DIY mill...

[Lee Roberts]

Well the design progresses slowly...

Been thinking about ways to mount my leadscrews and looking at bearings for the same, namely the 608ZZ that Lee sells...

If its any help to you mate i have got a few of the 608zz's with the zz's missing, there is nothing wrong with the bearings other then the zz missing on one side. I get the odd one or 2 now and again like that so i just use them for testing out ideas as i wouldnt sell them on like that.

I'v also got x4 SK12's sat in my office as well, but i'v just had a look and they wont be big enuff for the 608zz's. i have got lots of SK20's so you could bore them out to 22mm for the 608zz's to fit into?

This is how i went about supporting my leadscrew:






So yea i fitted a bearing at each end of the leadscrews, then i used nuts to preload? the rod a little.

I'v now made this as another solution also in the hope people will find them usefull:

This is a bearing rod support i have designed for other builders and myself, use's a small grubscrew to hold the bearing in place. A key feature of the design I believe to be an advantage over some of the mass produced supports is that it is flush fitting to your machines; this means you don’t need to bore big holes to fit it.

If you want some SK20's buy them from my ebay shop and let me know your ebay ID and ill send you the bearings for free. Also wait for me to send you an invoice as i can more then likly do you somthing extra on the invoice as well :).

Cheers,
Lee

[Robin Hewitt]

What is the lead screw made of? There is a big difference between machining chrome moly steel and machining studding.

Best mount is one axial thrust race at either end and one powerful spring holding it in tension. This is very reassuring when you eventually chuck up your DTI, measure the backlash and have a panic attack. (Not a good idea if working in MDF though, it's about as resilient as marzipan when under load).

[irving2008]

If its any help to you mate i have got a few of the 608zz's with the zz's missing, there is nothing wrong with the bearings other then the zz missing on one side. I get the odd one or 2 now and again like that so i just use them for testing out ideas as i wouldnt sell them on like that.

I'v also got x4 SK12's sat in my office as well, but i'v just had a look and they wont be big enuff for the 608zz's. i have got lots of SK20's so you could bore them out to 22mm for the 608zz's to fit into?

[...]

I'v now made this as another solution also in the hope people will find them usefull:

If you want some SK20's buy them from my ebay shop and let me know your ebay ID and ill send you the bearings for free. Also wait for me to send you an invoice as i can more then likly do you somthing extra on the invoice as well :).

Cheers,
Lee

Lee, thanks for that, I may need some more 12mm round rail too so I'll be back to you when I've worked out exactly what I am going to do.

I was thinking that SHF20's bored out would also work as per your solution above (is that Delrin?)

I note your leadscrew runs the entire length of the axis. Although it doesnt need to - asuming the connection with the table is in the middle of the table then the leadscrew only needs to be 1/2 the table length plus supporting/machining allowance. This does mean that the end bearing in under the table and not at the end like yours. Apart from the more complex mechanical arrangement is there any other reason not to do this? I'd rather not buy any more 10mm x 2 leadscrew.

Robin - the leadscrew is whatever MarchantDice sell - carbon steel I think.

I'll admit I got a bit carried away and bought their package of 3 lengths - 400mm(X), 350mm(Y), 250mm(Z) - and 3 Delrin nuts before I really thought things through :), and I know know I could have got a better deal from Lee (may still do so), however this is my thinking so far:

I got some 16mm x 500mm rails for the table X traverse. decided to go for a fixed gantry, moving X table as I reckon I can make this more easily rigid than a full moving gantry. With 500mm rails I can get a full traverse on a 230mm table.

The 350mm leadscrew will give me about 200mm usable table width in the Y (assuming the motor mount is ~100mm wide) or I could use the 400mm here to give me 250mm usable table width (in which case I would buy another 600mm of leadscrew at £6 approx). I have some 12mm x 500mm rails which would need to be shortened as appropriate to "Travel + 125mm", i.e. 325mm or 375mm, the remaining rail being used for the Z i.e. 150 or 100mm of travel allowing for mounting.

The remaining leadscrew would give me up to 170mm approx Z travel but I will be limited by the rail length as above.

Here is a quick sketch of the base and X axis. haven't yet decided what to make this of yet... mdf or Ali...

[Robin Hewitt]

Can you re-arrange and centre support those rails? Would make it lots stronger.

Have you considered reinforced concrete for the base? A bit of granite worktop would be nice. Sand and liquid epoxy if you have the pioneering spirit.

Looks a tadge flimsy as is.

[irving2008]

Robin, how would you support a round rail with a linear bearing that fully encloses the rail?

these are 16mm carbon steel rails supported at each end - according to my calcs a single rail will deflect no more than 0.04mm under a 35kg central point load and a) I dont plan to load them that much and b) the load is distributed across two pillow blocks on two rails so expect the deflection to be <0.01mm...

Where's the flimsy bit come in?

Irving...

[Robin Hewitt]

Robin, how would you support a round rail with a linear bearing that fully encloses the rail?

Where's the flimsy bit come in?

It's not so much the load, it's sympathetic vibration that may give you headaches.

[irving2008]

It's not so much the load, it's sympathetic vibration that may give you headaches.

Robin,

Its an interesting point. a centre support will reduce the deflection but as we already know thats tiny. But I hadn't considered the resonance aspect. I'll do some calcs to find the resonant freq of the bar and see if theres any issue.

Irving...

[irving2008]

Robin,

Its an interesting point. a centre support will reduce the deflection but as we already know thats tiny. But I hadn't considered the resonance aspect. I'll do some calcs to find the resonant freq of the bar and see if theres any issue.

Irving...

Well the natural, undamped frequency for a single rail is roughly 287Hz so I'm not sure theres any issue here - all the possible drivers forcing oscillation are at at a considerably lower frequency (min cutter speed 2800rpm - 46rps) so the damping coefficients for forced excitation are large. Also with the two pillow blocks attached there is a significant damping coefficient. There may just be issues with a combination of cutter speed at 17000 rpm (single flute), 8500 rpm (2 flute) or (4250rpm (4 flute) and/or stepping rate of 172mm/min. (Its years since I did any of this stuff at college... makes my brain hurt just reading the equations!)

Of course the proof will be in the actual result...