[peteeng]

Yep and that's one way too. Peter

[Nikolaguca]

I'm having some trouble with Z axis design. As you've changed my mind regarding rail and carriage orientation, I now have no idea where to find space to make carriages fixed and rails moving...

I understand that this design is more rigid since this way the bearing blocks stay at the closest point possible to the table.

Should I make X axis plate wider or maybe put carriages and rails on their side (90 degrees to the Z axis plate) like Datron did on their machines?

I plan on using 15mm/~3/4 steel plate as my X axis plate and maybe 12mm or 15mm plate as my Z. I would make a sheet metal cover for the spindle and thus make that Z plate much stiffer.

EDIT:

This is what I came up with for gantry risers. The red blocks are linear bearings. Everything is steel (20kg per side). Do you see any glaring mistakes or places where I can improve this design... I can't put any angled support between bearing block plate and vertical one because I don't have any space left. That plate is 30mm steel.

[peteeng]

Hi Niko - If you weld such a part it will warp so you will need to have it machined flat. Usually its welded then stress relieved then machined to spec. I suspect you can't do this... your saddle is very compact. Can you assemble it? You have used "square" bearings these can only be assembled from one side, be careful and consider the assemble sequence just not the geometry.... because you may not be able to adjust something or fit something. Plus if say a bearing becomes loose (which they may do when the machine is running in) you have to pull everything apart to get to the loose bolt.... so many rabbit holes. One solution for welding is to braze, or soft solder the parts together. No melted parent metal so the assembly stays very true to its set up position. Peter

https://www.cnczone.com/forums/diy-c...546-cnc-2.html may help with Z look at last post images

[Nikolaguca]

I thought about using thread loctite for all bolted connections to avoid such problems. Would that be enough?

I experimented with this Z axis design in Fusion... it would be as compact as ballscrew would allow and it would be very stiff because both X and Z plates would be C channels essentially. But it's hard to reach those Z axis carriages.

I have zero experience with router building so your help is essential during this phase.

I'm going to machine every surface that needs it. If I were to machine both X and Z plates, would I be able to somehow bolt those carriages down? I have around 4cm (1,1/2") of space between X and Z plate...

Other than that I think it's a stiff design... but everything is easy to assemble in CAD;

Thank you for your help!

[peteeng]

Hi Niko - If you assemble things as sub assemblies and check threads are clean and things bolt down correctly and stone or paper faying (touching) surfaces then pull it apart and reassemble with loctite it will be good. Use loctite on threads and on faying surfaces. Do you have access to heat treat? If you weld then machine, the part can still change shape. As an exercise you should put the bolts and nuts into the model then you can check they have top clearance for assemble. You may have to put holes in opposing parts so you can insert screws and drivers.

I don't think you gain anything with the channel design for the saddle, looks more complex then needed... getting rails parallel in two directions is tough mounted like you have. A flat saddle is better I think. Surprisingly it does not need to be very stiff. The Z axis plate (tool plate) does have to be very stiff....cheers Peter

[pippin88]

Re Z axis moving rails vs moving cars.

You need to think about tool length. More specifically getting your longest tool over your highest workpiece.

Example:
100mm thick workpiece.
50mm long tool.

If the spindle nose / tool can go above the gantry bottom then you only need 100mm gantry clearance (space between tabletop and gantry bottom). Moving cars makes it easy to get the tool above the gantry bottom - you can have a very tall Z axis no issues.

If the spindle nose cannot go above the bottom of the gantry (moving rails Z axis) then you need 150mm of gantry clearance to get the 50mm tool above the 100mm workpiece.

So you might gain stiffness at the Z and lose it at the gantry uprights.

(Numbers are just an example and not exact as you don't generally design to have the spindle nose able to touch the table, rather design for the shortest tool to touch the table)

(Essentially every design choice is a compromise and has pros and cons. There is no perfect choice.)

[Nikolaguca]

As I'm waiting for the main frame to get machined I decided to continue with the design of the Z axis.

I think I found the middle ground regarding Z carriage placement. What do you think? The spindle collet can clear the gantry bottom which is enough for me.

I can't decide if it's better to use steel or alu plates for X and Z. I used 25mm Alu plates here. Or should I go with 15mm steel plates? I don't have the time to do any analysis now regarding rigidity between the two.

I had to think about how to access all the bolts; I'll mill a shoulder on the backside of the spindle plate for rails to butt against. That way I'll be able to align Z carriages before mounting the X plate to the gantry. I'll mill a shoulder for the X axis blocks so I don't have to worry much about alignment.

[peteeng]

Hi Niko - The 25mm Al will be lighter and stiffer then the 15mm steel. But the webs on the Z plate could be steel to take advantage of its 3x stiffness. The al is half the weight and 1.62x stiffer. The webs will need several screws (as large as poss) to friction couple the web to the Z plate....cheers Peter

[Nikolaguca]

Hello guys, I'm making some progress on the project but I have a question for you.

I've mounted the ballscrew (2010) and when I put some force on the gantry side I get some movement along the Y Axis (you can ser my test setup here https://youtu.be/7OeQ09rGWKA)

For 150N of force I get around 2 to 3 hundredths of a millimeter (roughly 1 tou) movement.

Mind you, this is one side of the gantry, on the other side I have another ballscrew.

I've taken the end support bearings from consideration, so the entire backlash comes from the ball nut.

Is this 'acceptable' for wood/alu router?

As you can see on the video I've tried using 2 nuts to preload but without much success...

Sent from my SM-G970F using Tapatalk

[routalot]

Very little wood machining requires an accuracy of much greater than 0.1mm.What you don't yet know is what level of accuracy your machine can cut to because the flexing can be introduced from movement or distortion at any point from the spindle mounting backplate to the shank of a small tool.