CNC router on the cheap

[eried]

Sometimes when I get into some problem, I just go mad trying to solve it.

Weeks ago I bought a second 3d printer (this time a reprap, for an added challenge).


Short history, the build platform it is a little bent and it is just infuriating. I bought several more and they all come bent. After that I added an induction sensor to try to correct the bending via software but it is not perfect (the bend is a curve and the sensor would have to scan millions of points to correct the bending):


Anyway, my post is to share this nice CNC router kit that it is available in http://www.inventables.com fully customized to get any wood/aluminium/etc 3d carving needs fulfilled:

Parts:


X and Y axis assembled:


X, Y, Z and belting done. Missing the bottom structure:


About 5-7 hours of work.

Oh!! and the link with the 3d printing haha... it is that my intention is to 'flatten' it cutting a thin part of the top of the printing bed

[mfleming]

I sympathize with your frustration over accuracy. I have a 13" Southbend lathe that requires about an hour to adjust out any bed twist to minimize taper in a turned part. And this is a 1300 lb (~600 kg ROTW) rigid assembly.

Consider borrowing a machinist level and setting up your assembly on a solid platform, then make sure you are level in the X and Y axes. Even a small displacement in one corner could affect your accuracy. Another lathe setup technique is to turn a bar between centers, measure taper, then adjust leg height accordingly. Perhaps you can do the same with pegs printed a fixed height in a NSEW pattern and measure their height with calipers or depth gauge.

Many of the 3D printers these days claim accuracies near or beyond what I can do by hand. Then I see their shipping weight and I say "Oh my, my Aunt Sally!"

That's a very interesting looking device at the link you gave. Please post a picture of some test object you make. Maybe a case to hold a variety of batteries to power the calculator you find at the next flea market? I think I'd buy one of those

~Mark

[Graan]

Wonder if it can be used for PCBs as well?

[eried]

(04-17-2016 06:53 AM)mfleming Wrote:  Many of the 3D printers these days claim accuracies near or beyond what I can do by hand. Then I see their shipping weight and I say "Oh my, my Aunt Sally!"

Yeah, usually any printer says 50 micron per layer... THEORETICALLY*

This is based on the stepper itself totally isolated. And there are like 10 other layers of flimsy components.

(04-17-2016 06:53 AM)mfleming Wrote:  That's a very interesting looking device at the link you gave. Please post a picture of some test object you make. Maybe a case to hold a variety of batteries to power the calculator you find at the next flea market? I think I'd buy one of those

~Mark

Some of my designs: http://www.thingiverse.com/eried/designs

(04-17-2016 07:22 AM)Graan Wrote:  Wonder if it can be used for PCBs as well?

Sure you can: https://discuss.inventables.com/t/milling-pcbs/8503

[Graan]

Ok seems to work for pcb. But I guess not for more dense layouts. And only for through hole components.

[Maximilian Hohmann]

Hello,

very interesting projects. But do you really think the heated plate of your RepRap will not just take the shape of the rails of your low-cost milling machine instead of becoming flat? It's hard to imagine that this mill offers more precision than the RepRap.

Aren't there heated borosilicate glass plates for the 3D printers to solve that issue with the curved surface?

I am very curios about the outcome, however way you handle it, because I am thinking about buying/building a 3D printer myself and every experience is a welcome bit of information.

Regards,
Max

[eried]

(04-19-2016 01:14 PM)Maximilian Hohmann Wrote:  Hello,

very interesting projects. But do you really think the heated plate of your RepRap will not just take the shape of the rails of your low-cost milling machine instead of becoming flat? It's hard to imagine that this mill offers more precision than the RepRap.

Aren't there heated borosilicate glass plates for the 3D printers to solve that issue with the curved surface?

I am very curios about the outcome, however way you handle it, because I am thinking about buying/building a 3D printer myself and every experience is a welcome bit of information.

Regards,
Max

Yes, I expect the mill to have the same precision than the reprap. But the heated bed was not built by the reprap

I am going to start small trying to fix my old makerbot bed, that is also bent! (but = smaller objects, smaller problems)

[Vtile]

What is the material of the bed? Try to use tempered or boron glass?? If the bed is metal and needs to be I would try to search some material that have extremely low heat expansion coefficient and even then counter the natural heat bending by making it floating and heated evenly by some way. The glass bed is something I have seen in the pictures of some supposedly? high dollar machines, not my invention in any means. One cheap source of nonstick and flat small bed might be also a bottom of a suitable teflon or ceramic frying pan.

[eried]

(04-30-2016 11:40 AM)Vtile Wrote:  What is the material of the bed? Try to use tempered or boron glass?? If the bed is metal and needs to be I would try to search some material that have extremely low heat expansion coefficient and even then counter the natural heat bending by making it floating and heated evenly by some way. The glass bed is something I have seen in the pictures of some supposedly? high dollar machines, not my invention in any means. One cheap source of nonstick and flat small bed might be also a bottom of a suitable teflon or ceramic frying pan.

The bed is made from aluminium. I can't use glass because I need an induction sensor to sense the bed for the auto leveling.

Cheap flat things aren't that flat as you imagine. This aluminium bed looks quite flat to the eye. I need it to be flat up to 0.05 mm or at least 0.1 mm in heights differentials

[eried]

It's ALIVE!