Sturdy and Fast all Steel CNC , my first build

[Boyan Silyavski]

Would it be possible to lock one lead screw and put a dial gauge on the other side and measure how much (little) it can be moved?

As Geoffrey said, a purposeful looking piece of kit - very nice :-)

Interesting. I will do that tomorrow and report back. The gantry will not deflect for sure :-).

silyavski very good work and very helpfull thread.
I am planing a machine a little bit bigger than yours (X:1500mm Y:750mm Z:130mm real travel).
I have been inspired by your plans for my machine frame.
I have some questions about what i see.

1. How did your succeed to have Z axis rails in the same plane? Did you use epoxy, or did you mill the 6mm plates?

The Z axis rails are in the same plane for 2 reasons:
1. When i mounted the bearing blocks first, i checked the plate with straight edge. luckily it was almost straight. Some hits with big rubber mallet made it perfect.
2. Now having achieved to mount them on a plane surface, using precision square i squared 1 raw of them and tightened them. The rails were in the bearing blocks but not attached to anything else. cause when you measure long things, its easier to detect any error.Then using digital caliper i measured 150mm, tightened the bolt on the caliper and from both sides made sure the distance is the same . Tightened a bit, then with the digital caliper made sure its parallel to 0.01mm. Little bumps with small plastic hammer helped. Then tightened it finally.

2. Why you use 2x6mm plates instead of one plate 12mm?

I did not want to wast more time. That's what they had at hand at the metal shop. Luckily, cause they were not perfectly flat, so it took a lot of hammering to flatten them against straight edge.

3. What clamping system are you going to use?

Its up to my friend. On my small machine i have cast aluminum grind flat plate. With a lot of tapped holes. On top of it i mount 10mm hard plastic sheet using bolts at the corners and very very thin double sided glue sheet, the ones they use in photography printing to mount photos to lexan support. I Hot melt glue all jobs. When finished, i use alcohol to un glue the detail.

But with a strong and fast machine that will not always be possible. I know your doubt. I have it for my next machine and still not decided. My best bet will be to use plastic bed. I have to make some visits to the junkyard or an recycling center. And on top of it 5mm expanded PVC . However i plan to invest some money and maybe use steel channel , like a mill, but not on all the bed, just at the nearest side, so to say for fine jobs.

So that's my suggestion. Make 2 areas/ or 3 areas as i will do but my machine will be 3m long and 1.5 wide/ . One with a proper aluminum or steel bed and the other plastic.

Cause MDF will constantly need surfacing, i believe investment here will pay very fast in the future.

Vacuum i am not planning to use. The strength of vacuum is if you have production which is repeatable, or you work with sheets only. However in the future if needed you can make some removable vacuum fixtures.

4. What is the clearance from the lower part of z axis to the cutting surface. Also what is the real traver for your z axis.

260mm and real travel of 215mm

Now, careful cause this is still unconfirmed info and just a point of view. Its just what months of mind games told me. As i said it many times in the forum, cause that's my view for a DIY really multitask machine, i seem to find that the Z to bed/ frame without anything else/ 250-260mm is perfect. With the trade off that when alu sheet is routed it had to be risen another 100mm at least, using MDF, wooden or so, ribbed or solid, removable bed or fixture.

Below is how my z plate ended. As you see the solid bars combined with the rails reinforce the plate and is very doubtful it will bend, think of bending a solid steel bar ~50mm thick and at least 50mm wide, 500mm long. Another pint here is that i use 2 spindle mounts separated at spindle both ends. So the mounts and the spindle further reinforce the thing. No it will not bend in any way.

5. Finally what is the horizontal distance ( from left to right ) for y,z axis carriages and the vertical distance ( from bottom to top) for y and z axis carriages ( outer side of carriages)?
Thanks for you time
Vagelis

What i believe is the absolute minimum for a good strong machine- 240mm for the Z Spacing left right



The vertical is my minimum for such a long Z plate. The width of 150mm is the minial as i see it , again having in mind the length




Hope that helps.




Vagelis, something else.

I have been following your build with great interest, please let me tell you something. This also to the other people that aspire to build a machine. You will need to spend at least 100eur on the following things if you don't have them, to be able to accomplish the precision mounting successfully. They are a MUST. I don't see how it can be done without them and especially by one person.
Believe me on this. Not to mention the satisfaction you will feel when all is mounted precisely.


-1m straight edge or longer.Stainless better

The Must here is to be at least long enough that it can lay on both parallel mounted rails that the gantry slides on. In other words wide as your machine. Without this you will run at the moment of mounting and later at countless problems. The best will be if it has length of the diagonal of the bed or at least of the rail.

Many will disagree but Rectangular section is better than knifed edge, especially if you work alone. Much more cheaper also.

Here is my initial experience with links from where to buy. http://www.mycncuk.com/threads/6546-...o-straight-%29
To summarize it DIN 874 is the magic word. Whatever accuracy. Or Din 866 rule, especially if you go bigger machine build. Like my next one. You guessed right, i will buy at least 2 meter one.

-2x 30cm precision squares . Din 875 is the word here. these will help solder , square the gantry, the gantry to the bed, etc. Yes, 2 of them because many times you will clamp them together in clever ways.


-1x <10cm precision square , Din 875 is the word here. Stainless betterfor the tight fits, squaring the spindle, etc.

-2x cheap chinese squares

Once you have bought a really cool reference square, you go with it to the shop and buy cheap squares to use for soldering and damaging jobs. Once you reference them against the good one.

[EddyCurrent]

Nice looking machine but the eye is drawn to those brackets from the gantry to the X ball nuts. I know from your posts that you will have made sure these are correct but I have to say they look flimsy in comparion to the rest of the machine.
You made a good job with the control box considering what you had to work with, it's been informative and a pleasure to follow your build.

[Boyan Silyavski]

Nice looking machine but the eye is drawn to those brackets from the gantry to the X ball nuts. I know from your posts that you will have made sure these are correct but I have to say they look flimsy in comparison to the rest of the machine.

On the photos maybe so, but in real life they look normal. Furthermore the actual distance where they are not attached to something is very very short, so they are reinforced by the ball screw nut and the gantry.


Now i want to say it:
The machine is ready and running

Thanks to all who helped and all who watched the progress.


I would like to write soon a more detailed final conclusions, as these days again learned many things.

But for now i would say that i am extremely happy with the result.


Now about the final accuracy i achieved. Today i run it for first time cutting. My typical starting test is 25mm OD circle in 50mm square

Pictures speak better than words but lets say that all the careful fiddling payed itself and the final/accumulated/ accuracy of the machine without any compensation on mach 3 motor tuning and so is 0.01-0.02mm or less , having in mind that the digital caliper measuring may be not perfect. As you can see perfect super clean circle cut. No mismatch or whatsoever.
PS.The bed is 5cm thick so the Z at the test was extended at its lowest point.



All motors are tuned at 8m/min but i believe it can do much much more. I will test it further when on the floor, cause its quite dangerous now to move very fast, the supports flex like crazy.

After 1h of continuing run/ i am surfacing the 10mm plastic bed i fitted over 50mm wood and composite board/ the motor temperature at Y and A/the gantry motors/ is 55C and the X and Z motors is 35C . Its 24C outside.

Soon will post video and some further notes .

Thanks again!

[TonyD]

Well done. :)

[JAZZCNC]

Hi Boyan,

Just seen your Pm. Sorry not been watching your thread so much as I tend to not to drop in much on those who are experienced or getting along good.

Well you certainly have progressed buddy and done an absolutely cracking job my friend. Looks excellent withTank like strength. . . . . Health and safety will like that colour scheme certainly won't miss it . . Lol

Look forward to seeing video..!! . . . . . WELL DONE.

[Wal]

Excellent. Really nice build. Well done..!

Wal.

[kingcreaky]

Excellant Machine buddy, a real credit to you.

I am very jealous.

keep us posted with all the things you make on it!.

[njhussey]

Cracking build thread, really solid well built machine and an inspiration...

[EddyCurrent]

On the photos maybe so, but in real life they look normal. Furthermore the actual distance where they are not attached to something is very very short, so they are reinforced by the ball screw nut and the gantry.

I knew you would have got them right, excellent machine and accurate too, just as you intended, spot on.

[Boyan Silyavski]

Just installed it at its new home. The precision continues to surprise me. Instead of accumulating, all small imprecision neutralized each other, may be pure luck but still under 0.01 -0.03mm, it seems i can not measure it properly without introducing error. No need though, i am happy.

Surfacing aluminum resulted in glass like surface, small marks are not seen by naked eye and scratch test with fingernail leave impression of trying to scratch glass.

The C23 Bob works like a charm. But the integrated spindle control is erratic. I supply it with 12v stable voltage and it plays the output to spindle with 0.2v witch translates +-200rpm, which makes the spindle sound annoying. Will contact the maker. As i see it i will change the small pot with a big one external, in the hope to cure this problem. All PSUs i use are overrated at least by 100% and brand ones, so may be its the pot.

Next machine - no Flimsy: 8 wire alarm cables, connectors to box, 1 component paint.

The 0.8kw spindle seems now/ as i knew from the beginning/ way underrated. This baby deserves 3kw one and will be called - The Reaper :-). I am boasting :-)
But for its purpose is ideal.

Here is a video of it working. Unfortunately i forgot to make video when it was slicing aluminum and took it when we were testing some 3D jobs. I am surprised from RhinoCam, quite capable and i was able to figure in a minute how to program the 3d job.

A little noisy video, but i wanted to be heard what sound it makes. If i dont forget in a week or 2 will be making some aluminum parts there, so will make another one.