Feedscrews

[Roger YarwoodParticipant @rogeryarwood66088]

Hi all, i'm working on a project that is basically a custom CNC lathe for a specialised and unusual purpose.

One of the key elements is the feedscrew and its drive/control.

I'm not a mechanical engineer so i don't (yet) know all the correct terminology.Can anyone recommend any further reading on the selection and use of feedscrews and associated drive. Plus the technology behind positioning/measurement eqpt used on cnc equipment.

Any terminology or keywords for research would be much appreciated.

Thankyou

[Roger YarwoodParticipant @rogeryarwood66088]

[JasonBModerator @jasonb]

Well you could start by Googling "Ball Screws" there are what most people use for CNC conversions to eliminate the backlash associated with the usual Acme type feed screw and nut. Then "Stepper Motors" to surn the screws.

J

[David JuppParticipant @davidjupp51506]

Or Servo Drives instead of steppers where higher performance may be required.

[RobboParticipant @robbo]

Roger,

for details of ballscrew feedscrews etc, have a look at Marchant Dice Ltd -www.worldofcnc.com.

Phil

[Chris SheltonParticipant @chrisshelton11794]

A wealth of information, and advice can be found on MYCNCUK and CNCzone.com in USA.

Hope this helps.

[John McNamaraParticipant @johnmcnamara74883]

Hello Roger

A little more information would help?

Does your particular application require high accuracy say .002mm or less or will within a millimetre do?, and what speed in millimetres per minute does the feed or feeds's need to attain? Knowing this will make it easier to determine a drive system. What power is the lathe main drive spindle motor, to give an idea of the cutting forces? Assuming it is not a winding lathe.

It would also help to know the working environment. screws and ball screws need to be lubricated and they need to be kept clean.

If your application involves wood the ball screws in particular will need to be kept away from wood dust or metal swarf, wood dust will form a cloud around the machine, Acme screws also will not function well unless kept clean,

As your application is "specialised and unusual" maybe a conventional approach is not appropriate. For example for some applications a simple toothed "timing" belt drive is sufficient, or rack and pinion.

For drives; meaning a motor and the necessary driver electronics, the line between servos (brushless AC preferably or DC brushed) and steppers (with or without encoders that close the positioning loop thus avoiding missed steps), is becoming blurred. Steppers will most of the time cost less overall. I would look at steppers first.

Will special software be needed or can a standard package like Mach3 or linuxCNC provide the functionality you need? Artsoft Mach3 in particular has an active user group,

If it is a one off special machine that only does one thing there is not a lot to be gained from exceeding the requirements of the product it is making.

Regards
John

Ah Ha! Just noticed your call to engineers post Roger… can you define the machine a little more clearly please.

Edited By John McNamara on 22/07/2013 16:55:34

[Tony JeffreeParticipant @tonyjeffree56510]

Posted by JasonB on 22/07/2013 10:17:14:

Well you could start by Googling "Ball Screws" there are what most people use for CNC conversions to eliminate the backlash associated with the usual Acme type feed screw and nut. Then "Stepper Motors" to surn the screws.

 

J

I know that somehow it has become ingrained in the collective conciousness that ballscrews = zero backlash, but that is just not the case. Any screw system can be made zero backlash with suitable anti-backlash treatments – double nuts, preloaded nuts, etc. etc., but you don't get it for free, even with ballscrews.

Ballscrews come in a wide variety of grades and accuracies, just like conventional screws; for example, the mechanism used for raising and lowering hospital beds often uses ballscrews, but these are installed because they are more efficient than conventional screws, not (for obvious reasons) because high accuracy or zero backlash is needed, and you certainly don't get either with that grade of screw.

The two primary advantages of ballscrews over conventional screws of equivalent accuracy are high efficiency and low wear rate. If you want low backlash as well, then, just as you do with conventional screws, you pay a premium.

With lathe use, for most purposes you don't actually need zero backlash anyway – on the cross-slide, if you are turning the outside of a part, and you are actually cutting metal, then there is always a component of the force on the tool that is pushing the tool away from the work in the direction of the cross-slide feedscrew, so any backlash is always taken up in that direction. The only problem you get is if you change from cutting on the LHS of the tool to cutting on the RHS, as would happen (for example) in machining a complex profile like a chess piece. In that case, as you change flanks, any backlash taken up in the leadscrew would reverse direction also. However, if what you are doing doesn't involve such a change of cutting flank, then you can cope with backlash in the leadscrew with little problem.

My ML& CNC conversion uses the standard screws both for the cross-slide and the leadscrew – I have not had any problems from that source so far.

Regards,

Tony

 

Edited By Tony Jeffree on 22/07/2013 17:15:55

[John Stevenson 1Participant @johnstevenson1]

Posted by John McNamara on 22/07/2013 16:42:25:

Hello Roger

A little more information would help?

Does your particular application require high accuracy say .002mm or less

If it does he's on the wrong forum, 0.002mm is about 0.00007" which is far greater than the thermal expansion of steel.

[John McNamaraParticipant @johnmcnamara74883]

Hi John…. The size of a large pothole to these machines … **LINK**

[John Stevenson 1Participant @johnstevenson1]

Damn,

That's what it was.

Scrapped one of those last week, found it in this pensioner's old shed up in Rochdale.

He told me it was useless and would only hold 0.0006"

[Roger YarwoodParticipant @rogeryarwood66088]

Thanks for all the responses, i'm already losing track of what i've PM'd to who but i will try and PM all interested with an outline of what i'm trying to do. If i missed anyone please PM me.

I've already researched "off-the shelf" leadscrews at places like Zapp automation and others, but i do not yet have the knowledge to specify what i need.

I'm basically engraving a flat plastic disc with a spiral groove approx 0.002" wide that is modulated by a special dual voice coil cutting head. The material being removed (using a heated diamond tipped stylus) is a fine thread like hair. So the cutting forces are minimal. Modern stepper motors have been tried in this application, but even the best steppers modulate the groove which is unacceptable for this application. One key goal is to make a fairly heavy/solid machine in order to achieve the necessary accuracy/ detail.

My machine is dedicated to one purpose only, operated in a clean environment so there are no issues with dust or swarf. No cutting lubricant required. There will be a vacuum unit that removes the fine thread of material as it is being removed from the disc

Anyway i will try and PM you all an outline of what i want to do, PM me if i missed you

Thanks again all. And PS please see my other thread for more info: http://www.model-engineer.co.uk/forums/postings.asp?th=85858

Rog.

Edited By Roger Yarwood on 25/07/2013 03:20:04

Edited By Roger Yarwood on 25/07/2013 03:30:37

[russellParticipant @russell]

sounds suspiciously like a lathe for cutting LP records (remember them? they were the 'big black CD's' my son used to beg me to play….)

-russ

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Hi Roger ,

As Russel says your system is very like a stereo LP master disc cutting machine . The drive in a CD unit is one level more sophisticated but again basically the same .

Design and detail choice of components just comes down to simple sums – basically what resolution and accuracy are needed .

You don't give any general dimensions but for illustration lets examine a track running at about 50 mm radius with data in a track 0,1 mm wide and another 0,1 mm solid wall between tracks .

Lets assume for simplicity mono recording with a rather poor 10 bit data resolution .

10 bit is approx 1000 levels so your 0,1 data track has to have dimensional variations on it amounting to 1/1000 of 0,1 mm . That is 0,1 of 1 micron .

Interpretation of variations will have to be a bit vague because it could be depth , single sided width or double sided width – but it doesn't matter much for a general discussion .

Anyway – drive system for disc has to advance cutter head with an accuracy and resolution of better than 0,1 of 1 micron for no degredation of recorded data to occur .

In practiceyou would probably want to do better than this .

There is also probably a requirement for strict synchronism – accuracy in time – have to look at that separately .

Accuracy and resolution of 0,1 of 1 micron is achieved in scientific equipment and in the very best of CNC machine tools by essentially mechanical means but anything better requires sophisticated adaptive positioning systems ..

A separate consideration is jitter of the cutting head . No good achieving fantastic accuracy with a non rigid set up so design of spindle bearing systems and slide way systems will be critical .

MikeW

Edited By MICHAEL WILLIAMS on 25/07/2013 09:36:38

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

 

 

Edited By MICHAEL WILLIAMS on 25/07/2013 09:31:10

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

You haven't told us what you want to do in detail but if your system is anything like an acoustic recording system then electro-mechanical systems really reached the end of the road a while back .

Perhaps an adaptive system using sophisticated position monitoring and feedback would be best option if you want to stay with mechanical based systems .

Either that or design a system where intrinsic accuracy doesn't matter so much – possibly by :

(1) Pre-cutting the essential groove to only reasonable accuracy and then using it as a guide track which the data cutting head can follow .

(2) Distributing data along lengths of track rather than in spot positions and by using encoding algorithms .

Personally I would start in a different place and look at things like scanning laser and photomask etching or alternatively see if I could adapt computer disc technology .

MikeW

Sorry if answers are a bit scrambled – I've only got odd minutes to think about your project this morning – happy to discuss further later on .

[Roger YarwoodParticipant @rogeryarwood66088]

Yes it is a recording lathe i am building. The machine consists of two main parts:

1)The cutting head which is two voice coil motors that control the cutting stylus.

2)The lathe that has a turntable (or chuck) that rotates the workpiece (plastic disc) at constant speed, and a carriage that moves the cutting head across the disc at variable speed.

If you look at this link you will see the similarities with an engineering lathe

**LINK**

I'm hoping i can use parts from an exsisting lathe to provide me with a carriage and slideway (apologies if the terminology is incorrect). The chuck/ turntable part will have to be custom made as will the cutting head. But i'm reasonably confident that i can find the other parts in exsisting engineering lathes, and just modify them for my purpose.

But we digress: The purpose of this thread is for me to understand leadscrews, their different types, pro's and cons etc

Can we keep the discussion on leadscrews please. I need to learn about them so that i can specifiy what's needed. I do not need or want to develop a new method of doing this as what already exsists does the job perfectly and will not benefit from being changed

Thank you all for your input. Is there any reading material anyone can suggest i read to further understand leadscrew types, applications and pro's/cons etc

Pre-cutting a groove will not be suitable as the groove spacing is different for each job and depends upon the material being recorded. (louder music has wider grooves and quiter music smaller grooves) thats why the feedscrew has to be variable speed, controlled by the amplitude of the sound being recorded.

Basically i need a leadscrew that can travel about 5" in about 30 minutes in one smooth operation, and be moving all the time. I need to learn how to achieve this.

Thankyou all

R.

 

Edited By Roger Yarwood on 26/07/2013 07:06:27

[John McNamaraParticipant @johnmcnamara74883]

Hello Roger

Gee The last time I recommended this book it cost less… 70 bucks now, secondhand… New is off the planet! 1.5k it is becoming hard to get.

Precision Machine Design By Alexander J Slocum

**LINK**

Anyway it is the book you need. you local university library will have it or your local libraray?

It will answer all your questions.

Regards
John

[John McNamaraParticipant @johnmcnamara74883]

Hello again

This paper (free) is an excelent read, also It deals with low vibration design, Clearly a priimary objective of your lathe design.

Note the heavy construction of a Scully lathe in the link below.

Paper:
**LINK**

Scully Link:
**LINK**

Regards
John

[WeldsolParticipant @weldsol]

Hi Roger

For slides I use Moore international for ball screws and Marchant Dice for the bearings and rails and make the rest

As for control systems I use Siemens Logo units ( some machines I build I use the Siemens touch screen along with the Logo unit

For motor drives servo's with encoders I use Lojer units

Paul

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

Inventors and mousetraps again ?

MikeW

[KWILParticipant @kwil]

For those interested in the real (ie old) recording world, try this http://www.manley.com/lathe.php

[Sub MandrelParticipant @submandrel]

Hi Roger,

The positional accuracy of the spiral scratch is not hugely critical. You have to fit about 800 grooves into 3 1/2 – 4" for a well packed disk. That's about 4 thou per revolution and an ordinary 1mm pitch screw would provide that easily.

Unlike the motion of the cutter or the speed of rotation, the rate of sideways motion of the cutter to form the spiral is not critical, as long as it is smooth. Bear in mind that, to maximise the amount of recorded time on an LP the rate of sideways movement is varied when cuttinga master, to make the quieter/less bass-rich passages closer together. Such sophistication is probably a later step for you, and a simple geared motor arrangement to operate a 1mm pitch screw at around 10-12 rpm may well be adequate for proodf of concept, or even actual production use.

Neil

[AndyfParticipant @andyf]

It occurs to me that if the requirements for no appreciable "noise" getting on to the recording are to even approach those mentioned by Michael Williams yesterday, you may be hard pressed to find a screw which has no tiny imperfections on its threads which take the noise outside those limits.

Andy

[jason udallParticipant @jasonudall57142]

….and since advance is unidirectional backlash can be ignored.
Btw radial or tangential arm tracking.?…. (ducks for cover)

[Author]

Posts