box ways for pcb mill z axis

I was just reading the build thread for the swiss mak and noticed he is using box ways with a polymer insert. I've been considering taking my spare parts from my previous machine builds and building a dedicated pcb mill. Wondering what people think of using this type (delrin/hdpa on aluminum) of box ways for this application and how best to keep the grit out of them? I may just do the box ways for the Z axis since I only need about an inch of travel. i have some sbr12 rails I can use for the other axis.

My machine build is likely going to have the following:

200x200x30 or 200x150x30 working area in x,y,z

500 w brushless spindle with crappy deep groove ball bearings, seems to mill a pcb ok.

nema17 motors and 2gt belts 0.9 degree and 27 oz/in. I also have 3 nema 23 200oz/in motors I can use if the nema 17 motors are too wimpy.

What would be the advantage of box ways vs. rails for this application?

the advantage is that I can make them. using a rail for 1" travel seems like a waste. I also understand box ways are more rigid and provide better dampening which should help a high speed spindle cut a PCB with a small diameter cutter.

I also want the experience of designing and making a box way.

Please don't make this thread a rail vs way thread.

I made a prototype.

https://youtu.be/jvfQ-infxxE


a ton of questions:

how much adjustment play is typically needed? I was thinking something like 1 or 2 mm.

What ratio of lenth to width is typically considered? I used the 2:1 bearing rule as a guide for this but I was wondering if there were some ratios typically used for ways?

Tapered gibs are ideal but I can't find any guides on how to design them beyond the general rule of thumb that the taper is 1/4 over 12".(~1.193°). What is the best way to retain and adust them via screws? Looking at my gibs on my mill the screws seem very big and eat up lots of the corner. How much of the size is ease of adjustment vs holding the gib in position? I can't imagine a tapered gib exerts much force on the adjustment screw!

Looking closely at the swiss mak mill the pads are very very small and thin. I can't see how they provide a large contact surface and they look like they would make bulges. Since these are so far out on the edges of the ways I can see how it still works but i have not seen any design like it. Has anyone seen a swissmak in person?

I like how the lead screw nut is also trapped by the ways. i wonder how he made the nuts, did he tap them or melt them in?

I read that delrin is 9000 psi shear force and with 3/4" of thread contact on a 1/4 20 thread it would take over 3k lbs of force to strip it. I think that m5 screws should be enough for a small 500w spindle. Should I use M8?

my previous video received some great comments and it was suggested that Rulon XL be used (PTFE). I may order a small chunk since I do not have any onhand. I also want to test out Acetal. I have some seaboard which is Acetal and very slick.

i made a test taper on my main cnc machine and did some measurements on it. Trying to figure out how to design the adjustment screws...

I measured the angle and it is pretty close +- .02mm. Two of them measuring parallel is roughly +- .015mm I'm not sure what is needed and if i can try to get away from trying to scrape them vs lapping them.


https://youtu.be/Un4rJbn3Tgc

I think the taper adjustment is easy as long as the taper is longer than the ways. A simple washer seems like it would work.

Here is a sketch i did.

https://www.cnczone.com/forums/attac...d=407782&stc=1

I did some tests last night on a mdf torsion beam I made for this pcb mill. It was constructed with 1/2" mdf skin and .2mm plywood torsion braces. the dimensions were 400x80x80. I measured around .03mm of deflection with a 40lb weight on the middle. I'm not sure how to measure twisting deflection but it seems very rigid. It is not flat but I can shim it flat.

Here is my progress so far. It has worked out well and I think the box ways will out perform the round rails and the lead screw. Once I complete the frame i can tackle the electronics and belts.

https://youtu.be/dCVRs4excUY

the torsion beam testing is here:

https://youtu.be/NAvOUASyeiQ

and the beams look like this:

https://www.cnczone.com/forums/attac...d=408960&stc=1

Happy to share any of the designs, the torsion beam is semi parametric.


I've also been looking at some hand shaper designs and I think it would be interesting to make a very small hand shaper using this plastic on aluminum method. That sure would put it to the wear test quick.

I found this great project which has a really nice rack and pinion design.

I was a bit skeptical on how easy this would be to re-create but the initial results are impressive and I'm going to be using it on both my x and Y axis.

https://github.com/fellesverkstedet/...er/chamferrail


I'm wondering what method folks use to test backlash? I can see this rail react with .01mm moves so I'm not really sure how to test it.

My longer travel tests had it hitting ~9000 mm/s and repeating to +/- .015 mm moving it 80mm back and forth.