you can straighten it but you need to take more measurements to see where that movement is coming from. it could be on the ends, bearing race not turned true etc. also you can use a floating joint to drive your machine. also you get what you pay for. try buying your components from reputable companies and pay 10 or 15 times as much
nikre This is a hardened rod, so using heat methods would affect the rod so I used hydrolic left and tried my best until I get it down to less than 0.05mm in total and it is hard to do it right because you need good controlled environment to do this.
I got 4 of these 1605's....all bent 0.20.... The moment you tighten everything the ballscrew vibrates a lot... I tried bending it back... It was slightly better... But far from good...
actually the precission class is not about the bending, but on the exact lead. lol. but it should be relatively neglectable. of course it depends on what you are doing. might also be a cartel thing, that the big four deliver lower grade stuff to the masses to keep their profit stable and led their customers shine. thats how one thinks in capitalism. "(w)holistic marketing", lol.
Hi, I just bought 3 big ballscrews from aliexpres and all 3 seems to have bends up to 0.3mm in middle of screw. Do you think I should open dispute and ask for some refund, and do you think its could be corrected by bending it back with some press to use it for high precision CNC. They are 2 DFU2505 1500mm and 1 DFU3205 2500mm, nuts have also huge backlash, but that may be because they are not tight enough because spacer between them is too thick. Thanks.
@@NourMuhammad I wanted to achieve 0.05mm precision on each axis with finished machine. Rigid steel construction and hiwin rails. Later it could be changed for better parts, but I need to know should I go into that project with this ballscrews or should I stop it and look for other solution. Thanks for your help man.
@@eugenega7494 I think it is achievable, but you need to get rid of the bent in the ball screws and also get yourself a backlash nut for each one of them, if have access to hydraulic press and a good setup you can remove most of the bent by trail and error and get them to -+10 micron
Just want to say: I ordered 3 from aliexpress: don't know whether they got bent in shipping or what but they were 1/2"+ out and 1 was 1/8". I'd wet my pants if they were only .002 out. In fact, I recently demolished my POS thom carr CNC router and the carriage had 1" of deflection and easily .75" in the gantry.
I think your test method is just not accurate enough, dial guage is good but the block in between is allowing for pressure changes from the needle to the screw, best way to show the accuracy of that screw is to measure peak to peak with the dial and the valleys if it allows, I suspect you actually have a straight screw there and the block is obscuring your results - I know this video is old but it also has potentially and unfairly called out products that could well be accurate, I have 3 from them that are very close to perfect, in fact better than a Hiwin one I bought for twice the cost, freight could be the issue with the Hiwin but that's impossible to tell.
Well, that is simply not true, we are not talking microns but 100s of them and the way the setup is done is simply to show the problem. The tension created by the dial gauge is enough to place the block under almost a constant pressure and to transfer the force without any major shift The way the block is placed is good enough to show the problem and not for taking accurate measurements or a repeated readings! Also this is a known issue and you can find others having the same problem. You just got lucky and if it is the case you should not compare what you saw with what you got!
@@NourMuhammad I'm not trying to argue with you, I just think a better test would be to eliminate the buffer block between it as its a clearly repeating error, if you ran directly to the bearing surface it would be more convincing for us watching your video, perhaps the video depicts the error accurately but it's a little inconclusive for us as the viewers and not the in person. Didn't you find it strange that the error was so repeatable with identical values at each revolution? I'm only trying to help, I've checked bearing surfaces myself and thought they were incorrect but then realised another influence was upsetting the results. Just a thought Muhammad 👍
@@wayneshephard It is not an argument, I am not trying to say this is the most accurate test and also there is no need for that, I am simply sure that this is not a reason for inaccuracy, I have repeated this test with other ball screws (4 of them) with different blocks and I was getting exactly the same results, I assure you that the block is not a source of any major errors in the readings, actually in the matter of fact, this is the way this should be done because you want to measure the over all bind in the rod not the constancy of the grooves by simply eliminate the inconsistent measurements you would get from the grooves. "Didn't you find it strange that the error was so repeatable with identical values at each revolution?" That's really very strange to ask such a question when the video title clearly says "Bent SFU1605 Ball Screw" what did you expect from a bent rod? A bent bar will do exactly this behavior man! Which really prove that the used method to measure the range of the bent radially is correct! Anyway, I came to conclusion after reading more about ball screws and saw several people having the same issue that a rolled ball screw will always have this issue to certain degree so I am not complaining about it anymore!
@@NourMuhammad ok, so you ended up using them, worked out ok in the end by the sound of it, I've got one that is a bit out and definitely didn't try to straighten it but I can certainly sympathise with the desire for accuracy in such items.
you might be right, but i would not generalize that. as long a certain quality is not guaranteed in a business contract, it is not part of the product, right? rather more in the head of someone, who insited that other would like to work for next to no margin.
@@erwinz5926 I usually buy cheap chiese parts for prototypes, then talk to suplier what i am going to do and my needs, they usually respect and just sent parts in my specs.
and. ohh. you get very standardised quality, especially in this terrain. linear rails for example, but also screwdrive screws, ballscrews.... that comes in 8 CLASS es. With subclasses. Showing the exact deviance. Yes this products are more expensive. though the high end class as rexroth serves and so... thats worse the extra penny, when you are totally in that chasing the micrometer scene....
@@lUnderdogl 👍thanks for the reminder. communication is very important. in communication and service, most alibaba firms at least claim being online 25 / 8.
you know it could habe been just bad bearings at the ends right or a deviation in the OD / ID at ehe ends? There are 1000 variables if you start counting bent screw being the less likely it is HARDENED STEEL you don't bend that
Regardless of the other 100000 variables, this is the system I have paid for that I supposedly going to use as is without any modifications!!! and I am expecting something to work not to suffer from! And BTW, I have tested the system in other conditions and always getting the same results! And also BTW, it is not only me that has this issue, youtube full of those bent ball screws! There is a stage that the ball screw should go through before finishing which is after Harding and Grinding is straightening!! and obivesly they didn't do it!
you can straighten it but you need to take more measurements to see where that movement is coming from. it could be on the ends, bearing race not turned true etc. also you can use a floating joint to drive your machine. also you get what you pay for. try buying your components from reputable companies and pay 10 or 15 times as much
have you fixed the bent mill? if so, how? same situation here
nikre
This is a hardened rod, so using heat methods would affect the rod so I used hydrolic left and tried my best until I get it down to less than 0.05mm in total and it is hard to do it right because you need good controlled environment to do this.
how much heat? can i do it at home?
nikre
No heat just use hydrolic left or hydrolic press and try to bent it in the reverse direction.
Search TH-cam about this process.
I have a question. When the ball screw is sold as
1000mm is that end to end or just the thread part ?
End to end usually.
i agree, usually end to end. some outline "travel distance" additionally.
I got 4 of these 1605's....all bent 0.20.... The moment you tighten everything the ballscrew vibrates a lot... I tried bending it back... It was slightly better... But far from good...
actually the precission class is not about the bending, but on the exact lead.
lol.
but it should be relatively neglectable. of course it depends on what you are doing.
might also be a cartel thing, that the big four deliver lower grade stuff to the masses to keep their profit stable and led their customers shine. thats how one thinks in capitalism. "(w)holistic marketing", lol.
Hi, I just bought 3 big ballscrews from aliexpres and all 3 seems to have bends up to 0.3mm in middle of screw. Do you think I should open dispute and ask for some refund, and do you think its could be corrected by bending it back with some press to use it for high precision CNC. They are 2 DFU2505 1500mm and 1 DFU3205 2500mm, nuts have also huge backlash, but that may be because they are not tight enough because spacer between them is too thick. Thanks.
You need a good and powerful hydraulic press to undo such a large diameter ball screw
@@NourMuhammad
So you think its not good for high precision?
Thanks.
@@eugenega7494
It really depend up on your definition of precision!
In most hobby applications it is good enough though!
@@NourMuhammad
I wanted to achieve 0.05mm precision on each axis with finished machine. Rigid steel construction and hiwin rails. Later it could be changed for better parts, but I need to know should I go into that project with this ballscrews or should I stop it and look for other solution. Thanks for your help man.
@@eugenega7494
I think it is achievable, but you need to get rid of the bent in the ball screws and also get yourself a backlash nut for each one of them, if have access to hydraulic press and a good setup you can remove most of the bent by trail and error and get them to -+10 micron
Just want to say: I ordered 3 from aliexpress: don't know whether they got bent in shipping or what but they were 1/2"+ out and 1 was 1/8".
I'd wet my pants if they were only .002 out. In fact, I recently demolished my POS thom carr CNC router and the carriage had 1" of deflection and easily .75" in the gantry.
Muskoka Mike
I am quite sure they were bent from the seller.
I have bought again from Banggood the same part and it came exactly bent the same way!
yeah, I'd be tickled pink with .002" out.....
fucking learn metric you primitive fuck
I think your test method is just not accurate enough, dial guage is good but the block in between is allowing for pressure changes from the needle to the screw, best way to show the accuracy of that screw is to measure peak to peak with the dial and the valleys if it allows, I suspect you actually have a straight screw there and the block is obscuring your results - I know this video is old but it also has potentially and unfairly called out products that could well be accurate, I have 3 from them that are very close to perfect, in fact better than a Hiwin one I bought for twice the cost, freight could be the issue with the Hiwin but that's impossible to tell.
Well, that is simply not true, we are not talking microns but 100s of them and the way the setup is done is simply to show the problem.
The tension created by the dial gauge is enough to place the block under almost a constant pressure and to transfer the force without any major shift
The way the block is placed is good enough to show the problem and not for taking accurate measurements or a repeated readings!
Also this is a known issue and you can find others having the same problem.
You just got lucky and if it is the case you should not compare what you saw with what you got!
@@NourMuhammad I'm not trying to argue with you, I just think a better test would be to eliminate the buffer block between it as its a clearly repeating error, if you ran directly to the bearing surface it would be more convincing for us watching your video, perhaps the video depicts the error accurately but it's a little inconclusive for us as the viewers and not the in person. Didn't you find it strange that the error was so repeatable with identical values at each revolution? I'm only trying to help, I've checked bearing surfaces myself and thought they were incorrect but then realised another influence was upsetting the results. Just a thought Muhammad 👍
@@wayneshephard
It is not an argument, I am not trying to say this is the most accurate test and also there is no need for that, I am simply sure that this is not a reason for inaccuracy, I have repeated this test with other ball screws (4 of them) with different blocks and I was getting exactly the same results, I assure you that the block is not a source of any major errors in the readings, actually in the matter of fact, this is the way this should be done because you want to measure the over all bind in the rod not the constancy of the grooves by simply eliminate the inconsistent measurements you would get from the grooves.
"Didn't you find it strange that the error was so repeatable with identical values at each revolution?"
That's really very strange to ask such a question when the video title clearly says "Bent SFU1605 Ball Screw" what did you expect from a bent rod?
A bent bar will do exactly this behavior man!
Which really prove that the used method to measure the range of the bent radially is correct!
Anyway, I came to conclusion after reading more about ball screws and saw several people having the same issue that a rolled ball screw will always have this issue to certain degree so I am not complaining about it anymore!
@@NourMuhammad ok, so you ended up using them, worked out ok in the end by the sound of it, I've got one that is a bit out and definitely didn't try to straighten it but I can certainly sympathise with the desire for accuracy in such items.
@@wayneshephard, I was able to straighten them to -+0.02mm by using simple DIY hydraulic press !
The reason why Made in China is cheap because quality control is not part in their production cost
you might be right, but i would not generalize that. as long a certain quality is not guaranteed in a business contract, it is not part of the product, right? rather more in the head of someone, who insited that other would like to work for next to no margin.
@@erwinz5926 I usually buy cheap chiese parts for prototypes, then talk to suplier what i am going to do and my needs, they usually respect and just sent parts in my specs.
and. ohh. you get very standardised quality, especially in this terrain. linear rails for example, but also screwdrive screws, ballscrews.... that comes in 8 CLASS es. With subclasses. Showing the exact deviance. Yes this products are more expensive. though the high end class as rexroth serves and so... thats worse the extra penny, when you are totally in that chasing the micrometer scene....
@@lUnderdogl 👍thanks for the reminder. communication is very important. in communication and service, most alibaba firms at least claim being online 25 / 8.
you know it could habe been just bad bearings at the ends right or a deviation in the OD / ID at ehe ends? There are 1000 variables if you start counting bent screw being the less likely it is HARDENED STEEL you don't bend that
Regardless of the other 100000 variables, this is the system I have paid for that I supposedly going to use as is without any modifications!!! and I am expecting something to work not to suffer from!
And BTW, I have tested the system in other conditions and always getting the same results!
And also BTW, it is not only me that has this issue, youtube full of those bent ball screws!
There is a stage that the ball screw should go through before finishing which is after Harding and Grinding is straightening!! and obivesly they didn't do it!