PDA

View Full Version : Why aren't ballscrews more common? The tech is old and no one likes backlash, right??



abn
07-14-2006, 12:07 PM
Title pretty much says it, the tech has been around for a long time especially in steering gears etc. I guess cost is the obvious factor...But I've read plenty about the lengths manufacturers have gone in the manufacture of precision lead screws from master screw generation, to complex temperature compensation and control, form grinding, error mapping, error compensation mechanisms, etc. Seems like after that kind of investment the cost of making a ballscrew actuator would be negligible.

Evan
07-14-2006, 12:27 PM
Ball bearings don't handle shock loads at all well.

Ball bearings don't handle contamination at all well.

Ball screws will reverse drive under load.

Lubrication is critical, not self lubricating.

Wear compensation is difficult.

Load carrying capacity is less than acme.

Expensive.

SGW
07-14-2006, 12:57 PM
I imagine a ballscrew is just as difficult to make accurately (maybe more so) than a regular screw.

Rustybolt
07-14-2006, 01:17 PM
I imagine a ballscrew is just as difficult to make accurately (maybe more so) than a regular screw.

Ground precision acme screws can be just as expensive. What Evan said.

Milacron of PM
07-14-2006, 01:24 PM
Evan points out some good reasons. The bottom line is that ballscrews move too easily on a manual machine that has no servo motor connected to give it some 'drag'. Thus making climb milling problematic without clamping down on the gibs. But then you don't want to do that as it accelerates way wear...thus ball screws are undesirable in a manual machine.

SGW
07-14-2006, 01:26 PM
Maybe my previous replly was ambiguous. I was trying to suggest that ballscrews aren't magically more accurate than a regular screw; they are subject to all the errors in manufacture as are regular screws. They don't have backlash, but that's not a source of error as long as one always takes up travel in the same direction. And they do have all the problems Evan listed.

BobWarfield
07-14-2006, 02:49 PM
Actually, ballscrews are very common. So much so that they're readily available in the surplus market for a fraction of their original cost. From a design tradeoff and cost standpoint, I think they're used everywhere it makes sense. They're very common on CNC machines as well, but not so common on manual for all the reasons Evan mentions.

I think a good understanding of backlash also helps explain this further. Backlash is just not as pernicious to the manual world as it is to the CNC world for a variety of reasons. Yes, it can contribute to chatter and other ills, but there is a lot going on to offset that. The much greater friction of the ACME screws versus ballscrews for example.

On the flip side, CNC really has a problem with backlash and there aren't many factors that mitigate this. You can do things while profiling on CNC that just are not really possible to do manually and while doing them it is impossible to compensate completely for backlash. It's okay to take up the backlash in manual style, but what if the part you are machining requires arbitrary direction changes in mid cut with no opportunity to take up that backlash? Can you turn the handwheels exactly right to cut a smooth circle, ellipse, or other exotic shape directly? Probably not, but it's routine for CNC.

Remember too that while many CNC machines are closed loop, meaning they can sense what the machine does when a screw is turned, it is a very limited closed loop compared to a manual operator's ability to tell what's happening. Heidenhein has a fascinating article on the value of using linear (DRO) scales with CNC machines to augment their basic closed loop encoder capability that helps illuminate some of these issues. Backlash is a source of unintended motion that lets the controller's notion of where the machine is subtlely "drift" off target. The linear scales bring it back to reality, but if you can just eliminate the backlash directly, it's a much cheaper solution than trying to bandaid around it.

Best,

BW

John Stevenson
07-14-2006, 03:08 PM
Actually, ballscrews are very common. So much so that they're readily available in the surplus market for a fraction of their original cost.
BW
Very true, a guy local to me has got the job of refurbing a load of hospital beds.
These have ballscrews in with double nuts on the raise lower and tilt.
They have to come out and be replaced as part of the contract.
Looks as if some have never been full travel in their lives, some still have the plastic transit wrapping on them but they will have to be swopped.

[edit] just remembered there are some behind the seat of the truck, where's me camera ?

I have even seen cheap black rolled ones on window opening mechanisms as they need fast travel with little effort.



.

Milacron of PM
07-14-2006, 04:37 PM
Another point to consider is there are all grades of ballscrews...El Cheapo ones with backlash for simple operations like automated close/opening machine guard doors, instrument grade precision ones, preloaded machine tool grade, etc.

It is common for the newbie machinist Bridgeport owner to dream of replacing his standard screws with ball screws, but those standard screws are there for good reason.

retep
07-14-2006, 06:18 PM
Very true, a guy local to me has got the job of refurbing a load of hospital beds.
These have ballscrews in with double nuts on the raise lower and tilt.
They have to come out and be replaced as part of the contract.
Looks as if some have never been full travel in their lives, some still have the plastic transit wrapping on them but they will have to be swopped.

My granmpa has one at his house that he, and my late grama, use. It moves a little wheelchair lift up and down on the side of the stairs outside. The lift is really cheap, well made, but ugly as heck. The ballscrew is completely exposed to the elements and is visibly rusting. Works just fine though and has for about 5 years now. Pretty sure they bought it used too, so tack on another 5-10 years of usage.

I'll bet ya though, it probably wasn't a low backlash model!

Millman
07-14-2006, 06:52 PM
It sure would be nice to have Zero ,Perfect backlash. Just don't think it will ever happen in our lifetimes because of frictional resistance. It's always close but not perfect. Years ago I installed ballscrews in old Bridgeport NC'S and tolerances were greatly improved------but not what the company expected. Perfection is out there...just waiting on someone to defeat gravity and resistance. The Upsadasium didn't even work!

Mcgyver
07-14-2006, 06:58 PM
It is common for the newbie machinist Bridgeport owner to dream of replacing his standard screws with ball screws, but those standard screws are there for good reason.

For a long time for the reasons mentioned herein i'd held the belief that ballscrews, while having their place, were often falsely ascribed magical properties, they have their use but aren't the second coming.

The subject of ballscrews on manual machines came up at cnczone and to my surprise one of the very knowledgeable posters is a staunch advocate of ballscrews for manual machines. He says with abec7 with a high preload, the holding power is about the same as acme and with anti-backlash they are much nicer than acmes and the preferred way to go, i can see that for example climb milling or working with coordinates - as he says when you move the handwheel the table moves. Like most of the posts here, I'd not have thought them a first choice for manual machines, but this poster is not one who's expertise I'd quickly dismiss. i'm wondering if we should be so sure?

Millman
07-14-2006, 07:07 PM
How about double reverse Buttress threads?

Evan
07-14-2006, 07:12 PM
You can stop the backdrive problem with enough friction but ballscrews will inherently drive back through the screw. The reason is the high lead required to give enough room for the ball bearings. This gives a large helix angle. When the helix angle on a screw form is above about six degrees then it becomes possible to back drive the screw.

To overcome this problem on a manual machine will require some source of friction instead of the inherent self locking of a low helix angle acme thread. This is why they aren't suitable for manual machines.

BobWarfield
07-14-2006, 07:39 PM
The subject of ballscrews on manual machines came up at cnczone and to my surprise one of the very knowledgeable posters is a staunch advocate of ballscrews for manual machines. He says with abec7 with a high preload, the holding power is about the same as acme and with anti-backlash they are much nicer than acmes and the preferred way to go, i can see that for example climb milling or working with coordinates - as he says when you move the handwheel the table moves.

Yes, I know just what you are talking about. I have observed that particular prescription for the ABEC7 "made for ballscrew" angular contact bearings being made repeatedly. In fact, one will always read from the same poster about the magic of 20TAC47's (these are the ones that cost about $1000 unless you know the right people), the might be barely tolerable but still very expensive 7204CtYDUHP4's (if you don't know those right people), and the "I'd never use them, but they're still better than where you started," 7204BYG's. These bearings are always the cure for greater accuracy, and now I see they are the cure to apply ballscrews to a manual machine.

Call me a rebel, but when something gets to be so single mindedly and determinedly recommended, it just makes me want to look for other alternatives. These bearings may indeed have those properties, but the prescription is presented with such rigid consistency and scorn for any other possibility, that one almost has to wonder whether things can be so black and white or bearings so magical. I guess you can't really say unless you try them, or unless you have found an alternate solution that works equally as well. I will say that it is interesting to read the bearing catalogs, which the same prescriber often refers people to. There are differences in these magical bearings, but they do not look to me to be orders of magnitude differences at all when you look over the numbers. It is also interesting to read something like the Heidenhein linear scale paper where they estimate the inaccuracies just due to temperature change as a CNC machine heats up and compare those to the bearing differences documented in the catalogs.

Speaking of alternate solutions that work well, the Tormach mills, which are getting good reviews everywhere I have read, do not use ABEC7 bearings of any kind, let alone ABEC7 ball screw bearings. They don't even support the screws at the end, which I found particularly surprising. Now I have seen endless threads from any number of posters that argue there is a huge difference between ABEC7 and the other grades and that you just can't use those other grades on machine tools. You may as well just put the inferior bearings on a ("Gasp!") woodworking machine and not bother cutting metal.

Could be there are a lot of factors involved in getting an overall machine system to perform at its peak. Millman has tactfully pointed out that the reality does not always live up to the marketing hype. You may not need ABEC7's in your application, or the rest of your machine may not be up to what they are capable of anyway.

Best,

BW

lazlo
07-14-2006, 07:45 PM
They don't have backlash,

Yes, ballscrews have backlash, for the same reason that every other mechanical assembly does -- it wouldn't move it there wasn't clearance between the mating parts.

The whole reason you see fancy preloaded duplex ballnuts on high-end CNC setups is to adapt for the backlash in the ballscrew assembly.

Millman
07-14-2006, 07:48 PM
Sure, but don't you just love the salesmen...theirs are always the best.

lazlo
07-14-2006, 07:51 PM
He says with abec7 with a high preload, the holding power is about the same as acme and with anti-backlash they are much nicer than acmes and the preferred way to go,

That doesn't make sense. ABEC doesn't apply to ballnuts, so I assume he's referring to the angular contact bearings holding the driving end of the ballscrew?

The drive-side ballscrew bearings take up axial and radial forces due to the servo mechanism. It has virtually nothing to do with the extremely low friction of the ballnut on the ballscrew, which has the undesirable effect of allowing the table to move when the servos/steppers are not holding the screw (during manual operation).

The holding properties of an Acme screw are the result of the friction inherent in the trapezoidal screw form -- doesn't matter if you have Babbitt bearings holding the drive screw.

Millman
07-14-2006, 07:54 PM
Yes, but it's still friction and gravity as the determining factor in all movements.

lazlo
07-14-2006, 07:58 PM
the Tormach mills, which are getting good reviews everywhere I have read, do not use ABEC7 bearings of any kind, let alone ABEC7 ball screw bearings.

The Tormach is a nice setup at a great price-point, but it's still a Chinese machine. Putting 3C ballscrews and ABEC 7 bearings on a Tormach wouldn't make sense -- the accuracy of that gear would be superceded quality of the dovetail ways.

Don't get me wrong -- it's pretty amazing that you can buy a 1500 lb turn-key CNC system for $8,000, but you can't expect it to perform like a $100,000 CNC machine center.

Now, if someone were bored and hand-scraped and re-aligned the Tormach dovetails, that might be interesting ;)

lazlo
07-14-2006, 08:00 PM
Yes, but it's still friction and gravity as the determining factor in all movements.

Sure, but there's no way that increasing the accuracy of the drive bearings will suddenly make the ball nut have less friction, which is the only thing that's going to make the ballscrew "hold" for manual operation.

Millman
07-14-2006, 08:09 PM
OK, that goes back to manual table locks...which change the actual position of the XY axes in relation to the face of the cutter. Or point of the cutter. It's all irrelevant as far as accuracy is concerned because that accuracy depends on 100 other factors, such as how accurately the mechanism was machined and assembled at the point of manufacture. NOW, all that is in relation to the location and setup of the machine as you view it.

lazlo
07-14-2006, 08:13 PM
I have observed that particular prescription for the ABEC7 "made for ballscrew" angular contact bearings being made repeatedly. In fact, one will always read from the same poster about the magic of 20TAC47's (these are the ones that cost about $1000 unless you know the right people),

Bob,

20TAC47's are an NSK ballnut, not an angular contact bearing.

The 7204 is the angular contact bearing he's talking about. I'm guessing, based on his post, that the BYG suffix is a NSK designation for ABEC-7 tolerances.

I don't understand his point in the Ballscrew 101 post, but you can certainly get a C2 or C3 ballscrew and corresponding nut, and use a pair of ABEC-3 or ABEC-5 7204's for about $50 pair. Cheaper if you're patient on Ebay.

ABEC-7 bearings and super-precision ground ballscrews just don't make a lot of sense for a hobby CNC machine. There will be so many other inaccuracies in the design that overwhelm the precision of a ABEC-7 bearing.

Cheers,

Robert

lazlo
07-14-2006, 08:18 PM
that goes back to manual table locks...which change the actual position of the XY axes in relation to the face of the cutter.

But you can't lock the table in the direction you're cutting. That's Don's point about climb milling with ballscrews.

By the way, one high-tech way to overcome the ultra-low sliding resistance of the ballscrew is with a "manual override" -- as done on the Hass TL-1 lathes, for example. You have what looks like a traditional cast-iron hand crank, but in reality, the stepper or servo is holding the screw for you, and the cast-iron hand-wheel is driving a potentiometer input to the CAM software.

I have no idea how this "feels" -- I doubt the software can make it feel like you're turning the crank on a manual lathe or mill, but it's a good solution if you really want ballscrews on a manual machine.


It's all irrelevant as far as accuracy is concerned because that accuracy depends on 100 other factors, such as how accurately the mechanism was machined and assembled at the point of manufacture.

My point exactly!!!

:o

Millman
07-14-2006, 08:20 PM
Robert, you like throwing names and numbers around with authority...sure you aren't a salesman?

lazlo
07-14-2006, 08:24 PM
Bob,

If you really want to aggravate NC Cams, here's a pair of match-ground NSK ABEC-7 7204's for $199:

http://bearingsdirect.com/products/index.php?action=item&id=289&prevaction=search&previd=&prevstart=0

I've used these guys many times before -- they're great.

And if you're willing to "tolerate" ABEC-3 or ABEC-5 tolerances, here's a matched pair of Japanese 7204's for $23.88

http://bearingsdirect.com/products/index.php?action=item&id=325&prevaction=search&previd=&prevstart=0

Cheers,

Robert

lazlo
07-14-2006, 08:26 PM
Robert, you like throwing names and numbers around with authority...sure you aren't a salesman?

Huh? A salesman for what? What am I pitching?

I'm actually a microprocessor architect, if anyone cares...

Evan
07-14-2006, 08:32 PM
I'm not even using angular contact bearings for the thrust bearings in my mill. I'm using deep groove NSK ball bearings, type 6200-ZZ. These have a dynamic axial load capacity of about 300 lbs and I have one on each end of the X and Y lead screws. They are fully metal shielded and are being used for axial load only. I use a separate ball bearing on each end for radial positioning. Deep groove ball bearings have an axial load capacity of about 1/3 of the radial load capacity and make fine thrust bearings as long as the load isn't exceeded.

I have mounted up the acetal leadscrew nut I made and have run it back and forth at about 3 inches per second quite a few times with the table linear bearings set tight for extra load. After perhaps 50 trips back and forth the acetal leadscrew nut has about .003" of backlash when turning the manual handle. This is with only a single leadscrew nut and no backlash adjustment. I will be using paired leadscrew nuts with backlash adjustments and should be able to maintain close to zero backlash with the acetal nuts on acme leadscrews.

Millman
07-14-2006, 08:37 PM
I agree with the acetyl idea, but the Zero backlash also depends on the load. On your mill, that will work just fine because you are not going to have sudden side thrust loads applied instantaneously

Mcgyver
07-14-2006, 11:00 PM
That doesn't make sense. ABEC doesn't apply to ballnuts, so I assume he's referring to the angular contact bearings holding the driving end of the ballscrew?

The holding properties of an Acme screw are the result of the friction inherent in the trapezoidal screw form -- doesn't matter if you have Babbitt bearings holding the drive screw.

lazlo, yes, I should have been more specific, its the matched pair of preloaded angular contact bearings anchoring the screw, i know its not the nut. to your second point though I'm not sure thats right, the screw-nut and screw-anchor are both potential sources of backlash so babbit would be a problem - i guess your point more is on holding power more on that in a sec.

I agree these things can be overkill on a homemade cnc, but that's not the same as saying ballscrews aren't a good replacement for acme (a point which i am not fully convinced myself mind you). also the guy who proffered this is in a commercial setting.

Evan I thought the same (re helix) to which I was told the helix on the screws are the identical (was thinking resolution issues as well), so the only drive force through the screw is because it's more efficient, not because of the helix. The heavy loaded angular contacts add holding power such that its not much different than acme. the only time it would matter though is climb milling (in which case the level of anti backlash you get with ball screw and anitbacklash nut would be so much better), otherwise why to you care about the holding power? ie its your force moving the table against the cutting force. anyway supposedly with the right set up it acts similar to acme holding power, . I can't verify that, I can only state that it comes from someone who i think of as being credible. not sure if you have to lock the axis you're not cutting on, although that's good practice anyway i would think

The points you guys are raising are mostly things I was aware of and had thought true, i only wanted to note that the reasoning is maybe not absolute.

I suppose i inadvertently started a discussion about NC's views etc behind his back. that wasn't the intent, imo he is knowledgeable and I raised it as when a credible person holds a different view, perhaps there's something to learn.

Evan
07-15-2006, 12:44 AM
I was told the helix on the screws are the identical

Identical to what? Acme? That isn't correct. A single start acme thread in a .5" diameter will typically have a lead of .1", 10 tpi. A ball screw of the same diameter will have a lead of from .2" to .5". This puts the helix angle over six degrees and allows back drive to take place.

The efficiency of ball screw isn't mainly due to the use of ball bearings, contrary to popular belief. It's due to the higher helix angles. The efficiency of a screw drive increases as the helix angle increases, up to about 60 degrees. This is the primary reason that ball screws are more efficient.

See Nook ballscrews here:

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

abn
07-15-2006, 03:35 AM
Excellent converation, I'm learning a lot of factors...

I do notice the difference on the chart provided but it seems to be less pronounced as you get into the larger 1" - 1.250" ballscrews (vs. a typical 5 tpi leadscrew on say a Tree or Bridgeport manual mill).




Identical to what? Acme? That isn't correct. A single start acme thread in a .5" diameter will typically have a lead of .1", 10 tpi. A ball screw of the same diameter will have a lead of from .2" to .5". This puts the helix angle over six degrees and allows back drive to take place.

The efficiency of ball screw isn't mainly due to the use of ball bearings, contrary to popular belief. It's due to the higher helix angles. The efficiency of a screw drive increases as the helix angle increases, up to about 60 degrees. This is the primary reason that ball screws are more efficient.

See Nook ballscrews here:

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

John Stevenson
07-15-2006, 06:45 AM
Just to upset the apple cart a bit more the standards have changed over the years on ABEC bearings.

I had RHP's development centre check me some figures.
In 1938 when Bridgeport started fitted ABEC 7 to their heads they were the best they could get.
By 1953 the current crop of ABEC 5 had surpassed the original specs.
By 1968 the ABEC 3 was equal.

So if you have a bridgy before 68 and you fit ABEC 3's you are equal to what they first specified.

Bridgeport in the UK did make mills with ballscrews as an extra.
A guy at a plastics factory just up from me has one.
It doesn't unwind when milling but it's only on small mould work.
Don't know how they worked the friction part out but they do exist as factory machines.

I've fitted matched pairs of ABEC 5's to the ballscrew's on the CNC I'm converting.

http://homepage.ntlworld.com/stevenson.engineers/lsteve/X3/new%20mill2.jpg

Prior to fitting the nut block and wipers.
The reason to use these was space saving, two bearings instead of three and to be honest at today's prices as an OEM they only cost a few pence more than standard bearings.

.

Mcgyver
07-15-2006, 10:23 AM
Identical to what? Acme? That isn't correct. A single start acme thread in a .5" diameter will typically have a lead of .1", 10 tpi. A ball screw of the same diameter will have a lead of from .2" to .5". This puts the helix angle over six degrees and allows back drive to take place.
http://www.nookindustries.com/ball/BallInchAvailability.cfm

Identical - the context is the ball screw he replaced the acme with, not ballscrews and acme in general, that would have no meaning.

I raised the point that the helix angle would be different (at time suffering from the erroneous belief that ball screws usually have a higher lead) leading to back drive problems and low resolution on the handwheel dial, he responded he used the same pitch screw.

I'm not sure why you are are saying the helix can't be the same as the acme - 5tpi is what's on my excello and i believe is quite common (dials end up a nice .200 per ref). the link you gave confirms that the screws to convert a mill at the same helix as the acme are readily available? The context i gave was a mill conversion in a commercial setting so I don't absolutely know that its a 5tpi thread, but its a reasonable guess.


The efficiency of ball screw isn't mainly due to the use of ball bearings, contrary to popular belief. It's due to the higher helix angles. The efficiency of a screw drive increases as the helix angle increases, up to about 60 degrees. This is the primary reason that ball screws are more efficient.

So given that it can be the same helix angle, you now are in agreement that the holding power is the same :D I don't think you're right on this point, efficiency has nothing to do with pitch, that's just gearing. the efficiency is how much of the energy put into turning screw ends up moving the table and the ball screws are 80-90% efficient.

as i said before, i'm not completely converted to this idea (ballscrew in a manual machine), just arguing or rather debating to make it interesting

Evan
07-15-2006, 10:42 AM
So given that it can be the same helix angle, you now are in agreement that the holding power is the same :D I don't think you're right on this point, efficiency has nothing to do with pitch, that's just gearing. the efficiency is how much of the energy put into turning screw ends up moving the table and the ball screws are 80-90% efficient.
Not correct. The efficiency of a screw drive is primarily determined by the lead angle, not the type of drive. This applies equally to ball screws and acme screws. The pitch of a single start acme thread is determined by the thread form and to obtain higher leads multi start threads are used. Multi start acme screws have higher efficiency than single start screws.

http://vts.bc.ca/pics/driveeff.gif

Read here for a detailed explanation:

http://www.roton.com/index.php?section=59#11

Milacron of PM
07-15-2006, 11:03 AM
The subject of ballscrews on manual machines came up at cnczone and to my surprise one of the very knowledgeable posters is a staunch advocate of ballscrews for manual machines. He says with abec7 with a high preload, the holding power is about the same as acme and with anti-backlash they are much nicer than acmes and the preferred way to go, i can see that for example climb milling or working with coordinates - as he says when you move the handwheel the table moves. Like most of the posts here, I'd not have thought them a first choice for manual machines, but this poster is not one who's expertise I'd quickly dismiss. i'm wondering if we should be so sure?

If he hangs out at CNC Zone that makes his knowledge suspect right there ! ;)

Mcgyver
07-15-2006, 11:13 AM
The efficiency of ball screw isn't mainly due to the use of ball bearings, contrary to popular belief. It's due to the higher helix angles. The efficiency of a screw drive increases as the helix angle increases, up to about 60 degrees. This is the primary reason that ball screws are more efficient.
http://www.nookindustries.com/ball/BallInchAvailability.cfm

I agree a thread changes efficiency with the lead angle, but your post was regarding ball screws. if you're claiming the (mainly) efficiency from ballscrews comes from a higher helix angle, that's saying the efficiency of ball screws over acme would not exist if they had the same helix - this doesn't seem right and is contrary to the link you provided

in the link you gave (good one btw) it says


Ballscrews, which use rolling friction in place of the sliding friction of the other screw series, exhibit efficiencies that do not change with lead angle

which seems to be the opposite of your above position that the ball screw efficiency comes from the increased helix? they go on to say that low lead screws exhibit a slight efficiency advantage?

that chart you posted is not well presented or I'm just not seeing it - its not clear what line is is the ballscrews. the text would lead you to believe that the ballscrew efficiency drive and backdrive were horizontal lines of 90 and 80% respectively. the power screw line is clearly shown but the ball screw is not.

A.K. Boomer
07-15-2006, 12:37 PM
Gotta go with Mcgyver on this one,,, lead angle is just the standard on which to judge the different types of drive mechanisms, it has nothing to do with efficiancy and almost everything to do with efficiancy,,, but that does not matter due to both systems in compairison having to use identical lead angels, if there not then this is an unfair compairison --- the fact that some systems may exhibit more drag at higher angles is testimonial to them being less efficient at transmitting motion, generally a system that shows slightly more drag at lower lead angles will show much more as the angle is increased (i.e. acme vs. ball) ((unless there is some kind of a pre-load factor involved))
Bottom line is is you move the handle one rev. and the table moves from point A to point B,,,, what system goes this identical distance with the least amount of drag, and if you want to get fancy what about under load, Neg. load? and point A to point B in a certain time frame, this will mimmick "spike loads"
Lead angle is not a factor because its the standard from which all these measurements are being judged, the comparison is between how different mechanisms handle the identical given lead angles....

BobWarfield
07-15-2006, 02:11 PM
Bob,

20TAC47's are an NSK ballnut, not an angular contact bearing.

The 7204 is the angular contact bearing he's talking about. I'm guessing, based on his post, that the BYG suffix is a NSK designation for ABEC-7 tolerances.

I don't understand his point in the Ballscrew 101 post, but you can certainly get a C2 or C3 ballscrew and corresponding nut, and use a pair of ABEC-3 or ABEC-5 7204's for about $50 pair. Cheaper if you're patient on Ebay.

ABEC-7 bearings and super-precision ground ballscrews just don't make a lot of sense for a hobby CNC machine. There will be so many other inaccuracies in the design that overwhelm the precision of a ABEC-7 bearing.

Cheers,

Robert

Robert, the 20TAC47 is an NSK angular contact bearing. It does not appear in their regular bearing catalog, but rather in their "Super Precision" catalog, page 150. They refer to it as a ballscrew ac bearing, meaning it was specifically intended for the purpose. Their primary magic bullet spec is a limiting axial load of 26 Newtons versus 6 or 7 for the other bearings mentioned.

My only two points were that it probably doesn't make sense to spend $1000 on bearings unless the rest of your machine is up to it (I think we agree based on your $100K CNC comment!) and that to always view these bearings as the solution to every problem just seemed a bit suspicious to me. There have to be a lot of solutions to the problems.

As to whether these bearings can prevent ballscrew back driving, I don't have difficulty believing they might. Given enough preload they are going to increase the resistance to turning the screw. While it is true that the ballnut's friction will not have increased, the overall friction will. There have definitely been some manual machines made with ballscrews as was mentioned. My understanding is that this works because provision has been made to provide the backdrive-resisting friction in another way, even using an adjustable friction lock on the axis.

In any event, this discussion now sounds almost exactly like a CNCZone discussion, which is probably nobody's intent! LOL

BW

PS I did recently find a nice set of Naachi ABEC7 bearings on eBay that I bought for $200. My thought was to use them to make a spindle rather than on ballscrews since there were only 2. I do believe the HSM crowd is more likely to need precise bearings if working on high speed spindles than ballscrews.

Evan
07-15-2006, 06:39 PM
Mcgyver is correct about the lead not affecting the forward drive efficiency of a ball screw. This only applies to the forward drive though and does not apply to acme and other similar thread forms. With any non-ball screw drive the efficiency is proportional to the lead.

When backdriving the lead does affect both acme and ballscrews and that is where the problem lies. As I said earlier ball screws need room for the ball bearings. With the smaller sizes such as are found on the average home shop machine the lead must be made larger than will be needed for a single start acme thread. It is possible but entirely impractical to make a .5 x 10 tpi ballscrew as the balls would need to be very small and fragile.

This means that especially in the smaller sizes a ballscrew will back drive where an acme screw will not because of the difference in lead. If the lead on a small diameter ball screw is reduced to the same as the lead for an equally sized acme then the backdrive problem is minor to non-existent for the ballscrew. You will not find this as the photo posted by John shows and the link to Nook also confirms.

Using preloaded bearings to provide friction is perhaps the dumbest idea I have heard of in a while. Ball bearings have a much reduced load capacity when static, around 50% of the dynamic load. Combining such a high static load with any shock loads that may be present the bearings may well be destroyed in short order.

lazlo
07-15-2006, 07:14 PM
If he hangs out at CNC Zone that makes his knowledge suspect right there !

Ain't that the truth. ;)

There seems to be two groups of folks over there -- the guys who have built hobby CNC machines (some of which are simply amazing) and the guys who enjoy arguing about fantasy machines, granite surface plates as machine surfaces, polymer concrete, c3 ballscrews, ABEC-7 bearings,...

When I'm bored, I sometime enjoy reading through the build logs, which are often from the "quiet" members. There was a recent revelation in the gantry camp to use a torsion box, which resulted in several outstanding designs, many of which were constructed entirely in wood, but were still very rigid.

I was particularly impressed with this guys' table-top mill, made entirely from steel weldments. One of 2 or 3 scratch-built machines that I've seen that are capable of cutting steel:

http://home.earthlink.net/~gmduke/

lazlo
07-15-2006, 07:34 PM
In 1938 when Bridgeport started fitted ABEC 7 to their heads they were the best they could get.
By 1953 the current crop of ABEC 5 had surpassed the original specs.
By 1968 the ABEC 3 was equal.

Wow, I didn't know that. Has the ABEC standard itself changed?

Now that a lot of the bearings are made in China, I hope the ABEC standard isn't going backwards :)

Bob:


I did recently find a nice set of Naachi ABEC7 bearings on eBay that I bought for $200. My thought was to use them to make a spindle rather than on ballscrews since there were only 2.

The most common "end fixity" for a ballscrew is "Fixed - Simple" -- a pair of angular contact bearings on the drive side, and a plain thrust bearing on the opposite end.

But I agree -- if I happened upon a pair of ABEC-7 angular contact bearings, I'd use them in a toolpost spindle, not on a leadscrew.

Cheers,

Robert

Mcgyver
07-15-2006, 07:55 PM
Using preloaded bearings to provide friction is perhaps the dumbest idea I have heard of in a while. Ball bearings have a much reduced load capacity when static, around 50% of the dynamic load. Combining such a high static load with any shock loads that may be present the bearings may well be destroyed in short order.

not sure what you mean, the bearings are used because it eliminates backlash from the screw’s mounting not to provide friction, any friction is the result from the factory preload, I think that the friction results in some holding power (or so i gather) is a by product, not the reason for the setup.


I'm not even using angular contact bearings for the thrust bearings in my mill. I'm using deep groove NSK ball bearings, type 6200-ZZ.


The deep groove should be fine. On the super light duty (super cheap) mill I’m building, I’m using cheapo angular contacts, ready rod and plastic anti- backlash nuts. The angular contact setup is convenient way to eliminate a major source of backlash – granted not the only one, but the point is you don’t have to be using expensive abec 7’s to take advantage of these setups.

Just for kicks, here some photos of my home brew, design as you, go setup

http://i20.photobucket.com/albums/b201/michael0100/cnc%20mill/feedscrewcomponents.jpg

http://i20.photobucket.com/albums/b201/michael0100/cnc%20mill/feedscrewassembled.jpg




Ain't that the truth. ;)


Now you guy’s aren’t playing nice :D are these comments directed at me, I’m a regular poster there?

lazlo
07-15-2006, 09:26 PM
Now you guy’s aren’t playing nice are these comments directed at me, I’m a regular poster there?

No, you make nice stuff Mcgyver :)

Your machine-work is impecable. Could you explain a couple of features?

You have a jam-nut to compress the angular contact bearings in that spindle cartridge-looking bearing mount, what does that massive knurled nut do? Is it putting tension on the threaded sleeve (or is that an extender for the threaded rod)?


the point is you don’t have to be using expensive abec 7’s to take advantage of these setups.

Agree completely -- that was my point earlier in the thread! Although I don't think Nachi 7200's qualify as cheap bearings :)

Mcgyver
07-15-2006, 10:02 PM
thanks for the compliment - i know you guys both used smilies when taking a poke and cnczone so i was just poking back, no stress here.

in addition to design as you go, I'm home shop guy learning as i go as well, so no claims that this is the right way to do it, just the best i could do with where i was on the curve at the time :)

the big knurled 'nut' is a sort of handwell. it just goes on the end of the feed screw secured by a set screw onto the shaft. the super cheap mini mill is cnc so I thought it might be handy for positioning purposed to be able to manually move the table.

Evan
07-15-2006, 10:54 PM
On mine I am using the 6200zz bearing at each end and preloading them lightly with a simple acme thread nut at each end of the leadscrew. This method also eliminates all lash in the leadscrew position.

When I refered to using preloaded bearings as a friction device as stupid I am making the assumption that they would be preloaded well beyond spec. That is the only way they will provide sufficient friction to prevent back drive with a ball screw.

I used preloaded roller bearings on the polar axis of my telescope. Even with a heavy preload they do not provide significant friction. I provide holding friction in that system with an adjustable teflon clutch disk between an aluminum and brass plate. That provides the right amount of friction but with no stick/slip.

To provide enough friction to make a usable difference with a ballscrew would mean loading the bearings close to the point of deformation of the balls and races. This is assuming the rest of the linear motion system has low friction. Something similar to the teflon clutch in my telescope would be a much better idea.

J Tiers
07-16-2006, 01:08 AM
I must be missing something here.....

There seems to be an assertion that the back-driving is solely a function of the angle....... not the technology.

The angle will naturally affect it, but the coefficient of friction has to be part of it also. Assuming the screw is perfectly free to rotate aside from nut friction, if the friction due to the pressure against the screw by the nut is large enough, the sliding (or rolling) won't happen.

Same as setting a box on a slope. If it overcomes friction it slides, If not, it won't. On a smooth teflon surface, it will slide easily and at a shallower slope. Put it down on ball bearings, and it will probably move with a very slight slope.

Should be very similar with screws..... ball screws have less inherent friction than a bronze nut on acme screw. So a far larger helix angle (versus axis of screw, meaning more tpi) should be required to prevent it from rotating an unrestrained screw than would be required for a bronze nut on acme.

Sounded like there was an assertion that ballscrews or acme would have the same critical helix angle.....

Evan
07-16-2006, 02:13 AM
Sounded like there was an assertion that ballscrews or acme would have the same critical helix angle.....

They do, approximately. It's geometry. Below a certain angle the inclined plane begins to approximate a flat surface. A little simple trig will give the force vector.

Ballscrews are not 100% efficient. A usual figure is 90% efficiency. The sine of 6 degrees is .1045. So, at 6 degrees the force is 90 percent axial and 10 percent torque which is nicely cancelled by the 10 percent inefficiency (friction) of the ballscrew. Coincidentally or not acme screws have the same approximate critical angle.

JRouche
07-16-2006, 03:20 AM
"Why aren't ballscrews more common?"

Cost...JRouche

Milacron of PM
07-16-2006, 08:35 AM
"Why aren't ballscrews more common?"

Cost...JRouche

Wrong. Lower grade ballscrews (which still have way less backlash than conventional screws) are cheap.

J Tiers
07-16-2006, 10:44 AM
They do, approximately. It's geometry. Below a certain angle the inclined plane begins to approximate a flat surface. A little simple trig will give the force vector.

Ballscrews are not 100% efficient. A usual figure is 90% efficiency. The sine of 6 degrees is .1045. So, at 6 degrees the force is 90 percent axial and 10 percent torque which is nicely cancelled by the 10 percent inefficiency (friction) of the ballscrew. Coincidentally or not acme screws have the same approximate critical angle.

I think we are talking about two different things....

It reads as if both will allow the load to "back drive" the screw at the same angle.

That sounds completely wrong, as it takes no account of the wild difference in friction of "nut" on screw between the two.

With no bearing friction in the screw supporte, it is a problem similar to the angle at which a mass will slide on a slope. A mass on ball bearings will "slide" at a very small slope, a mass with significant sliding friction takes a much steeper slope....

So the force is the same in either case.... so what? In one it CAN move the parts against the remaining friction, in the other it may not due to the remaining friction being so much more.

Re-read what I wrote in original question........

MikeHenry
07-16-2006, 10:54 AM
The Tormach is a nice setup at a great price-point, but it's still a Chinese machine. Putting 3C ballscrews and ABEC 7 bearings on a Tormach wouldn't make sense -- the accuracy of that gear would be superceded quality of the dovetail ways.

Tormach refers to ABEC bearings in their design document, presumably for the spindle, but don't mention what grade. They use a cartridge approach for the spindle and claim that makes it easier for the user to change the bearings or simply remove the cartridge so that it can be sent back to Tormach for replacement. Is that a reasonable design approach?

They do mention using P4 grade ball screws and that standard apparently translates to 0.0006"/foot in any 300 mm section and 0.0008" over 35" of travel. Where does that grade fit on the p*ss-poor to fantastic scale?

This is an interesting discussion for me as I'm seriously considering purchaing a Tormach mill.

Mike

John Stevenson
07-16-2006, 11:04 AM
Mike,
That is a very reasonable approach and is one that is quite standard in industry to save downtime and allow non critical trained personel to swap cartridges out.
As regards the grade of ball screw to be honest it's probably to a greater tolerance than the machine can handle
At these tolerances you are close to needing a climate controlled environment to house and use the machine.

To give Greg Jackson his due, this guy seems to have done a lot of homework before bringing this machine out.

.

wierdscience
07-16-2006, 11:28 AM
There are quite a few of these installed on B-port mills,they don't self-feed and they have the exact same helix angle as the original acme screws(.200"RH).
They make climb cutting a dream since they virtually eliminate backlash chatter.

http://www.rockfordballscrew.com/preload-bridgeport-clonemill.htm


They have been around for years and the last price I noticed was $1850 for a 9x42 kit.Sounds expensive,but worth the money IMHO.

BobWarfield
07-16-2006, 12:59 PM
There are quite a few of these installed on B-port mills,they don't self-feed and they have the exact same helix angle as the original acme screws(.200"RH).
They make climb cutting a dream since they virtually eliminate backlash chatter.

http://www.rockfordballscrew.com/preload-bridgeport-clonemill.htm


They have been around for years and the last price I noticed was $1850 for a 9x42 kit.Sounds expensive,but worth the money IMHO.

Although that same sight on their Ballscrew tech data page had this to say about back driving:

Back Drive Torque: The torque produced through the screw shaft by a thrust load on the ball nut. Ball screws can coast or backdrive due to the high efficiency of the mechanism (90%). If back driving is not acceptable, a method to resist the overturning backdriving systemic torque, such as a brake, will be required to hold the load. If backdriving is desired, the lead of the screw should be at least 1/3 of the screw diameter. Ideally the lead should be equal to the screw diameter. This calculated torque is the minimum amount of braking torque to hold the load in position.

Sounds like the machines you mention must have provided such a mechanism.

Best,

BW

BobWarfield
07-16-2006, 01:13 PM
Tormach refers to ABEC bearings in their design document, presumably for the spindle, but don't mention what grade. They use a cartridge approach for the spindle and claim that makes it easier for the user to change the bearings or simply remove the cartridge so that it can be sent back to Tormach for replacement. Is that a reasonable design approach?

They do mention using P4 grade ball screws and that standard apparently translates to 0.0006"/foot in any 300 mm section and 0.0008" over 35" of travel. Where does that grade fit on the p*ss-poor to fantastic scale?

This is an interesting discussion for me as I'm seriously considering purchaing a Tormach mill.

Mike

Mike, check their exploded parts diagram, available here:

http://www.tormach.com/document_library/Exploded-View-Parts_Lists.pdf

It lists the angular contact bearings for the ballscrews as being "ABEC7202B/P5", hence they are P5 DIN = ABEC5 quality bearings.

Here is the interesting thing about the ratings, BTW, if you've never looked at the numbers:

ABEC 1 Radial Runout = 0.00012"
ABEC 3 RR = 0.00008
ABEC 5 RR = 0.00006
ABEC 7 RR = 0.00004 (Classy machine tool ballscrew bearings)
ABEC 9 RR = 0.00002 (High speed spindle bearings)

These are for a 20mm diameter, and they will vary with larger and smaller bearings, but they are indicative. Now are we really going to decide that 2 of these nth's is the only difference between a junk Chinese machine and a $100K CNC center?

No, of course not. There are many many differences between the two machines, and that is but one of them. I think John's remark that the old ABEC7's are the new ABEC-I-Wouldn't-Use-Them-On-An-Electric-Screwdriver is the telling remark. Use some decent bearings, heck use ABEC7's if you can get a deal and it makes sense, but don't assume it's a silver bullet or an absolute requirement for many HSM applications.

Best,

BW

PS: The "B" on the Tormach bearing is probably the more important thing than the P4. It specifies a 40 degree contact angle which will give you a lot better result in the dual bearing ballscrew application. Those really expensive "TAC" bearings have even more angle there as I recall. As NC Cams has said, this is all in the bearing catalogs which are free to download and make fascinating reading.

Milacron of PM
07-16-2006, 01:39 PM
There are quite a few of these installed on B-port mills,they don't self-feed and they have the exact same helix angle as the original acme screws(.200"RH).
They make climb cutting a dream since they virtually eliminate backlash chatter.

Are you basing this on actual experience, or just rehashing the ad copy ? And are you saying that you can climb mill with these ballscrews on a manual machine without using power feed, and without tightening down the gibs, and have no table "runaway" issues whatsoever ?

wierdscience
07-16-2006, 01:55 PM
Although that same sight on their Ballscrew tech data page had this to say about back driving:

Back Drive Torque: The torque produced through the screw shaft by a thrust load on the ball nut. Ball screws can coast or backdrive due to the high efficiency of the mechanism (90%). If back driving is not acceptable, a method to resist the overturning backdriving systemic torque, such as a brake, will be required to hold the load. If backdriving is desired, the lead of the screw should be at least 1/3 of the screw diameter. Ideally the lead should be equal to the screw diameter. This calculated torque is the minimum amount of braking torque to hold the load in position.

Sounds like the machines you mention must have provided such a mechanism.

Best,

BW


Notice the sentence in the above quote-
"If backdriving is desired, the lead of the screw should be at least 1/3 of the screw diameter."
The mill conversion I linked to has a .200" lead which on a 1" diameter screw works out to be 1/5 of the screw diameter which means backdriving either isn't possible or is at a minimum unlikely at that diameter/lead combination.

The mills I have seen these installed on didn't have any type of brake or other drag device installed.The drag from the ways was enough by itself.

Backdriving would be an issue if the screws had some ridiculus lead like .750"/rev,but since it would be ridiculus to use a .750"/rev screw on a manual mill it's a non-issue anyway.

The point I am making here is,ballscrews are availible and are used on manual machines,it's even an option on some mills,it's just not particularly cheap.

wierdscience
07-16-2006, 02:15 PM
Are you basing this on actual experience, or just rehashing the ad copy ? And are you saying that you can climb mill with these ballscrews on a manual machine without using power feed, and without tightening down the gibs, and have no table "runaway" issues whatsoever ?

Yes actual experience,and yes you can climb cut without the feed,but as with any climb cut on a B-port you must be within reason on your cutter diameter,feedrate and Doc since the typical B-port frame is ridgid as a wet noodle in a climb cut anyway.

You could also drop a feed in gear with the motor off and use it as a drag brake,but I think for that to be needed the cut in question would be outside what the machine is capable of making safely.

Evan
07-16-2006, 02:35 PM
"If backdriving is desired, the lead of the screw should be at least 1/3 of the screw diameter."


Anybody care to calculate what helix angle that is? Hint: the sine is .1061

Although 1/5 is below that if enough force is applied some backdrive may be possible depending on the system friction and ballscrew type. A lead of 1/2 dia which is common with ballscrews ensures backdrive will happen.

Today with high precision thread rolling and polymer feed nuts ballscrews are being superceeded in small to medium load applications. Acme screws with polymer nuts have significant advantages over ball screws and also provide extremely low lash. That is one reason I am using them.

John Stevenson
07-16-2006, 02:54 PM
Although 1/5 is below that if enough force is applied some backdrive may be possible depending on the system friction and ballscrew type. A lead of 1/2 dia which is common with ballscrews ensures backdrive will happen.


Not on the type of machine, read home shop, we are talking about here.
Most are 2mm, 4mm, 5mm or 0.2" pitch

1/2", 10mm or 12mm is common on larger machines but we are talking larger diameters here.

My nephew's CNC router has a pitch of 50mm on about a 30mm screw but that's a different ballgame.

Hang on....................................

Just a mo ...............................

OK Just been out to the shop and tried to push the X axis on the little CNC mill.
5mm pitch screws on 14mm diameter, stepper on direct drive, no electrics connected, loose wires at motor.
Put a catalogue against my chest and bear hugged the mill and can't move it.
Probably need a bigger bear :D

OK not very scientific but it's a test.

.

Evan
07-16-2006, 03:00 PM
Disconnect the stepper and try it, I'm curious.

John Stevenson
07-16-2006, 03:03 PM
K, in a bit, dinner has just been thrown on the table and that damn dog is getting as big as me ................

.

Milacron of PM
07-16-2006, 03:31 PM
OK Just been out to the shop and tried to push the X axis on the little CNC mill.
5mm pitch screws on 14mm diameter, stepper on direct drive, no electrics connected, loose wires at motor.
Put a catalogue against my chest and bear hugged the mill and can't move it.


.

I could push the ballscrew equipped table of a 7,000 lb CNC EDM machine (sinker type) fairly easily by hand. Axis have to be secured when shipping ballscrew equipped CNC machines or the tables, turrets, etc move all over the place during trucking...and that's with the servos still attached. Of course those machines have linear ways as well, so they do move easier than a machine with dovetail or box ways.

John Stevenson
07-16-2006, 04:33 PM
I agree with Don on the linear ways and high lead screws, they do move.

Well took the stepper off and tried to move the bed.
Took a fair bit of effort to get moving but once moving it was easy.
This is on dovetail ways and it's still a bit tight from new.

Stepper on it's own is easy to move so it must just be borderline when the motor is on.

.

Evan
07-16-2006, 04:39 PM
Interesting and it matches the information. Your ballscrews are a hair over 1/3 dia lead and that happens to be just over six degrees helix angle.

lazlo
07-16-2006, 08:52 PM
They do mention using P4 grade ball screws and that standard apparently translates to 0.0006"/foot in any 300 mm section and 0.0008" over 35" of travel. Where does that grade fit on the p*ss-poor to fantastic scale?


That should be "C4", right? C4 ballscrew lead accuracy is around +/- 0.0006"/foot, but those are very, very expensive lead screws.

Tormach really has C4 ballscrews for $8,000???

Another possibility is that they're quoting a lead accuracy of +/- 0.0006"/inch, which is C2 accuracy, and
which would be a lot more cost effective.

Edit 1: Just saw Bob's answer: that's the spec of the angular contact bearings, not the ballscrew.
Edit 2: Bob, I just looked up the P4 DIN spec, and it's supposedly ABEC-7 (!)

ISO DIN AFBMA
class 4 P4 ABEC7
class 2 P2 ABEC9

Edit 3: Tormach's web page seems to confuse the angular contact bearing ratings with the ballscrew ratings themselves.
They talk about the Chinese factory that makes the Tormach ballscrews having a P4 ballscrew grinder, but the
DIN 620 and ABEC specs apply to ball bearings, not ballscrews.

http://www.tormach.com/quality_overview.htm

lazlo
07-16-2006, 08:59 PM
Well took the stepper off and tried to move the bed.
Took a fair bit of effort to get moving but once moving it was easy.

That's stiction, one of the downsides to box ways versus linear ways.

BobWarfield
07-16-2006, 09:22 PM
That should be "C4", right? C4 ballscrew lead accuracy is around +/- 0.0006"/foot, but those are very, very expensive lead screws.

Tormach really has C4 ballscrews for $8,000???

Another possibility is that they're quoting a lead accuracy of +/- 0.0006"/inch, which is C2 accuracy, and
which would be a lot more cost effective.

Edit 1: Just saw Bob's answer: that's the spec of the angular contact bearings, not the ballscrew.
Edit 2: Bob, I just looked up the P4 DIN spec, and it's supposedly ABEC-7 (!)

ISO DIN AFBMA
class 4 P4 ABEC7
class 2 P2 ABEC9

Edit 3: Tormach's web page seems to confuse the angular contact bearing ratings with the ballscrew ratings themselves.
They talk about the Chinese factory that makes the Tormach ballscrews having a P4 ballscrew grinder, but the
DIN 620 and ABEC specs apply to ball bearings, not ballscrews.

http://www.tormach.com/quality_overview.htm

Whoa there Lazlo, slow down, take a breath there fella. Let's keep the posts straight. I agree completely that P4 = ABEC7, never said otherwise. What I said was that the AC bearings they are using, courtesy of their exploded parts list, are P5's. In fact the exact quote from my post was:

"ABEC7202B/P5", hence they are P5 DIN = ABEC5

Indeed, this is sounding more and more like a CNCZone thread. Amazing how easy it is to go off down that track!

Best,

BW

MikeHenry
07-17-2006, 12:35 AM
Bob, John, and Lazlo,

Thanks for the replies on the Tormach questions. The bottom line seems to be that the bearings and ball screws are adequately specified for a hobby/prototype CNC mill, or at least that they shouldn't be a major limitation for that class of work.

My apologies if that mistates anyone's opinion.

Mike

lazlo
07-17-2006, 02:57 PM
Whoa there Lazlo, slow down, take a breath there fella. BW

We're talking past each other Bob.

Mike's original post/question stated (correctly) that Tormach indicates that they use P4 ballscrews. ABEC/DIN 620 applies to ball bearings, not
ballscrews, and if you read that Tormach link I posted, Tormach is actually claiming that they're using P4 angular contact bearings, not ballscrews.

Since P4 == ABEC-7, my point was that Tormach is actually claiming that they're using ABEC-7 bearings on their ballscrews.

I think this is all a moot point though -- since all their components (castings, ballscrews, bearings, ...) are made in a Chinese factory, they can stamp whatever accuracy grade they want on the parts, and we would never be able to tell.

Harbor Freight/Sieg has "UL Approved" stamped on a lot of their motors, but I would sincerely doubt they really went through UL Laboratory approval process :)

lazlo
07-17-2006, 03:04 PM
The bottom line seems to be that the bearings and ball screws are adequately specified for a hobby/prototype CNC mill

My synopsis is that Tormach is going for "Bang for the Buck" -- and whether or not they have ultra-spec bearings or screws, it looks like they did an outstanding job pulling this machine together.

Based on the deliriously happy new owners on CNCZone, it looks like the perfect machine at this price-point.

Get the rotary table too Mike -- I'd fly out to Illinois to see the Tormach running a 4-axis CAM program ;)

MikeHenry
07-19-2006, 02:38 AM
Get the rotary table too Mike -- I'd fly out to Illinois to see the Tormach running a 4-axis CAM program

Start thinking about ticket prices. An unexpected financial windfall has me planning on a fully tricked out Tormach and the 4th axis is on the list. Just need to sell off a few things to make room before ordering.

Millman
07-19-2006, 03:40 AM
There you go again guys; making things overcomplicated; without touching any milling table manually; how can you have this "Matter of Fact Views"? If it's not sitting in front of you, to manhandle it properly; how can you state exactly what is going on? If it's not 'Touchable', then it's all theory. And...you know what I think about theories.

lazlo
07-19-2006, 10:21 AM
There you go again guys; making things overcomplicated


Mike and I are both full-time engineers. We make things overcomplicated for a living :D


without touching any milling table manually; how can you have this "Matter of Fact Views"? If it's not sitting in front of you, to manhandle it properly; how can you state exactly what is going on?

Most definitely true. Unfortunately, Tormach has only shipped an handful of their CNC machines, and I don't think they have a showroom. That's why I want to go see Mike's machine :)