-
Scam Alert. Members are reminded to NOT send money to buy anything. Don't buy things remote and have it shipped - go get it yourself, pay in person, and take your equipment with you. Scammers have burned people on this forum. Urgency, secrecy, excuses, selling for friend, newish members, FUD, are RED FLAGS. A video conference call is not adequate assurance. Face to face interactions are required. Please report suspicions to the forum admins. Stay Safe - anyone can get scammed.
Weiss VM32 CNC Conversion
- Thread starter David
- Start date
slow-poke
Ultra Member
Suggestion, you might want to start with a cheap (non angular contact bearing in that location, yes you might have a few thousands displacement, then when everything is working install the good bearing to get rid of that slop. I don't see why the retainer would need to be so tight, you can generate a LOT of force with even mild torque. In theory the bearing shells should be touching and they should be able to take a lot of force, I torqued the retainer to 4Nm.
That's what I did when I was setting up my mill. I'm using cheap ball screws and I had about 0.002" initially from both the cheap bearing and the ball screw. It required some effort to get it under 0.001, however with a bit of tweaking final was 0.00039"
That's what I did when I was setting up my mill. I'm using cheap ball screws and I had about 0.002" initially from both the cheap bearing and the ball screw. It required some effort to get it under 0.001, however with a bit of tweaking final was 0.00039"
I just pulled it apart. Again.
I clamped a vise grip around the ball screw under the upper bracket where the ball nut will never reach so no worries about scratching the thread.
Used my dad's 14" long Crescent wrench and it took maybe 15-18 lbs of force to loosen the nut. There looks to be remnants of Loctite on the threads which helps explain why it was so hard to remove.
Here's a shot of the top of the bracket.
The bottom. The circular scratches were there when I got the parts.
Top of the bracket with the ball screw inserted.
Bottom with the ball screw inserted. The bearing is just there for illustration.
I'm a bit muddled in my thinking here. What is the bearing orientation that is required?
Me thinks that the top bearing should be like this so it can resist downward pull when the servo motor is pulling the head up.
And the bottom like this to resist upward force when the servo is pushing the head down.
I need help on this one. I think the top bearing may have been in upside down because what was left on the end of the ball screw was just the inner race.
I clamped a vise grip around the ball screw under the upper bracket where the ball nut will never reach so no worries about scratching the thread.
Used my dad's 14" long Crescent wrench and it took maybe 15-18 lbs of force to loosen the nut. There looks to be remnants of Loctite on the threads which helps explain why it was so hard to remove.
Here's a shot of the top of the bracket.
The bottom. The circular scratches were there when I got the parts.
Top of the bracket with the ball screw inserted.
Bottom with the ball screw inserted. The bearing is just there for illustration.
I'm a bit muddled in my thinking here. What is the bearing orientation that is required?
Me thinks that the top bearing should be like this so it can resist downward pull when the servo motor is pulling the head up.
And the bottom like this to resist upward force when the servo is pushing the head down.
I need help on this one. I think the top bearing may have been in upside down because what was left on the end of the ball screw was just the inner race.
Last edited:
So are the bearings in 2 pockets with a step between them? If so there needs to be a spacer bush on the shaft for the inner races to keep them from being overloaded. Its not the best arrangement, as the tolerances are usually really tight, hence the reason angular contact bearings are sold as matched sets, or universally matched sets that can be installed with the correct preload.
Agree with Slow-poke - replace with 6202-2RS for now, you'll have a small amount of backlash due to it but should be manageable. Really what you need to know is the size of that step so you can make a spacer bush to control the preload, probably something in the order of a few tenths to a thou less than the step size (thats a total WAG).
Agree with Slow-poke - replace with 6202-2RS for now, you'll have a small amount of backlash due to it but should be manageable. Really what you need to know is the size of that step so you can make a spacer bush to control the preload, probably something in the order of a few tenths to a thou less than the step size (thats a total WAG).
Yes there's a small step between the bearings (for the bloody life of me I could not think of how to describe that feature
...)Is my thinking correct on the direction of the axial forces?
(No idea why I'm struggling to picture this in my head)
They can be either back to back or face to face for bi-directional force which is what they will see in this application. This link has more than you ever wanted (or needed) to know about the bearing arrangements: SKF Bearing Preload
That said, with the ballscrew being a relatively long and bendy thing compared to say a lathe spindle, stiffness of the bearing arrangement to off axis loads is probably irrelevant so either way is probably "right".
But if you put them back to back then you won't get as many haters screaming at you online that you did it wrong.
That said, with the ballscrew being a relatively long and bendy thing compared to say a lathe spindle, stiffness of the bearing arrangement to off axis loads is probably irrelevant so either way is probably "right".
But if you put them back to back then you won't get as many haters screaming at you online that you did it wrong.
Great read. Reminds me of the old MGB wheel bearings. There the pre-load was done with shims. Put in the right amount of shim and then torqued the wheel nut really tight.
Nowadays we use crown nuts and use either a key'd washer or cotter pin to hold the nut at the preloaded point.
I used angular contact bearings for X and Y on my system and a spacer to provide the load from the nut.
slow-poke
Ultra Member
When I was shopping for angular contact bearings my cheap side provided a few lessons.
1) The Amazon special dirt cheap AC bearings are useless. How bad can they be? BAD!
2) Non matched pairs are no bueno.
3) Prices quoted from most of the bearing places can be very expensive > $300 per
4) There are cloned fakes (just like the Mitutoyo calipers), in NSK boxes that look legitimate but are not to specification, choose your source carefully.
5) I found a bearing shop in NewYork (eBay) that specializes in quality NOS bearings for affordable prices $50 pair, every pair they sent me was as per description and high quality.
All that being said, chasing tenth's is time consuming and expensive and if you have not closed the feedback loop with linear scales is probably a waste of time because the ball screw error will change with temperature anyways so your open loop numbers might look good but your actual won't be quite as good. The good news is most of us don't need to make parts to that accuracy, but it's fun trying.
1) The Amazon special dirt cheap AC bearings are useless. How bad can they be? BAD!
2) Non matched pairs are no bueno.
3) Prices quoted from most of the bearing places can be very expensive > $300 per
4) There are cloned fakes (just like the Mitutoyo calipers), in NSK boxes that look legitimate but are not to specification, choose your source carefully.
5) I found a bearing shop in NewYork (eBay) that specializes in quality NOS bearings for affordable prices $50 pair, every pair they sent me was as per description and high quality.
All that being said, chasing tenth's is time consuming and expensive and if you have not closed the feedback loop with linear scales is probably a waste of time because the ball screw error will change with temperature anyways so your open loop numbers might look good but your actual won't be quite as good. The good news is most of us don't need to make parts to that accuracy, but it's fun trying.
Last edited:
I'm going to talk to a local supplier about what they stock for matched 7202's. Also getting a pair of 6202's from them on Friday.
Given that I got such a deal on the whole kit I think spending a few dollars on replacement bearings makes sense.
Do you have a link to your eBay seller?
slow-poke
Ultra Member
NDH 20202DTT5CC Bearing 7202-C-T-DUL-P4 20202-DT-T5CC 15x35x11 mm | eBay
Find many great new & used options and get the best deals for NDH 20202DTT5CC Bearing 7202-C-T-DUL-P4 20202-DT-T5CC 15x35x11 mm at the best online prices at eBay! Free shipping for many products!
www.ebay.ca
Might be worth giving him a call 585-617-3500
Lot's to choose from, prices are up a bit.
Another option
NACHI 7202CYDU/GL ANGULAR CONTACT BEARING, P0291 | eBay
Find many great new & used options and get the best deals for NACHI 7202CYDU/GL ANGULAR CONTACT BEARING, P0291 at the best online prices at eBay! Free shipping for many products!
www.ebay.ca
Curious to here about prices from your local vendor, in Ottawa they were quite expensive.
Last edited:
@slow-poke
Just back from my supplier. Picked up the NSK 6202's. $26 for the pair.
The matched 7202's (FAG) are $94 each from my local supplier.
Just back from my supplier. Picked up the NSK 6202's. $26 for the pair.
The matched 7202's (FAG) are $94 each from my local supplier.
Last edited:
It is actually. The Nachi's from Bearings Canada start at $32ea on sale for C3 grade and then jump to $175ea for ABEC-5 and go up from there.
Don't spend money on a matched set until you have a plan for mounting them, your current setup won't make any use of a matched set.
I see a few ways forward.
1. Do your best to determine the space between the 2 bearings and make a spacer bush to fit, minus a tiny amount for preload. Give it a try with cheap, non-matched 7202's
2. Modify the current mount setup to use a pair of matched 7202's back to back, with a separate shaft seals (i.e. no seals on the bearings like your originals).
3. Use an off the shelf FK15 from a reputable supplier - comes with 2 angular contact bearings installed, just open up a large hole for it to drop into and bolt it down. May require spacing your servo motor out a bit) - note that some "FK15" are sold with regular bearings, avoid those. Get from a good source.
FK15 bearing mount
I'd be leaning to #3 these days, easiest to do, and results are expected to be good. Not the cheapest, but solves a bunch of issues
I see a few ways forward.
1. Do your best to determine the space between the 2 bearings and make a spacer bush to fit, minus a tiny amount for preload. Give it a try with cheap, non-matched 7202's
2. Modify the current mount setup to use a pair of matched 7202's back to back, with a separate shaft seals (i.e. no seals on the bearings like your originals).
3. Use an off the shelf FK15 from a reputable supplier - comes with 2 angular contact bearings installed, just open up a large hole for it to drop into and bolt it down. May require spacing your servo motor out a bit) - note that some "FK15" are sold with regular bearings, avoid those. Get from a good source.
FK15 bearing mount
I'd be leaning to #3 these days, easiest to do, and results are expected to be good. Not the cheapest, but solves a bunch of issues
Couldn't the bearings be taken out and flipped?
They could, but as suppliers of equipment designed for this application I'd trust them to have installed them that way on purpose.
No point messing with a quality part, and perhaps they are getting bearings that are matched for face to face preload only.
The SKF document I linked mentioned that back to back is stiffer for resisting bending. If side loads aren't what you care about, but absolute precision location is, then face to face will result in higher preload when thermal expansion occurs, versus back to back will loosen up with thermal expansion. I would not expect that either condition is a concern in your CNC (minimal thermal expansion and no side load like a spindle), so back to my original statement that I don't think it matters in this case.
Every time I look into this stuff (CNC design) I'm amazed at how much behind the scenes engineering is going into the off the shelf bits and pieces we take for granted these days, even if I don't understand it all.
No point messing with a quality part, and perhaps they are getting bearings that are matched for face to face preload only.
The SKF document I linked mentioned that back to back is stiffer for resisting bending. If side loads aren't what you care about, but absolute precision location is, then face to face will result in higher preload when thermal expansion occurs, versus back to back will loosen up with thermal expansion. I would not expect that either condition is a concern in your CNC (minimal thermal expansion and no side load like a spindle), so back to my original statement that I don't think it matters in this case.
Every time I look into this stuff (CNC design) I'm amazed at how much behind the scenes engineering is going into the off the shelf bits and pieces we take for granted these days, even if I don't understand it all.