Understanding the bearing block

[Jonathan]

A pair of, or a single double row angular contact bearing, is generally used to resist axial forces and eliminate end float. I think the problem with thrust bearings is they have to be mounted accurately - each side of the block they are mounted on must be parallel otherwise they will periodically bind. Angular contact bearings probably have a lower coefficient of rolling resistance, though I've not checked.

[Ricardoco]

A pair of, or a single double row angular contact bearing, is generally used to resist axial forces and eliminate end float. I think the problem with thrust bearings is they have to be mounted accurately - each side of the block they are mounted on must be parallel otherwise they will periodically bind. Angular contact bearings probably have a lower coefficient of rolling resistance, though I've not checked.

Thats the word.. Axial... i just couldnt think what it was at 4 in the morning lol..

As for type of bearing I think Jonathan has hit the nail on the head. from a self fabrication point of view, I found that angular contact bearings and the housings were easier to use.

[Rogue]

As for type of bearing I think Jonathan has hit the nail on the head. from a self fabrication point of view, I found that angular contact bearings and the housings were easier to use.

If my current thinking is on-track then I'm hoping the basic homebrew block would look something like this:

[Jonathan]

No, not like that. Move the 'wall' in the middle of the block to the edge, and have one plate on the other side to press the bearings against that. Have a look at the BK/BF blocks on eBay/Zapp if you're not sure...

The way you've drawn it makes it harder to machine concentrically. The standard way only requires one hole to be bored.

[Rogue]

...this would appear to be the cheapest way to avoid the latteral load from the axis being transmitted into the Motor bearing. the fact that the drawing above does not show a ballscrew is a little missleading

...

The radial bearing is not fixed to the drill rod but just a nice tight fit as a support for it. as you will know the shaft couplers are fixed to the rods they couple together

...

there you go... Clear as mud

Actually yes, that does make it a lot clearer! I was originally thinking in terms of a load perpendicular to the shaft which was probably what threw me. I'm currently considering starting with Acme thread which is why I was looking at that particular diagram.

I was planning to just use the axis rather than introduce additional drill rod unless there is a good reason not to do so?

A pair of, or a single double row angular contact bearing, is generally used.

So one double row ACB (such as this: http://www.worldofcnc.com/products.asp?recnumber=147), or two back to back single row ACBs (such as these: http://www.worldofcnc.com/products.asp?recnumber=73) mounted in a block would perform that same function as the combination of elements discussed above?

In the original block idea I posted (not the one from the website, mine) using two ACBs, would this mean that the load is being transferred into the coverplates that are holding the ACBs in place? And as they are pressing against the plates from the inside heading out, it would be the bolts holding the plates that are actually taking the strain? Number 1. in the picture below.



If that's the case, presumably Number 2 would be a better option as the bearing block then takes the strain from the bearings.

On a side note, I'm posting links to particular products that I'm looking at for the sake of clarity and safety - ie in case someone says "that bit doesn't do what you think it does, don't buy it" or "that really is too cheap to be worth using". It's not with any intention of advertising. I hope it's ok to do this

[Jonathan]

The bearings should be like in 1. i.e pressing against the plates. Fix the bearing(s) to the end of the screw by having a thread and a nut on the screw to lock it in place.

Those bearings you linked to are fine - but expensive! There's plenty on eBay...just search for the bearing number.

[Rogue]

Jonathan, thanks for the quick reply, I need to learn to type faster!

I'm surprised that No.1 would be the better option. May I ask why that is?

If the screw has been turned down by that stage then would a shaft collar be sufficient for keeping it in place?

Edited to add: and again with the fast reply! The single hole on one side was the original design idea, just like the other blocks I'd seen. The reason for trying to a different approach was because I thought they were using standard radial bearing in the block and thrust bearings on either side, and using the angular bearings would benefit from a different approach. Clearly not!

[Jonathan]

If the screw has been turned down by that stage then would a shaft collar be sufficient for keeping it in place?

Possibly, but it's not ideal. The problem with a collar is, similar to the thrust bearings, it may be skewed on the shaft. A collar that clamps onto the shaft (ring with slit, bolt through slit to squash it) would be better. I would stick with having a thread. It depends what you're doing with the machine. If you're only cutting PCBs then pretty much anything will work ...

[Rogue]

"It depends what you're doing with the machine..."

I think, like most people, I want something that is accurate and repeatable to within 0.000001" and can easily and cleanly mill everything from paper to adamanitum, while costing <£100 to build

PCB milling and aluminium/acrylic sheets for panel making is high on the list. The X/Y work envelope of 10"-12" is based on this. Small aluminium pieces (by small I mean 1"x1"x1") would be very useful.

I would really like to work on larger aluminium pieces, primarily in order to make things to improve the mill/build the next, but also for other occasional projects. For this reason I'd like 5"-7" inch Z axis, though I appreciate this is quite significant for a small DIY mill. I might end up sacrificing this functionality but it would be nice to have it.

I'm working on the basis that if I can build a solid enough design then I can work on a spindle motor solution (swappable mounts? pulley system? variable speed motor?) that will let me cover the range of work. I'm also not too concerned with the overall process speed - this is for personal projects. Fine if someone else can do the job in an hour start to finish. If I'm spending 30 minutes for each cut and the whole job takes a week of evenings, such is life.

Yes, I would like to buy once/cry once in terms of ballscrews/nuts etc, but I've weighed it up carefully and it's not within my reach at the moment.