My Ballscrew Conversion

I would estimate total cost was between $800 and $900. About $500 for the screws/nuts, $200 for bearings, and other bits and pieces here and there (seals, felt, raw material). The lowest price I got for a kit was $1800, and that used the stock bearing mounts which I wasn't psyched about.
 
That's a pretty neat solution. I looked at the McMaster-Carr site and those square ball nuts are a pretty good price ($51) compared to the other nuts there http://www.mcmaster.com/#ball-screws/=r90flj

I'm trying to understand how this goes together. It looks to me as if you drop in the left hand nut with the white Delrin ring screwed on its end and the Belleville washers in the middle. Then drop the other nut into the remaining space - it looks as if it is already screwed into the rectangular plate on the right. Then tighten the bolts on the plate on the right so the springs are compressed and the right hand nut bottoms out against the white ring. It means that the 200lb preload is carried by the threads in the Delrin ring (shouldn't be a problem if it's done right) and the right hand nut is constrained by the Delrin ring and the rectangular plate.

Looks as if you got the leadscrew with the premachined end. Was the whole bar hardened or just the thread surface? If I bought a piece of unmachined leadscrew, I'm worried I wouldn't be able to machine it down without going through a lot of carbide inserts.

The yoke was a worry to me and this looks easier than boring out a couple of cylindrical holes. Do you have any drawings for what you did or was it made to fit? The McMaster-Carr dimensions are very basic and not much to work from. Like you, I'm expecting to have to remove the table on a frequent basis, as I will be using the machine to make parts for itself! Mine's a Taiwanese BP clone.

Murray
 
The price of the nuts and the screw stock is the main reason I went with this. I figured it was low enough that if it didn't work out, I would have learned something, and not been out too much $$$.

The rectangular plate (the flange, as it is referred to) only attaches the assembly to the yoke. Please refer to this picture

20131020_194122.jpg

The flange is on the right side. The threads from the right-most ballnut are screwed into the flange, and a set screw in the flange locks the flange to the ballnut (there is some blue loc-tite in there for good measure). This nut is fixed to the yoke by 4 socket head cap screws that run through the flange, parallel to the ballscrew, and into the yoke.

The second ballnut (left-most) is free to float. The white delrin ring (the adjusting nut) is threaded onto this nut. During assembly this nut was loaded onto the ballscrew first, then the adjusting nut as screwed onto the threads from this ballnut as far as it would go. The disc springs followed (greased), then the second ballnut. The ballnuts are actually arranged "back to front". The tension on the springs is pressing on the back of the right-most ballnut (the fixed one) and the adjusting nut. Once assembled, the adjusting nut is tightened until the gap between the adjusting nut and the right-most ballnut is around 0.004" (the rated accuracy of the screws). This allows for expansion, but limits total movement in the case that we overcome the spring tension.

I purchased the raw ballscrew stock and machined it myself. I annealed the screw ends first with a torch, wrapping a wet towel around the section I wanted to leave hard. You don't want to get this to orange, you just want to temper it enough that you have a hope of cutting it (if you get it too hot, it will self-quench, and get harder than when you started...). After that it wasn't too bad. I made up some soft jaws for my lathe to hold onto it properly, and the rest was just basic turning, facing, threading. Oh, I also had to make a spider for my lathe, as the X axis screw is around 5 feet, and I had to run it through the headstock and support it on the other side.

I have some drawings of the yoke that I did based on my stock yoke. I'll see if I can find them when I get home, but I would recommend you check your own setup before you trust that my drawings will work for you ;)

My machine is also a Taiwanese BP clone, but as I have found, there are a lot of things in common, and a lot of things that differ. It's not always clear which is which.

20131020_194122.jpg
 
Here's a set of screws with zero backlash double ball nuts. They are C7 1605 and 2005 based on Jumps4 RF45 diameters. Supplier is LinearMotionBearings on Ebay. $189 shipped for this set.
Dave
tmp_20140325_171708-1284194254.jpg

tmp_20140325_171708-1284194254.jpg
 
I'm thinking of using a 2505 ballscrew (25mm dia, 5mm pitch), for all 3 axes, possibly with a double nut like dave2176 showed. My lathe headstock will take a 25mm shaft but not a 32mm (which is the diameter of the original BP leadscrews). Can't imagine I'd need a 32mm version anyway.

One issue with using McMaster-Carr is that they don't ship to Canada where I live. It is do-able but adds another level of hassle. I may stick to Plan A, ie take my chances with a Chinese supplier.

That ebay store is based in China and there are actually loads to choose from if you look at AliExpress, somewhere like this: http://bit.ly/1iyRFMA although dave's looks pretty good.

BTW, there is a whole series of "white papers" published by Tormach which are really interesting for people like us. I'd probably buy one of their machines if I had the dosh but in the meantime it's encouraging to see people like them explaining how to size and test the system components, cutting through a lot of the crap and using good engineering practice. Have a look: http://www.tormach.com/engineering.html

My Taiwanese BP clone has a whole hotchpotch of metric and imperial threads and dimensions. It's a metric machine with 5mm leadscrews but the threads elsewhere are a mixture of UNC, metric and possibly even some BS Whitworth threads. I've learned to measure them very carefully when I'm planning to replace any of them! If you have a DRO and/or CNC system it doesn't matter in the end if the leadscrew pitch is metric or imperial.

Murray
 
I'm thinking of using a 2505 ballscrew (25mm dia, 5mm pitch), for all 3 axes, possibly with a double nut like dave2176 showed. My lathe headstock will take a 25mm shaft but not a 32mm (which is the diameter of the original BP leadscrews). Can't imagine I'd need a 32mm version anyway.

One issue with using McMaster-Carr is that they don't ship to Canada where I live. It is do-able but adds another level of hassle. I may stick to Plan A, ie take my chances with a Chinese supplier.

That ebay store is based in China and there are actually loads to choose from if you look at AliExpress, somewhere like this: http://bit.ly/1iyRFMA although dave's looks pretty good.

BTW, there is a whole series of "white papers" published by Tormach which are really interesting for people like us. I'd probably buy one of their machines if I had the dosh but in the meantime it's encouraging to see people like them explaining how to size and test the system components, cutting through a lot of the crap and using good engineering practice. Have a look: http://www.tormach.com/engineering.html

My Taiwanese BP clone has a whole hotchpotch of metric and imperial threads and dimensions. It's a metric machine with 5mm leadscrews but the threads elsewhere are a mixture of UNC, metric and possibly even some BS Whitworth threads. I've learned to measure them very carefully when I'm planning to replace any of them! If you have a DRO and/or CNC system it doesn't matter in the end if the leadscrew pitch is metric or imperial.

Murray
You may want to consider using a different pitch on the Y, reason being is that there is more inertia, weight and momentum on the saddle vs. the X only moves the table....Tim
 
There's not a whole lot of choice when it comes to pitch at this diameter. It's either 5mm or 10mm, so I'd go for the 5mm, which is just under 1/4". Pitch translates into what torque you need from your motor for a given thrust (reaction + inertia) load on the table. I'm not after very fast feeds so the smaller motor / finer pitch / slower feeds is a better trade off for my needs (including cost) than the bigger motor / coarser pitch / faster feeds option.

Having had this machine in pieces before, I know that the 42" table is a lot heavier than the saddle, so I'm thinking that the difference in inertia won't be a big issue for me, even if it means I have to limit the Y feeds slightly lower than the x feeds. Something like 60ipm will be fine for me. When I worked out the loadings and torques, it seemed to me that a "true"(?) 650ozin would be pretty handy. If the worst came to the worst, I could always go back later and fit a larger motor to clear the bottle neck.

I'll let you know how I get on, although I'm always open to hearing of others' experiences and suggestions.

Murray
 
You will get the most drive force from the 5mm anyway. You max speed will be about 1/2 of the 10mm. Probably not a problem.
 
Looking at how you have to assemble the yoke into the saddle then fit the Y ballscrew from the front of the machine, did you have to carefully feed the screw in while removing a cardboard or plastic tube from the nut(s)? And do that the other way round each time you needed to withdraw the screw? I'm imagining the balls getting loose if you aren't careful. You must have done this a few times. Makes me nervous!

Doesn't look as if you can insert the screw and nut (already assembled together) from the front and then fasten them to the yoke, which I believe is possible with a single piece flanged double cylindrical nut. Even so, I think they install it with the flange at the back - which means they need an open yoke to make it possible to assemble (eg Hiwin).

Murray
 
The ballscrew and pre-loaded ballnuts can be removed wouthout dis-mounting the ballnuts. First the end supports are removed, then the table is removed, giving access to the yoke. 4 Socket head cap screws are then removed from the flange, and the whole assembly just lifts out. Easy peasy (well, relatively easy).

I did manage to unintentionally unload the ball-nuts several times ;). It's a real pain in the rear, and I recommend against it, but re-loading them is very doable. The hardest part is getting the ball-return tube back on. The trick is to pack the tube with grease, then fill the tube with balls so they stick in the grease and don't roll under the heaviest pieces of equipment in your shop, or into a nice pile of swarf. Unfortunately I never did find two of the balls, so one of the nuts is just shy of a full compliment. Still seems to work ok.
 
Back
Top