First Proper CNC Router

[Kitwn]

Generally you don't use a bottom limit switch but you can use soft limits. on all axis.

I've gone with common sensors for limits and homing plus soft limits. Sooooo easy in LinuxCNC. I have a low limit switch on my Z axis, but that's because on my design there's a crash risk with the gantry ends if it gets too low. There is a soft limit as well so it's really belt and braces.

In practice you should never need limit switches to prevent a crash as long as you have the soft limits set correctly and always home the machine before you do anything else after switching it on.

BUT. There may be a risk if you have a pendant for manual control of the machine which does not operate through the controlling software and therefore moves the machine without the software knowing. I think this is possible with some of the cheap Chinese break-out boards and is a train-wreck waiting to happen for the blissfully ignorant.

[diycnc]

-

[diycnc]

After welding the frame, i did a quick check to see how straight and planer the two Y axis tubes are using a small surface plate with a DTI on a long arm. The surface plate was placed on a piece of plywood with three jack screws used to level the plate to three reference points on the tubes at the front left, rear left, and rear right.

Both tubes are now convex, and not planer. Measurements are below. Max deviation is 3mm!
I cant tell you how much is from the welding, and much was there before, but given that both are now convex my bet is its dominated by the welding.


My original plan was to use shims under rails to bring them into alignment, but given the level of error I'm having second thoughts about that.
Is my only option epoxy?

[JAZZCNC]

I'm not sure I'd trust your method of measuring as it won't be very accurate, just the length of the arm its self will deflect a fair bit. Unless you have a flat reference surface to measure from then you'll never truly know for sure how far your measurements are off.

To me you have 2 choices, Fit the rails as it is and see how it cuts and then shim out any issues. Or just bite the bullet and go with epoxy.

What size machine and what's the main usage.?

[Neale]

The long rails on my machine are 1.8m, 100x50x3 box steel. One dipped in the middle by about 1.5mm, the other side a little bit more. Like you, don't know if that was welding or how it came, but in my case epoxy filled the hole

[diycnc]

I appreciate the concerns about the measurement method.
Did some more measurements today to check the repeatability and accuracy of the method.
I deliberately rotated by plywood platform 180 degrees, and positioned the surface plate differently to really test the approach, and pretty surprised to the 1st and 2nd measurements vary by only a max of 0.1mm, that more than good enough for me.
I also used the same method to measure the polished edge of a piece of granite upstand, its the straightest thing I have to hand. I got less than 0.1mm deviation along its 500mm length. Again I'm happy with this.

All in all im very happy with the measurements, and im confident they are accurate enough for me.



X and Y axis are 750mm.
Intended for wood and plastic, and hopefully aluminium. I know ill have to take it easy with aluminium, and I'm not expecting it hog out anything more than the smallest of cuts, or the highest of surface finish quality.

Going back to shims, my concern is will there be a noticeable lack of stiffness with the rails supported on just a handful of points?

[Boyan Silyavski]

Going back to shims, my concern is will there be a noticeable lack of stiffness with the rails supported on just a handful of points?

No problem if you use epoxy putty. I have never tried it but if you think of mounting the rails and then pouring epoxy, i believe it will be difficult to take the air out, though probably will work if you use West System epoxy as its quite thin.

To be 100% sure, best method is pouring the epoxy 5mm deep and 40mm wide, you could use only one channel to connect but make sure channels extend additional 20cm from the frame, there is a lot of info on forum. In my second build from signature i have explained in great detail how to do it properly and most economically for that matter. It cost me a couple of hundred $$ to understand the process in detail.


Another cheap option is use calibrated steel bar like 40x5mm or something, put epoxy putty on it and straighten it on place with straight edge. When dry, dill and mount. Its precise and cheap flat surface

At the end all depends on what shape you rails are in, in most of the cases they are perfect and easy to mount and shim, plus its a small machine, so easier




But epoxy makes that even easier, if you do the epoxy right . Most of the people i am 100% sure do the epoxy in the wrong way or have no way of measuring the result. As funny as it seems i went to that conclusion, measuring my results on a 3m rails and reading the build logs at that time





Similar machine like yours done with 100x100x3mm box , 40x100cm x/y , was well under 0.05mm frame bend when finished.

[driftspin]

HI diycnc

I can say epoxy is a bit of a mess, result is excelent.



But i do feel when you do 2 pours like 1 to get in to plane where thinnest film is 3mm thick or so and then do a final pour of another 3 or 4 will give you excellent results.

Please read up on best practices on this forum.

I had about 3 mm of warp over 1850mm
i did a single pour 4 to 7 mm

I did ignore the reports about epoxy shrink differences where layer thickness is different, which is solved by a second pour.
Maybe I will do a second pour some day...

I have no jobs that need such accuracy yet.

Grtz Bert




Verstuurd vanaf mijn SM-A505FN met Tapatalk

[diycnc]

I have persevered with the shimming approach, and after a couple of evening measuring then shimming, both my Y and Z axis are planer to within 0.05mm.
Measurements are 100% repeatable too, and I have no reason to believe the method has any error at all. For example, after adding a 0.1mm shim to under a bolt, the measurement increases by pretty much exactly 0.1mm at that location.

Only slight tweak to the method is I now do the leveling of the measurements in excel, rather then physically moving the surface plate.

Iv tried to solve the amount of support under the rails issue by adding more bolt, lots more bolt! I have 50mm spacing between bolts now, rather than 150mm as they arrived.

I initially started aiming for a 0.25mm deviation over the Y-axis but blew past that on the first attempt, so just did 1 more iteration to see how good I could get it, and ended up at 0.05mm deviation from a flat plane. It was almost 3mm when I started.
At the moment I reckon the play in the cheep bearings, and maybe even the deflection of the frame under the gantry weight will be more than this, so its total overkill!

Shims are just small pieces of aluminum sheet, cans, and foil. Nothing special.

Im pretty chuffed with the results, it's one of the few things that's gone better than expected.
I painted the frame too, blue Hammerite spray paint, and the paint is so soft I can scratch it off with my fingernail! I clearly didn't do a very good job of cleaning it before painting.


Any suggestions for how to measure straightness of these round rails?

[Voicecoil]

[QUOTE=diycnc;118473]
I painted the frame too, blue Hammerite spray paint, and the paint is so soft I can scratch it off with my fingernail! I clearly didn't do a very good job of cleaning it before painting.
/QUOTE]

It may not necessarily be your fault - over the years I've had very mixed results with Hammerite - it's never dried particularly quickly, but some coats/colours have hardened OK after a week or so, other were still soft months later even with lots of care with surface prep. I don't know whether temperature is an issue in your case, but I've heard other people say it's best to keep it warm to accelerate the drying.

Well done on the shimming BTW