I'm getting quite excited about what I see is being done and what is possible with the OpenBuilds C-beam system. I am an Electronics teacher at a local college, and we have the facility to produce small numbers of photo etched printed circuit boards, which students do as part of their courses. These have to be hand drilled, which for a small board of 20-50 holes (typical student work) doesn't take them that long. However, my designs for various things usually end up around 1000+ holes per board, and often 3 or 4 boards in a project of mine (I don't do small projects...). This takes some considerable time to drill, and some minor mistakes are inevitable. Due to our manufacturing process, the cut sizes of the boards can be up to 3mm variance, so automatic board drilling based simply off designed co-ordinates won't work. My idea: An XYZ CNC drilling machine, with a camera. Take a hi-res picture of the board in situ, find the holes and drill them. Several years ago I wrote a program that processed an image of a stock price graph and converted it back into a time indexed table of prices. Identifying a hole profile should be easy in comparison. Whilst I understand something about mechanics, I am NOT a mechanical engineer, so anything mechanical that I want to build has to be straightforward, not require me to drill accurate holes manually, and bolt / screw together easily. Openbuilds looks like it will allow me to do just that. I would like to be able to drill boards up to A2 size, which is 420mm x 300mm. I looked at the C-beam video on youtube, and decided that I didn't want a table like that machine has, I wanted something like a light duty OX. So I ordered some sample plates, wheels and a couple of pieces of C-beam to get my head around how things might line up. I was a little concerned with potential twist on the Z axis in the X plane, and thought that if Z were supported top AND bottom this would be very much reduced. This would mean two Y axis beams, which would need supporting a fixed distance apart. This in turn could transmit that twist back to the Y axis supports on the X axis, so the X axis rails would also need to be doubled up. I am useless at mechanical drawings, and whilst I can see it in my head, and have tried to describe it, I'm not sure how clear I am being. V-rail or C-beam could be used to provide the vertical spacing of the Y axes, as well as vertical spacing between the X axes, so no special plates or drilling should be required. I also wanted to have the lead screws not exposed to the work cutting area, therefore the X axes have the flat side of the C-beam facing inwards. Similarly Y axes have the flat side facing the Z axis, and the Z axis flat side facing the tool. This, however, means that the gantry plates are all wrong.... Unless I use the XL plates, which can run on the outside of the C-beam, and use two of them, with the C-beam sandwiched between them. This would also allow for greater bearing surfaces of the increased load of extra beams, and could also run 6 wheels per plate not 4. So net result is: X axis C-beam 1000mm, 4 off Y axis C-beam 500mm, 2 off Z axis C-beam 500mm 1 off plus framing members to hold it all square and rigid. X axis gantry: XL plates, 8 off. 48 wheels (6 wheels per plate) Y axis gantry: XL plates, 4 off. 24 wheels (6 wheels per plate) Z axis gantry: XL plates, 2 off. 12 wheels (6 wheels per plate). With a total of 84 wheels, that is something like £550-£600 just in wheels (using Xtreme wheels for accuracy)... I also then realised - this might actually be stronger than the OX, and with no custom plates. Sorry I can't provide mechanical drawings, I just have no clue how to draw it...I suppose it looks sort of like a cube, but with Y and Z axes moveable. Thoughts anyone? Regards, Slick Conway.
Don't exactly follow where you are headed with this but it sounds like you are way overdesigning. The minimal size spindle you need for drilling holes is not going to exert significant rotational force on the X-axis. Something along this line should be more than sufficient.
I don't think you need a high res camera to do what you need to do. This is what a cheap usb bore camera looks like on a 7" Raspberry Pi screen, in bCNC. Probe Camera Alignment · vlachoudis/bCNC Wiki · GitHub My PCB-Milling-Toolchain
Thanks for pointing that at me Rick Yes that is the basic concept that I want to build, and I agree for what I want to achieve, that is perfectly adequate. What I was perhaps less clear about was the potential for twist in ANY machine of this type, and how it might be resolved. I do want to achieve a very high accuracy (tens of microns if I can get it that good), and therefore was considering options to reduce twist to an absolute minimum. As I explored around the various openbuilds videos people have made, I did wonder about being able to expand the machine usage at a later stage. Once idea being putting a rotating spindle with chuck underneath the work area, and being able to convert it to a vertical lathe....at this point greater strength and accuracy for harder materials becomes more important, hence the extra considerations.
If you seek extremely tight tolerances it is best to resist the temptation of upgrading it to a jack-of-all-trades machine. PCB drilling doesn't require a hefty spindle and can be done with a fairly light and nimble machine. When you start increasing requirements you start adding weight to the machine and weight is the enemy of tolerances, affecting everything from backlash to wheel bearing wear. Increases in size are also the enemy of tolerances so making a machine any larger than necessary also has a negative impact. Even modest increases in size can have noticeably adverse impacts on the rigidity of an assembly. If you want a larger, stouter machine that will do anything, fine, but don't expect it to have as good of tolerances as an optimized, purpose built system. As far as resolving twist, I assume you mean minimizing twist as twist is something that by the laws of physics will always exist when a moment or eccentric load is applied to a beam, no matter how small of a moment or load that is applied. Torsional rotations can be reduced by increases in the beam size but again, this increases weight which adversely affects tolerances.
