X axis spring back on vice squaring ?

[handlewanker]

Old Mac that sounds quite good, I could start using that in about 20 years, when I catch up to where you are at, you have been saying you had parts designed for more than a year, I guess we will just have to wait another year or two for the Designed parts to be approved, I believe you will have a birthday coming up when your baby cnc mill will turn ( 2 ), can we have some chips please

What ever happened to all the things you where going to build, and started, we would love to see the progress

Well, I strongly subscribe to the phrase....."say something if you have something to say, not because you have to say something"......the same applies to the mill....metaphorically speaking......I don't HAVE TO make chips just to show it can make chips...... it's already been proved by Skyfire on UTUBE that chip making is a function it was designed for.

On the subject of X Y backlash that has now reared it's ugly head for Ti5m......correct me if I'm wrong but I think I did see a double ball nut on the X axis screw when the table was off to reveal the dovetails......yep, just went back to post #15 and it does have a double ball nut.

Shock horror......we have backlash and tight spots at the ends of the screw????.....that tells me the ballscrew could be a well used one and a double ball nut will not cure the backlash.

Double ball nuts are designed purely to apply preload to offset heavy cutter forces.....and for no other function......under normal cutting loads and conditions they will wear the screw more than if a single nut was used.

I strongly suspect the backlash is only in the Y axis ballscrew.....but it could be a case of slack thrust bearings in either one etc.......a retrofit does open a can of worm so anything is possible.

So Mac.....with your self professed infinite wisdom........what would YOU do to cure this problem?..............I have a pet theory on how to DIY cure it, but I might be accused of heretical thinking, so the ball's in your court.
Ian.

[handlewanker]

Mac...getting off topic for a sec.....if you have ideas that occur from time to time it's best to file them for future consideration even if they've been described and discussed.....it doesn't mean they have to be acted on immediately......I have many irons in the fire to work on than the swag of ideas that occur with passing time.
Ian.

[ti5m]

Shock horror......we have backlash and tight spots at the ends of the screw????.....that tells me the ballscrew could be a well used one and a double ball nut will not cure the backlash.

They are actually brand new TBI screws and nuts. The tight spots are in the mill ways not the screw. The double nuts have been set to take up any play in the screws so should be good.

The thrust bearings are these https://www.damencnc.com/products/me..._308_1433_GB_1

[handlewanker]

??.....how did you set the double ballnuts to take up any play in the screws?

The only way you can yourself set ball nuts to eliminate backlash is by using 2 individual nuts flange to flange with a predetermined spacer between them.

Double ball nuts are bought already set at the factory for a preload.....single ball nut screws are normally constructed to be tight same as a ball race.

You would only want to have a double ball nut if you anticipated a heavy cutting load where a preload offsets the force applied against it.........this is the same as applying a preload to the spindle bearings to offset the loading from the cutter forces..........correct me if I'm wrong.
Ian.

[mactec54]

Thanks mactec54,

I just watched a lot of YouTube videos on how to use Fusion 360 and Mach3. Don't really know what i am doing lol.

I still have a bit of a backlash issue you might be able to see in the part on the middle of the curved sections there is a little flat spot. Its where the direction change happens in the Y and X axis. I cant really tune it out with Mach3 as i can get the backlash adjustment set perfect in one spot but then in a different spot it will be out. This is the same on all 3 axis. Good in one spot, and a bit out in others. I know my X axis has tight spots at the ends of travel but this part was machined in the middle of the table where its free so ?

Also my steps per is a bit weird. I can get it set perfect with two 1 2 3 blocks so a distance of 6". but then when i try to step in 0.1mm increments its way out. I have to step 12 times to get to 1.0mm which is strange as it measures the 6" distance spot on.

You will get it, takes a lot of study, when you are just starting from scratch, you converted your machine from manual to CNC, that is a big step in itself, ( that's a lot more than our friend handwanker has ever done) just try one small thing at a time, Mach3 is not good at doing Backlash comp, and most leave it off, and try to correct where the backlash is

Do all your steps per setup with the backlash comp off, this may be why you are seeing this 12 steps of 0.1mm for 1mm, also run you machine cutting a part with the Backlash comp off and see how the direction change is, this may be where the problem is

The Ballscrew Pitch could also be off and that would cause the same problem with the 12 steps

[ti5m]

??.....how did you set the double ballnuts to take up any play in the screws?

The only way you can yourself set ball nuts to eliminate backlash is by using 2 individual nuts flange to flange with a predetermined spacer between them.

Double ball nuts are bought already set at the factory for a preload.....single ball nut screws are normally constructed to be tight same as a ball race.

You would only want to have a double ball nut if you anticipated a heavy cutting load where a preload offsets the force applied against it.........this is the same as applying a preload to the spindle bearings to offset the loading from the cutter forces..........correct me if I'm wrong.
Ian.

The TBI double nuts are adjustable. Its like two nuts with a spacer between them and some adjustment screws that turn the second nut up against the spacer.

[ti5m]

You will get it, takes a lot of study, when you are just starting from scratch, you converted your machine from manual to CNC, that is a big step in itself, ( that's a lot more than our friend handwanker has ever done) just try one small thing at a time, Mach3 is not good at doing Backlash comp, and most leave it off, and try to correct where the backlash is

Do all your steps per setup with the backlash comp off, this may be why you are seeing this 12 steps of 0.1mm for 1mm, also run you machine cutting a part with the Backlash comp off and see how the direction change is, this may be where the problem is

The Ballscrew Pitch could also be off and that would cause the same problem with the 12 steps

Thanks mactec54,

Yes i think i did have backlash comp turned on when i set things up. I didnt realize Mach3 had problems with backlash comp.

[mactec54]

Thanks mactec54,

Yes i think i did have backlash comp turned on when i set things up. I didnt realize Mach3 had problems with backlash comp.

Did you buy Mach3 license or just using the Demo

[Fastest1]

Mach 3's backlash comp can work pretty good. Set the "look ahead" with a higher value. Removing physical backlash is always the best approach however it will never be perfect. Hence why there is bl comp ;-)

[handlewanker]

The TBI double nuts are adjustable. Its like two nuts with a spacer between them and some adjustment screws that turn the second nut up against the spacer.

Hi, that sounds good.......to be adjustable is the ideal......that would be the same as having 2 normal ball nuts with a spacer between them......my opinion is that with fixed adjustment, even with screws, it's only good for getting the fit backlash free on a good screw from new.

I have to wonder why with adjustable nuts there is still backlash.....that would to be be a back to the dealer as it doesn't do the job.

When you say there is a spacer between them and adjustment screws.....is that a hard spacer? ......IE made from steel..... or a resilient one made from whatever to allow the screws to apply pressure to compress it?

Just browsed the TBI site.....very interesting range of stuff.

How do their prices (for what you bought) compare with the EBAY range of ball screws etc.
Ian.

[ti5m]

Hi, that sounds good.......to be adjustable is the ideal......that would be the same as having 2 normal ball nuts with a spacer between them......my opinion is that with fixed adjustment, even with screws, it's only good for getting the fit backlash free on a good screw from new.

I have to wonder why with adjustable nuts there is still backlash.....that would to be be a back to the dealer as it doesn't do the job.

When you say there is a spacer between them and adjustment screws.....is that a hard spacer? ......IE made from steel..... or a resilient one made from whatever to allow the screws to apply pressure to compress it?

Just browsed the TBI site.....very interesting range of stuff.

How do their prices (for what you bought) compare with the EBAY range of ball screws etc.
Ian.

Its a hard spacer made of steel. Prices are a fair bit more than the Chinese eBay ballscrews. I got all 3 ballscrews with custom end machining and the 3 double nuts for about A$1632. I got some of the ebay ones from chai that everyone on here get for A$300 and they where total rubbish.

Did you buy Mach3 license or just using the Demo

Yes i bought a license for Mach3

Mach 3's backlash comp can work pretty good. Set the "look ahead" with a higher value. Removing physical backlash is always the best approach however it will never be perfect. Hence why there is bl comp ;-)

Yes i set my look ahead to 200. I put a dial test indicator on the floating end of my X axis ball screw and i measure about 0.005mm of movement there. So i guess my AC bearing holder on the fixed end has a little backlash in it. Not real sure how to do anything with it as it doesn't look like it can be adjusted. The information on it says "Dont Disassemble the unit it has been factory set for preload".

I might order a new gib for the Y axis too. If its not flat i can have a go at lapping it on a flat surface. I wont touch the original gib.

[handlewanker]

Wow....5 X the difference......no wonder the EBAY stuff is so popular.
Ian.

[mactec54]

Yes i bought a license for Mach3

Yes i set my look ahead to 200. I put a dial test indicator on the floating end of my X axis ball screw and i measure about 0.005mm of movement there. So i guess my AC bearing holder on the fixed end has a little backlash in it. Not real sure how to do anything with it as it doesn't look like it can be adjusted. The information on it says "Dont Disassemble the unit it has been factory set for preload".

Yes 200 is the norm for look ahead, the Backlash comp is what it is, and has never worked very well, there is one zone person that has had it working better than normal, he has lots of videos look for Hoss machine

I was just hoping that you had not brought the license yet, as UNCNC have a better control software now, Mach3 is ok but the things that don't work quite right will never be fixed

Do you have a photo of the thrust bearing block, depending on the Bearings they used, they most likely have a some spacers that are used to create the preload on the bearings, again this is something that you could lap if they have spacers, it's a small amount, and may not affect the machining unless the Gib is to tight, which will cause extra loading on the thrust Bearings

Some of the time with these Thrust Bearing housings, it is not the Preloaded Bearings that is the problem, but the housing end cap is not clamping the Bearings, together in the housing tight enough, this can be fixed with a shim, or machining the end cap

[ti5m]

This is what the bearing blocks look like


I downloaded a CAD file for it and looked at all the parts inside it but there are no shims in it so not sure how they are getting the preload. They are SYK brand FK15. Made in Taiwan supposedly. They look well made.

[ninefinger]

To know if its your support bearings contributing to the backlash, use your dial indicator on the end of the screw. Now be aware that the end of the screw may not be perfectly flat and concentric, but if you turn the screw exactly 1 full rotation one way and then exactly 1 full rotaion back your indicator should fall back to 0 (i.e it should be at the same place on the screw), if not then you have play in the support bearing. (i.e. the screw should not be moving relative to the table on X axis or relative to the base on Y axis).
As far as a quality support bearings, the bearings are a matched set, ground with a certain amount of preload built in. Check out the SKF site Angular contact ball bearings for details on back to back angular contact bearings. A more crude method is to use shims to create the preload but its much more difficult to achieve a good result as shims are often not available in the correct thickness required so a "close enough" shim is used, resulting in too little or too much preload.

I personally had much difficulty with with my cheap (from Chai) end bearings - they had no preload and had a huge amount of backlash that I eventually removed through shims, but not without difficulty.
Those you have claim to be superior so the dial indicator test should be good
Mike

[mactec54]

This is what the bearing blocks look like


I downloaded a CAD file for it and looked at all the parts inside it but there are no shims in it so not sure how they are getting the preload. They are SYK brand FK15. Made in Taiwan supposedly. They look well made.

Most of the Made in Taiwan is good quality

The 4 SHCS on the front clamp the outside of the Bearings into the Housing, the nut on the end of the Ballscrew clamps the 2 inner Bearing races together, this is all there is to it very simple, so something to check is where the 2 parts come apart there may be a very small gap between the 2 faces, when I make them, I give them .05 to .1mm, if there is no gap then this could be where the problem is, they can be made with no gap, but all the part faces need to be Ground to achieve this, and most likely these are just turned or milled

You could try to tighten those 4 SHCS and the Ballscrew locking Nut, if the bearings are a matched set, you can not over load the Bearings by doing this, as they will already have a set Ground Preload, the preload is done by grinding the faces of the Bearings to get the correct Preload

When I make up thrust Bearings sets, I Grind the Bearing faces the same way, as the factory's do them, this way I can use any mismatched Bearings and turn them into a preloaded pair

[ti5m]

Most of the Made in Taiwan is good quality

The 4 SHCS on the front clamp the outside of the Bearings into the Housing, the nut on the end of the Ballscrew clamps the 2 inner Bearing races together, this is all there is to it very simple, so something to check is where the 2 parts come apart there may be a very small gap between the 2 faces, when I make them, I give them .05 to .1mm, if there is no gap then this could be where the problem is, they can be made with no gap, but all the part faces need to be Ground to achieve this, and most likely these are just turned or milled

You could try to tighten those 4 SHCS and the Ballscrew locking Nut, if the bearings are a matched set, you can not over load the Bearings by doing this, as they will already have a set Ground Preload, the preload is done by grinding the faces of the Bearings to get the correct Preload

When I make up thrust Bearings sets, I Grind the Bearing faces the same way, as the factory's do them, this way I can use any mismatched Bearings and turn them into a preloaded pair

I just had a look at mine and there dosent appear to be any gap between the two parts of the bearing housing. I checked the tightness of the 4 screws and there was no movement left in them. I have the locking nut on the ballscrew done up quite tight also. The bearings are arranged face-to-face. And yes they are a matched set. Advertised as C3 quality whatever that might mean ? Website for them https://www.damencnc.com/products/me..._308_1433_GB_1

[mactec54]

I just had a look at mine and there dosent appear to be any gap between the two parts of the bearing housing. I checked the tightness of the 4 screws and there was no movement left in them. I have the locking nut on the ballscrew done up quite tight also. The bearings are arranged face-to-face. And yes they are a matched set. Advertised as C3 quality whatever that might mean ? Website for them https://www.damencnc.com/products/me..._308_1433_GB_1

C3 grade bearings are the cheapest and are just a general purpose grade Bearing, but when it come's to AC Bearings to do job like this, they work ok and they can be adjusted to remove any backlash, they won't be matched, unless they have had the faces ground, which is what I do, to preload these cheaper bearings

If you can't tighten the Ballscrew Nut to eliminate the Backlash, then You will most likely have to use a shim, in front of the outside (Front Face ) of the Bearing, this will be where that front plate clamps the Bearing into the housing, if this does not work then you would have to shim in between the ( 2 ) Bearings on the inside, and only where the ( 2 ) inner races clamp together, this is how you would do it for your Bearings being Face to Face

The write up on there web site where you brought these is total BS, C7 through C9 are the highest quality Bearings, this high quality come's with a price, that is why they are using C3

This is where you would need to shim, I would be going more for the inside shim as to where it needs it

[ti5m]

Cheers thanks mactec54,

I just checked my ballscrew again for any end-float turning it by hand this time and i couldn't measure any so the bearings must be good. The end of the ballscrew was not quite flat so as the screw turns it has a 0.01mm high spot that was throwing my measurement off i guess. Turning it by hand i get repeatable 0 turning it one full revolution left and right it always falls back to 0 at the same point of the end of the ballscrew so thats good.

I always thought the lower the C rating on bearings was a tighter tolerance ? I dont think there is any C ratings above C5 is there ? Unless you mean ABEC ratings ?

Here is some info on bearing C ratings and what they are.
Bearing internal clearance

I dont know where Damoncnc gets there info from as it seems incorrect doesn't it. I think they might be confused with ballscrew C ratings where a C7 ballscrew is a lower grade than a C5 or a C3. And the Chinese ebay sellers listing there bearings as C7 which is wrong.

[mactec54]

Cheers thanks mactec54, I shall look into shimming the bearings. Maybe have to make some out of some alfoil.

I always thought the lower the C rating on bearings was a tighter tolerance ? I dont think there is any C ratings above C5 is there ? Unless you mean ABEC ratings ?

Here is some info on bearing C ratings and what they are.
Bearing internal clearance


I dont know where Damoncnc gets there info from as it seems incorrect doesn't it. I think they might be confused with ballscrew C ratings where a C7 ballscrew is a lower grade than a C5 or a C3. And the Chinese ebay sellers listing there bearings as C7 which is wrong.

Steel shim would be best, you should be able to get this from a local industrial supply, sometimes Bearing suppliers will also have shim they may even have some already the right size for your bearings, .05mm should be enough

Yes C Ratings have always been used for Ballscrews, and correct the smaller the number the better the Ballscrew Pitch Lead error is

Yes correct ABEC ratings, but C ratings are a little confusing in the SKF page as well, the bearing ratings are not only about the radial clearances as for AC Bearings this is adjustable, no matter what the grade is, the ratings are about the whole Bearing

Bearings rated with the ABEC rating system typically have five ratings in the class scale that range from the widest tolerances to the tightest tolerances of the bearings with rating numbers of 1, 3, 5, 7, and 9. Each number corresponds to a degree of roundness in the outer and inner races of a bearing. The higher the rating number of bearings, the tighter the component parts for tolerance, meaning that a higher ABEC grade is assigned to a ball bearing manufactured against a higher standard of precision. For instance, a ball bearing rated at a grade of 9 provides the highest precision and efficiency.