[handlewanker]

Hi Rokag, the problem is that you can't mesh a straight line with a helix angle....the corners of the Acme thread will foul the corners of the roller.

The bigger the roller the more of it's flanks will interfere with the screw flank corners at top and bottom.

What will happen as you have found out in practice, is that the steel Acme screw thread cuts into the bronze straight sided roller and wears it away......this only stops happening when the bronze roller hits the bottom of the Acme thread, and gives you sideways clearance called backlash.

Having studied this phenomena by meshing a screw thread with a straight sided single grooved roller for the study purpose it became apparent that the bigger the roller the harder it is to get the two to interface on depth....about 2/3 rds max.

The only time a screw thread will mesh with another screw thread PERFECTLY is when one is right handed and the other left handed, but it does not work as a drive as one cancels out the other for forward movement, but the two threads are meshing because the opposite sides of each one are at the same helix angle......too bad they don't work.

I suggesrt that the smaller the roller diam the sharper the curvature of the contact with the Acme thread which means that there is less chance of the corners of the Acme thread cutting the straight sided roller thread.

For instance if you had an Acme thread of 50mm diam and a roller of 15mm diam you would have less line contact at the helix/groove faces at the centre line.

If the 15mm roller was meshing with the Acme thread it would have to have bearings for radial support and also thrust races for axial thrust resistance against the frame of the nut, but this would make for very small bearings and even smaller thrust races....not very practical.

So to solve the problem I suggest that the roller is made a small diam, solid with grooves, no bearings or thrust races and held against the Acme thread by two more larger rollers.

The smaller roller is on the centre line and the two larger rollers also grooved are above and below the centre line, 30 deg approx, so trapping the small roller in a triangle....the small roller can now float there both radially and axially held in place by the big rollers.

Having big rollers you can now put substantial needle bearings and thrust races to carry the load which is transmitted axially by the large rollers to the frame of the nut.

To sum up, the Acme thread will apply pressure to the small floating roller axially, and this force will be transmitted by the small roller grooves to the large roller grooves and then to the frame.

The small roller will have a better time interfacing with the Acme screw thread due to it's small curvature that makes less contact with the screw flanks than a large roller.

The larger the roller the greater will be the interference with the screw at the thread top corners.

This is a study using a single roller pack on one side, but for a practical drive it would have at least 5 roller asssemblies spaced round the Acme thread.

It would be quite bulky, but you could have a low profile model by having just two assemblies one on either side of the Acme thread and to prevent the Acme thread from deflecting up or down under load a pair of plain ungrooved rollers would be placed in the frame to contact the top and bottom of the Acme thread.

Adjustment for mesh would be by one of the large rollers having an eccentric spindle, the other fixed.....the small roller floats of course, held in space by the two large rollers.
Ian.

[rokag3]

You focus too much on the depth

The roller i have and working perfectly is made of 3 rolls bigger than the rod.

The other one is working only in one direction (mean it's not usable) and the difference is the size of the rollers.

This problems have been exposed long ago and i regret very much that i did not read it carefully

If you revisit the European site you will notice the planetary threads are much larger than the rod ??? His system must work because he has it on his router. So the different diameters must come into play.

He was talking about hungarian site who as disappear.



and yes it come into play it is in fact the only think that come into play !

this is working


The original roller

[BillTodd]

Good to see some progress

It looks great

It'll be interesting to see how well your design works: It looks to me only one roller is taking the cutting force in each direction (Roller nut Version 5 ) . If so, I wonder if the stiffness of the roller may be an issue ?

I assume the little spring is only a light preload on the thrust bearing - I could not use needle roller thrust bearings on mine as it would have prevented the eccentrics from turning freely (the balls thrust bearing on my design allow the rollers to move radially even under quite a large pre-load)



Bill

[handlewanker]

Hi all, I admit I can't argue fully as I have not made a sample test piece to put my observations or suggestions to the test, but one inescapable fact still bugs the alternative to the ball screw, nobody has made an alternative to the ball screw that gives frictionless backlshless drive with any degree of live expectancy or continuous accuracy.

You can make a set-up that approaches the ideal, but so can a brand new bronze nut and Acme screw thread with compensating back up nut, but due to the forces and materials in use, the Acme thread and bronze nut, like the systems as yet designed and put forward, all wear to the point that eventually within a short time they have backlash and so fail to be alternatives to a ball screw.

The problem is and always will be the fact that when you attempt to mesh a straight face against a serpentine like grooved face, (it looks almost like a groove tilted at an angle) you will get point contact on the top and bottom of the roller when it attempts to mesh with the "angle" of thread.....and that has not been overcome to date.

Perhaps the latest offering as described by Bbutcher85710 in post #795 using the resilience of Delrin for the rollers will solve the problem......only testing will tell.

If it has been overcome and a system is now functional and in use, where is the design, why was it not patented, and why has there not been a groundswell of activity by all who need such a system.

We must not fool ourselves that a system appears to work at first running, when in actual fact it is still new and unworn by repetitional useage.....it must work and keep working without continual adjustment.

The ball screw, like a ball or roller bearing, is working under exact and ideal conditions, and the ball screw replicates the fundamental design of the ball race except it is a linear motion instead of a rotary one.....backlashless, frictionless but costly.

BTW, the design in post #803 by John Jardine will NOT work as an alternative to the ball screw, due to the slip the rollers (as much as one whole pitch) have on the screw....it will work after a fashion as a drive but does not return to zero once it has moved forward and backwards a number of times.....this has alreasdy been discussed in very early posts of this project.

The design uses All Thread on All Thread, which is steel on steel, and so is just an interesting design, but not practical for longevity or accuracy in CNC mode.

BTW, looking at the All Thread screw thread form, you have a 60 deg thread included angle.

This gives you a wide entry and exit "doorway" to the thread profile when meshed with a straight sided roller allowing the straight sided roller to only make contact with the screw thread on the centre line, whereas the Acme thread, with its narrow thread profile of 29deg included angle, makes any contact with a straight sided roller hazadous.

The other problem is we can buy or have precision Acme thread off the shelf, whereas we can only get All Thread off the shelf in non precision gen purpose quantities......pitching of no practical use in CNC work.

I would further say that if a precision screw thread was cut with a 90 deg included angle instead of the ordinary 60 deg angle, and mated with a straight sided grooved roller, the mating would be better in that there would be no point contact at the top and bottom of the thread/rollers interface like you would expect to get with Acme 29 deg thread.

This would then allow bigger rollers with the necessary bearing arrangments for radial and axial loading.

You don't need to build a ship the size of the Titanic to know it will sink if it hits an iceberg....the design will tell you that if you can analyse it.
Ian.

[svenakela]

...

The design uses All Thread on All Thread, which is steel on steel, and so is just an interesting design, but not practical for longevity or accuracy in CNC mode.

...

I disagree on that argument, We've got a "home made" roller screw in an industrial application (like this one), it has been running approx 2000 hours back and forth and is still tight. I think you misunderstand the design, there's (almost) only a rolling contact between the steel parts. They don't wear as a screw and nut does.

[handlewanker]

Hi Sven, I have once again read all the posts to see if I have missed any developments seeing as how the thread went quiet for about a year, but as the original development went......"when you use Allthread and also Allthread rollers, you might get smooth and almost frictionless running , but the accuracy of the nut is lacking due to the slip that occurs when the rollers are threads and not grooves".

This has already been "discovered" and the opinion is that the rollers must be grooved to prevent slip.

I can't remember which post noted this problem, but the slip will be as much as a full pitch which cannot possible be usefull for any CNC application.

You cannot possibly now state that the Allthread and Allthread roller design, although simple, is the answer, seeing as how it was dismissed for CNC use a long way back in the posts.

It is extremely good for simple frictionless transmission, but not accuracy of position.

The other problem is the attrition of the grooved roller "solution", which being straight sided and mating with a helix, will either erode the screw thread or the roller or both.

Therefore the development to date has not solved the problem, only the Delrin resilient roller design is a new development in the attrition of the roller situation, and only testing will see if they can "bend with the wind and last the day".

The Delrin development is a development in roller material only, not design, as the DIY design of the roller screw per se has already been finalised in the shape and format of the straight sided roller meshing with a screw thread, and the Delrin solution, while showing great promise, has yet to be proved.
Ian.

[handlewanker]

It would seem that the thread has gone back to bed to mull over the aformentioned observations.
Ian.

[rokag3]

Hi all, I admit I can't argue fully as I have not made a sample test piece to put my observations or suggestions to the test, but one inescapable fact still bugs the alternative to the ball screw, nobody has made an alternative to the ball screw that gives frictionless backlshless drive with any degree of live expectancy or continuous accuracy.

Sorry I did for my small lathe with big rolls.

You are talking about wear, but the point is that the roller screw roll and that is a problem to extract energy and if you have not much energy you cannot wear fast at all.


no backslash !!! - YouTube

[handlewanker]

will00ard10 is a spambot....don't open...delete.
Ian.

[dsquire]

will00ard10 is a spambot....don't open...delete.
Ian.

Ian

Thanks for the warning but where do we have to watch for it? Is it a file or what??

Cheers :cheers:
Don

Edit

I found out that will00ard10 was originally the owner of post #809 which has since been deleted therefore I didn't see it which caused me to ask the above question. All is good now.

Don

[handlewanker]

To whom it may concern, I have been told that you will normally see the spambots in your emails pretending to be a new poster to a thread you have been posting to, but the spambot is only in YOUR email, and as soon as you open the email to go to the thread on this site (and others you have been on) the spambot has got into your system....the purpose of which is to spread a net of advertising in which the originators of the spambot get paid for every email adress they contact.

The spambot is normally recognised by the stupid identity they post with, and if it gets onto the thread here, it has a short posting history.

The spambot's identity is generated by a computer and is pumped out by the million to automatically enter any email list it currently has on record.......and you can identify it by the peculiar sequence of alpha and numeric characters not previously having been on the site you regularly visit.

When it appears in your email list, it does not appear on the thread it says you are posting in, and it usually has a short meaninless line from one of the posts you have been on.....this is to arouse your curiosity and encourage you to open the email which gives it access to your mailing list and everyone gets inundated with spam spam and more spam.

The greater the mailing list it accesses the more customers the spammers can claim to "service" with junk mail.

Sometimes you will get several emails from the same spambot, posting on different sites that you have frequented, all with short meaningless messages that do not appear on those sites.

When this occurs, the trick is to go to the forum site either from the address you store in Favourites or on a previous post from a regular poster you recognise.

After a while you can spot the crap as it comes in and delete it immediately.

It doesn't matter if you misinterpret and inadvertently delete a genuine email notification from the forum, as when you go to the forum you will automatically see it there, it's only a notification for your convenience, but can also be a foot in the door from a spambot if you don't recognise it and open it.

To be 100% certain of not being spammed, don't open any email notifications that appear in your email list, but delete them and go to the forum and the thread.
Ian.

[phil c]

Haven't read the complete thread, it's been running too long. Basic physics, a screw has to mesh with some kind of screw to provide positive feed. How about a "string" nut- Think of a ballscrew with the balls replaced by a loop of plastic, something about the thickness of a weed whacker string. feeding through the ballscrew track and then welded into a continuous loop.

As the power screw turns it would carry the "sting" along with it. The string would come out of one end of the nut and loop back around to the other end. With the right size string made out of plastic with some lubricant additives, and a nut with a similar composition the friction would not be too great. Clearances would be controlled by the diameter of the string. The right size, determined by experiment, would minimize backlash.

Haven't built it, but feel free. As of now, if it works it isn't patentable. The idea is free.

[handlewanker]

Hi Phil, I like your "free thinking" attitude....many ideas come from what at first seems impractical, but you have to imagine the idea first.

I personally don't think the method will work....you have to transmit the thrust of the screw, be it a mill table (large heavy load) or a router (med load) or a lathe leadscrew, to the nut and thence to the machine.

Although you are replicating the motion of balls as in the ballscrew, the ballscrew is designed around ballrace technology, where the outer race and the inner race recieve and/or transmit the load to the static body.

It is quite possible that the drive that you envisage can be of use for some purpose, even if it is not practical for CNC mill, Lathe or Router useage.

It would pay to give the posts a quick glance to see the progress that has been made to date in the quest for an alternative to the ballscrew for Mill, Lathe or Router use.

Who would have thought that practically all compact cassette recorder/players would be driven by a rubber band, whereas all Compact disc players etc are direct motor drive, and many better quality turntables are also driven by a rubber belt or even a direct motor drive.

Read the posts, develop the thinking further and publish your findings, even if you don't go so far as to making anything yourself.
Ian.

[Zathras]

Phil,

Essentialy this is what you're talking about.

Patent number: 4272997
Filing date: Aug 17, 1978
Issue date: Jun 16, 1981

Patent US4272997 - Drive mechanism - Google Patents

[handlewanker]

Hi, that patented drive mechanism will work as a drive, but will not replace the ball screw for accuracy..... economy maybe, which is what this thread is about.

I think the focus on accuracy is the key word, because even though you can drive a router table forward to an accumulative error of +/- .010", using just off the shelf Allthread, the moment you reverse the drive and experience backlash, that's where it all starts to unravel and become a head scratching moment, and when in the quest for accuracy you go further down the path to more expensive but better made Acme thread with compensated bronze nut, the backlash makes the gains no better due to the backlash.

The whole point of CNC is to not only go from A to B, but to return to A....and if you have a return position of A+ .002", you have backlash......possibly acceptable for whatever purpose the drive is needed, but definately not a replacement for ball screws.

If in any drive the backlash can be tolerated or compensated for by program control, the simplest drive is then the best when all things are equal, and you can't beat the simplicity of an Acme thread drive with compensated bronze nut.

This is then where the roller screw drive lies, as it does not fullfill the need for anti backlash control, so why use it if it is no better than the Acme thread equivalent.

When I worked a horizontal boring mill the position of the table was decided by the readout of a light vernier, that is, you just kept winding the handle until you got from position A to position B by refering to the light vernier readout.....the turning of the screw's accuracy played no part as to where you started from and where you go to, you were at position A and you got to position B and then back to A without backlash being a factor for accuracy....it was absolute.

When a computer, and I don't know if it can, is able to read the position of A as zero and move the table to position B and back to A without error from backlash, the need for semi accurate ball screw substitutes will have been solved.

I would think that the digital readout with glass scales can be the answer to incremental movement control as opposed to obtaining position by the pitch of a screw, provided the computer control can sense the variations that occur from the digital read out head and know where it is at all times.

I don't know enough about computer control to know if this is a practical or expensive way to have absolute accuracy, but I do know that is the way I would prefer to be at when I want to go from A to B and back to A.....I got so used to the absolute accuracy of position from my horizontal boring mill days.

In temperature controlled environments the vernier or glass scale is absolutely accurate to within it's manufacturing tolerance, and that could be as little as an error of .02mm over a distance of many metres, and that is far more accurate than 99% of all people involved in CNC machining would or could work to ever.

Perhaps someone with the knowledge of computers and digital readouts could enlighten me on this subject, not just digital readout positioning, which is common knowledge, but the interacting of a digital readout and a computer control.

This would make the CNCíng of any machine possible without the need to replace Acme screw drives with ball screws, but only a stepper or servo motor to work with computer readouts.

As I see it, the computer would compare the reading it gets from the read head to the position it is at, and just keeps sending pulses to the stepper or servo motor until the position required is attained.....very much like a GPS where you move to a co-ordinate by whatever means you have either walking or driving a bus etc, the position is absolute no matter how many steps you take or revolutons the bus wheels make to get there.

If I were to simulate a computer, I would have my finger on the pulse button while observing the digital readout, and by pulseing the button get from point A to the point B position then back to point A....no error.

The pulse rates could be one pulse = .02mm of the screw, which is really irrelevenat for positioning, just as long as it turns the screw in fine steps, but as I'm driving a screw with bronze nut, the number of pulses might be more or less than the pitch of the screw, the final position would be the stopping point of the pulseing which I know from watching the digital readout.
Ian.

[Zathras]

I wouldn't use the patent either! I have thought about this type of idea but had to dismiss it in the end. Even the low cost C7 Chineese ballscrews would be better than this idea IMHO.

I think one reason someone would use the roller screw over an ACME say, is perhaps for high load and stiffness applications which require low friction.
To make a DIY roller screw, the more components you have, the more difficult it is to keep stiffness and backlash in check surely. I do like the roller screw idea (and the "Everman" type I have my eye on as well) but I'm not convinced it's worth it overall. It might be fun to try it one day. :wee:

Backlash and linear accuracy of whatever mechanical actuator can be compensated for using a linear encoder in the motion controller loop. The encoder feedback to a PID control loop does what you need and the actuator shouldn't matter that much (roughly speaking).
If you wanted to, the linearity can be mapped along the length and the backlash, even if it varies along it's working length, can be compensated for in the software. Now you could remove the encoder for open loop control. Obviously there are overheads doing this so not sure if anyone bothers to do this. It's possible in principle anyway.

I would use an encoder feedback for long actuator lengths anyway but that depends on the application of course.


-Mark

[handlewanker]

Hi Mark, the biggest problem with backlash and frictionless slideways is that when you come to the reversal of the drive there's nothing holding the table from being pulled by the cutter......with a ballscrew and stepper motor the combination of tight ballscrew and stepper holding power will ensure you don't wobble at the moment of reversal.

Imagine plunge milling out a cavity 100 X 100 X 50, and you have backlash with compensation.

With linear slides of the Hiwin type the sensitivity will be lost due to the need for some braking to prevent inadvertent cutter grabbing.

Once the roller screw design overcomes the self destruction rate, then the design will get somewhere, it has promise but no final solution.

Perhaps the Delrin roller in the design will work, but I have the feeling that the resilience of the plastic material under load might just add a small percentage of backlash, as in spring from compression, and that takes you back to the compensated spring loaded Acme thread backnut result.

It's no backlash, or no solution.

I can read a digital read out to go from A to B and back to A without worrying about lost motion from backlash, as I'm reading the position of the job, not the movement of a loose nut on a screw thread, and I can do it using an Allthread drive, so why can't a computer do the same but faster and on the move.....no need for a highly expensive screw drive.
Ian.

[BrianTee]

Quite the thread! I've thought about the challenges of precision for the DIYer to create the rollers for the roller screw and think this may be a different direction.

While i am a farmer and have a lathe and mill at my disposal often i do not have 'every' tool. This idea is for the tinkerer like me. Not an industrial solution, there are precision parts for that.

Putting a bearing on approx a 45 degree angle and running ACME thread inside the bearing creates an interesting design. I have made a little video to illustrate what i mean

Have a look at: Roller screw for DIY CNC - YouTube

To finish the idea i need to plant my crop first but basically it will consist of that angled bearing running in the grooves of the ACME thread and a roller to ensure the bearing stays in contact with the thread. Simple in theory and hopefully in practice too. Best part, simple, cheap readily available parts, no special tools, self cleaning, no backlash, and negligible rolling resistance

Any thoughts on this anyone? possibly 2 bearings at opposite angles to eliminate torque - like a V. Getting both bearings exactly in the groove may be tricky.

Brian

[handlewanker]

Hi, as gently as possible so as not to offend you.......this has already been done or rather "prophesied" as a possible solution, but while it will follow the Acme thread for a while it will erode the thread away until in the end all you have left is burred over rounded corners on the screw because the corner of the bearing, being at 90 degrees carves itself a form on the screw and eventually bottoms leaving you with backlash and a worn out screw.

It's nice to experiment, and we all keep an open mind to possible solutions that given more work could supply a cheap alternative drive to the ball screw.

Congratulations for getting involved in the attempt at a different screw drive.
Ian.

[BrianTee]

Oh hey, no need to be gentle, just couldn't find on the site. Its a tough one to make simple. Eliminate the scrub, eliminate the problem. Thinking.........