operation cnc

[Jonathan]

But I am a realist as much as I could probably make one, if I set my mind to it. The other 7 just wouldnt be the 'same'

I very nearly captioned that photo with 'can you spot the mistake'!

. :-D Just looking at them, their accuracy makes me feel sick. Its one of these situations where I need to see them before I know what Im doing

They're a copy of a common design on eBay. I started with 2" square aluminium bar, used the lathe to face both sides and bore the hole in the middle. Next I milled two opposite sides to the correct size using a fly-cutter (hence the nice finish). I then spotted and drilled the holes using the milling machine by just moving to the correct co-ordinates for each hole and finished off by milling the other 2 sides and the corners. The milling was done using CNC and g-code written then saved as I went along, but it could just as easily be done manually. I could have started with aluminium bar closer to the final size to save some milling, but since my lathe has a 4-jaw self centering chuck it's more convenient to use that than get a load of rectangles on centre with the 4 independent jaw chuck.

... For example you have mentioned seals, I've heard circlips mentioned.

The BK block has a rubber shaft seal on both sides to help prevent swarf etc getting into the angular contact bearings as that would rapidly destroy them. There is a groove for a circlip on the other end of the ballscrew, which doesn't seem to achieve much.

An example, if I need 1500mm of drivable screw. (inside the supports) then how do I calculate how much 'extra' I need at either end to go through the housing (your suggesting I do not buy from the same supplier) ... easier to just buy the supported bearing housings from the china supplier and if necessary re-make them based on the design on the ones that come with the screw?

I'm suggesting you do, but it's not compulsory by any means! Clearly you can get the standard mounts to work, it just may require adding shims between the bearings, machining a spacer so that the nut doesn't rub on the shaft seal and possibly opening the bearing bore out a little so that the bearings are not too tight a fit. Or they might be fine. To me that doesn't seem worth it when I can just make my own and know it's going to be accurate from the start, in addition to gaining flexibility with the design and learning something the process. You will need to have a look in other threads about how to make them - the drawings are readily available.

On previous threads on here I noticed your original build thread for your first machine. This machine was significantly different to your present one, and I guess you learnt from your mistakes on that build

Short answer is yes. My initial machine was constrained significantly by the cost of getting the proper components, hence I tried to make as much as I could to save money. Everything between the X-rails (so Y and Z-axis essentially) hasn't changed since the first build. I just removed the silly tall gantry sides and made a better frame. I'm glad I did it this way since it forced me to try new things, such as using a bearing riding on a threaded rod to make a cheap drive with negligible backlash.

You say i should upgrade my X axis rails, they are already SBR20... what is an upgrade to them?

SBR25, or minimum 15mm profile rails (linear guides). Latter is a bit too expensive though.

Get the Y and Z ballscrews with the standard end-machining since they are quite short, so critical speed is not a problem, but do ask Chai to increase the length of the cylindrical portion on the fixed end of the screw (dimension F on his drawing) from 15mm to 25mm to accommodate the pulley.

Since the X-axis ballscrew is quite long, to get the best feedrates you can put a pair of angular contact bearings on both ends, so you could get the ballscrew machined the same on both ends.

Here are the standard end-machining drawings from Chai:



You can see from those how much 'extra' length you need for each ballscrew. If you ask Chai for RM1610-1000mm for instance, he will machine it such that the total length is 1000mm and the actual ballscrew is 924mm with F increased to 25mm.

The entire point of me building a machine is to be able to get to a point where I can make 'things' more accurate, Its a chicken and egg situation. If I had a rig, this rig build would be as easy as you make it sound!

The thing is you have got the machine to make this machine. When I decided to make my CNC router I had a milling machine and mini lathe, so I used those to make all the parts, such as bearing blocks and ballnut mounts and thus have never had to pay anyone to make bits for me. I did mill the Z-axis plate square on the school Bridgeport milling machine, but that was mainly just for aesthetics.