use a servo motor and specific gearing that is only for this type of application.
Unless you get the parts on ebay, it will cost a few hundred just for the gearing.
Also fixing both ends of the ball screw offer similar results (Within about 25%)

Didnt think of that :eek:. Ist the main problem in this case the actual torque needed to spin the screw? Although the extra weight wouldnt be ideal surley it is a lesser evil ? wouldnt it be possilble to also use a smaller screw which would make it cheaper?

Sorry to hijack your tread :naughty:



Isnt that all part of the fun.....:heehee:

John S
18-08-2009, 10:10 AM
Waste of time on something like a router. You fit double thrusts on both ends, adjust to get a preload, and all that happens is the end frames deflect inwards so the screw is still whipping even with double bearings.

The design has to take this into account.
Bystronics have realised this on their new range of lasers, they have massive end castings and tension the ball screw to something like 70 tonnes tension [ no typo ] to stop it whipping.

Mind you they are using twin drives with 15 HP servo motors on each drive :whistling:

I feel the OP needs to sit down, count his pennies then come up with something that's workable instead of throwing down pie in the sky figures for his first build.

Why the insane rapids anyway ? The whole idea is to bury the tool and work at feed rate.

.

AdCNC
18-08-2009, 12:41 PM
iim not set on the smaller motors but as i have them now i may as well give them ago the machine does not have to work mega fast! its being made out of aluminium. ill give anything ago (within reason) if it works ok and saves abit of money! if not then ill have to go for some " muchas maracas ". that Z axis you got there gary looks a fair bit of kit what OD and pitch is that ball screw?

Gary
18-08-2009, 01:16 PM
This version has a 20mm ball screw with a 5MM lead, but can go up to 25mm diameter with any lead.
We are also working on a smaller version that uses a 16mm ball screw, so will be a lot smaller, but i wnat to keed the width to keep the ridgidity.
I only use profiled rail, so no round or supported round rail.
If you have a more details specification, email it over to me and i can work out what you need.


iim not set on the smaller motors but as i have them now i may as well give them ago the machine does not have to work mega fast! its being made out of aluminium. ill give anything ago (within reason) if it works ok and saves abit of money! if not then ill have to go for some " muchas maracas ". that Z axis you got there gary looks a fair bit of kit what OD and pitch is that ball screw?

AdCNC
18-08-2009, 01:33 PM
Thanks Gary ill have to go over a few things then once ive got all the axis up and running then ill drop you a mail with some photos and details. see what you think!

ptjw7uk
19-08-2009, 09:59 AM
If the problem with long ballscrews is whip why not fit traveling intermediate bearings as in travelling steady would only need a bearing in which the ballscrew would slide and then some sliding mechanism to pull the steady back into place when the axis moved to the other end. t least 2 steadies would be needed each side (if double ballscrews) plus space at each end to park the steady.

Just an idea!

Peter

Gary
19-08-2009, 10:36 AM
You would still have a bit of chatter and vibration and i think the cost of this would also be more than the cost of a larger ball screw.



If the problem with long ballscrews is whip why not fit traveling intermediate bearings as in travelling steady would only need a bearing in which the ballscrew would slide and then some sliding mechanism to pull the steady back into place when the axis moved to the other end. t least 2 steadies would be needed each side (if double ballscrews) plus space at each end to park the steady.

Just an idea!

Peter

John S
19-08-2009, 11:08 AM
If the problem with long ballscrews is whip why not fit traveling intermediate bearings as in travelling steady would only need a bearing in which the ballscrew would slide and then some sliding mechanism to pull the steady back into place when the axis moved to the other end. t least 2 steadies would be needed each side (if double ballscrews) plus space at each end to park the steady.

Just an idea!

Peter

It's only circumventing a bad design.

.

vre
02-02-2014, 02:49 AM
I want to retrofit a lathe with servo 1.8kw 3000 rpm.
lathe between center is 1.6m and the leadscrew for z axis is 1.8m diameter 30mm 4TPI
i will replace the leadscrew with ballscrew.
What diameter-pitch ballscrew will be optimal for this situation ?
Also what is better to drive the ballscrew with coupler directly from servo or drive it with pulleys & timing belts ?
What pulleys & timing twill be strong enough and without backlash ?

irving2008
02-02-2014, 08:45 AM
Please don't resurrect ancient threads, start a new one.

Anyway, to answer your question, there is no easy 'optimal' as it all depends. However the easy option is to go near like for like and use 30mm dia 5mm pitch.

However that screw is rarely going to be turning at more than a few rpm (<1 for metric fine threading, <100 for a finishing pass and maybe 1000 for a rapid traverse) and despite it's 11kg weight only needs <1Nm to accelerate it to cutting speed in under 1mm of traverse. So a servo is massive overkill and gearing it down while avoiding backlash is going to be tricky.

vre
03-02-2014, 06:01 AM
Thank you for your answer.
My lathe has prismatic ways not linear ways with bearings.
So you suggest me to go with 3205 ballscrew and direct driving with coupler ?
1000rpm*5mm = 5m/min rapids is 5m/min good linear rapid speed for a lathe ?
Because my servo is 3000rpm with this setup i will get max 15m/min rapids but the question is if the ballscrew can rotate at 3000rpm without problems.

Because of ballscrew inertia higher pitch ballscrew (10mm) at lower rpm and gearing down would be better ?
If i gear down what pulleys-timing belts will eliminate backlash ?