Hi, I have been lurking here for quite a while and am now thinking of taking the plunge (!) and building some form of cnc router. Before I dive in (and probably get it all wrong) I would welcome comments on what I am thinking of doing.

First though, some background.....
I have 3 3d printers, one of which I built myself from scratch so think I understand the principles of CNC. I also have a dilapidated single garage as a workshop which is so full of stuff I can hardly move around but there is a lathe, pillar drill, belt sander and a fair selection of hand tools in there. Space though is a problem!
I also have too many things to do at the moment so, if it is to have any chance of success, I need to keep any build as simple as possible!

To that end I set a basic spec for a machine:
- 800x500x500 machine dimensions with the biggest practical work area. Anything over 200 x 300 workable area would be acceptable though.
- workpieces likely to be more 2.5d rather than 3d, so a small (maybe 100mm?) Z is probably ok. I suspect allowing for tooling and workholding will impact the z height more than the workpiece itself.
- capable of cutting aluminium and brass. Steel would be a bonus but I have no illusions that that will be sketchy at best with what I am likely to make.
- cheap! Target cost is around £500 although I am fully aware it is likely to go massively over budget by the time I have made it do what I really want.

Also, I might upset a few people here with this but, based on my experience with 3d printers, I would rather start cheap and "just about ok" knowing I am likely to upgrade later than go straight in with an over-engineered solution that will do anything. I know this sounds strange but I somehow enjoy the challenge of making things work and finding out what are really the critical things that need to be done right. As an example, I started out with an Anet A8 3d printer. It was only by getting that working as best it could (and having bought an ender3 and also fixed that) I reckoned I knew how to make my own design. I still have the A8 though and it still prints as well as anything else (if a bit slowly).

Anyway, to the point.....

Given the small size and need to keep it simple and rigid I am thinking of going with a fixed gantry design. Sensible?

For rails I am thinking of going with SBR supported rods. Two reasons: 1) cost. 2) tolerance to mounting errors. My custom 3d printer was made using 2040 extrusion with linear rails and I was surprised how difficult it was to make the mounting surfaces coplanar. Even slightly distorted extrusion meant I had to mill then flat to stop binding.
Given how I am thinking of making the frame (next point) that puts me off linear rails and I certainly don't think unsupported rods will work. Will I regret not going with rails even on this first build?

Lastly, frame material. Again, this might be controversial - I am thinking of using plywood! My reasoning being that I initially want something quick and easy to build that is rigid enough to start making parts for a mk2 version (there is bound to be one so I might as well plan for it...). Steel would need welding to make it worthwhile (I haven't used my welder in 10 years) and would likely resonate unless filled with epoxy granite or sand. It is also hard to modify and experiment with, as is aluminium (unless I had a working cnc router). Wood is good for stopping noise and resonances, can be stiff if built properly (especially given it is a fixed gantry design) and is cheap an easy to work with. The down side is that it is almost impossible to make things accurately and it likely to distort and move over time. I am thinking that is not really a problem for my first build though as it should be "good enough". This is for the main frame and gantry (and maybe) the moving bed. The z axis assembly is likely to be aluminium. Obviously it will be thick laminations and box sections to get the rigidity, not just flat ply sheets.

So, am I being silly? ;)