View Full Version : Fixed Gantry small precision begineer project advice

22-09-2016, 03:32 PM
Hello everyone,

Most of the projects i have seen here are mobile gantry so i guess my machine will be uncommon due to the small size. I am completly a begineer so feel free to shout me if you notice something shamefully wrong, i appreciate any advice.

I wanted to build a machine for 3D work, mostly propellers and parts of small RC planes ,and maybe jewlery and engraving. So these are the constrains and requirements :

-Small cutting forces due low deep steps(400-800w spindle) ,
-High accuracy and no backlash
-Not any speed/aceleration requirement, not a problem if a job takes 8h or 20h .
-Weight that allow to move it with the table (no frame, clamped on table) withouth a lift machine or disassembly the stages.
-Arround 170x120x52 mm workspace
-Easy alignement or low amount of tooling required

This is the preliminary design (although the aluminum of the plates of the Z column is not ideal due to the fatigue problems of bending in a few years, my concern is more about the rigidity of the design ):

It has profiled rails Hiwin HGH15 , and ballscrews 1402 or 1605 L160/250/300mm, Nema23 , both ends fixed although i didnīt draw because i still have to choose the exact end fixing, nuts, motor support, pocket dimensions etc... But the thickness of the plates of the column is 30mm and the basement of x stage 25mm, total weight arround 25-30 kg withouth motors/spindle.

I have some doubts about the alignement of the stages:

-For the X and Y squareness i found a video that does the same steps i am going to follow https://www.youtube.com/watch?v=UETVKpmWTL0

My idea using a combination of machinist square and gauge block to square the X rail to de Y plate, instead of the ground angle plate he is using because it doesnīt seem to me enough accurate those that i have found.
-For the Y and Z squareness i have thought this, and i wonder i the machinist square is enough straight to alignement of the rail as a straight edge


Boyan Silyavski
23-09-2016, 12:23 PM
HI Yasu,

My advice:

-use 800w water cooled spindle, it won't waste too much electricity and is powerful enough for serious aluminum jobs. Even steel with miniature bits.

-if i am to machine the parts, as we spoke in PM , better design it so that the rails are say 2mm deep in a channel which is wider 2mm than rail. So when i machine all parts, you will not even need a square, cause you could align the rails to the edge of channel. Just check how deep they must be so nothing interferes / the carriage/ and have in mind the edge is beveled so check that out too, for the depth i mean.
Also when you draw the holes have in mind how rails are cut / both end equal spacing to first hole/

By the way i have somewhere here 2 aluminum mounts for the 800w spindle.

Thats what im refering to:


25-09-2016, 09:15 PM
Thats a great idea. I have intention to order some tooling that i could use for part of the alignement like a dial test indicator of 0.002mm and a precision machinist square but this simplifies the alignement by a lot.

But i have more doubts now. It wouldnīt be more hard to machining all these channels edges to tight tolerances of straighness same as a straight grounded edge?
Another question is what happen if a channel is a bit higher even if is 0.01-0.02, could it cause tilt? or it is removed by machining the sacrificial surface the the machine is assemblied?

The x and y alignement i have clear now but the z alignement not so clear.
I have seen a method of how a DTI is placed on the spindle and with the readings there are some shims put on the z plate to alignement. That way is it relevant that the column of the Z axis had to be machining to square tolerances and also the base plate to surface tolerance where it sits to have paralelism with the surface where it sits the rails? or it doesnīt matter and is just shimmed later?


I sent you a PM about the spindle mount Boyan.

I guess i should have started the thread on build log section, maybe i could move it

Boyan Silyavski
25-09-2016, 11:21 PM
Before continuing the conversation, tell me what you would be using that machine for exactly?

26-09-2016, 06:05 PM
For engraving and 3D small propeller parts

On my CAM simulations the surface finishing of the propeller to be accurate needs like 0.01-0.02 mm of resolution

Those are some jobs i would perform

Boyan Silyavski
26-09-2016, 07:07 PM
No need for such space technology as you are led to believe then . Otherwise prepare quite some cash, if you want the parts for the machine to be made by surface grinder. I think you are mistaking a normal over all accuracy with real laboratory surface accuracy which is measured at each point. In short, don't worry about that!

I will help you with advice if you dont mind, but first you need to set some priorities.

For example the next questions are:

What's your budget for that machine? What's your idea for that size working area? You dont need bigger, want to keep cost down, or want machine to be portable or...?

One thing that comes to my mind is that if i was doing that machine, i will make it properly so i could mill steel with miniature bits on it. Steel stamps and so. And would not worry too much about spindle, cause yes its capable, jobs would be miniature and 1 job probably will cover the cost of new spindle. A steel stamp for example. In other words i see that aluminum a bit flimsy to my liking. Considering the money you would spend on rails, motors and drivers. And by the way 20 size rails is better, same money, easier to design clearances for ball screws and so.

What CAD you are using?

27-09-2016, 01:40 AM
Hi Yasu,

Sounds like a cool build. I made something similar, but you're starting from a better place by opting for the profile rails (I used the TBR20 supported rails). Here's a 3D vis of what mine looks like:


...and a pic of how the stages looked before the build was finished:


As you can see, I used 20mm eco-cast plate (this means that it was machined flat and the surfaces were parallel) which was bolted to a welded box section (60x60x5mm) frame and yep, the 'squareness' was shimmed in. Compared to some of the master-pieces that have been built by folks on here my build's a bit agricultural, but hey, I had to start somewhere!

What I will say to you at this early stage (and it might not matter to you) is that you could buy yourself a slightly larger work-space by swapping your machine table (the bit that you fix the work-piece to) with your Y-axis stage. ie. make the machine table the long plate and mount the rails and the ball-screw supports to its underside and fix the pillow blocks it rides on to the Y-axis along with the ball-screw housing. I dearly wish I'd done this with mine for the extra travel/work area..! One possible disadvantage here is that it could be a more complex assembly and the motor will be mounted to the underside of the machine table which may introduce extra unwanted resonance (could be an issue for a lighter machine). I say extra resonance because with a machine this size/weight you will already have some resonance which could potentially create ridges on your surface finish that exceed the magical .01mm resolution you have in mind... A good way to combat resonance is to increase machine mass, although at this scale I doubt it's all that beneficial and I appreciate that you're trying to keep the weight of the machine down for the sake of portability, which leads me to my next bit of advice...

I built a bench for my machine and included adjustable feet and castors on each corner (see pics below), when the bench is in position all the weight is borne by the feet (the loaded bench is IMPOSSIBLE to lift) if I need to move it I retract the feet so the castors take the weight and let me move it to where it needs to go.

19339 19340

Good luck with your build, ask plenty of questions - there's a wealth of knowledge and experience to tap into here! Looking forward to seeing it come together.


Boyan Silyavski
30-09-2016, 06:28 AM
I am thinking also that if i was going to have 1 small machine only, it would be a mill 100%, not a router . Why just you don't look at what George offers (http://www.mycncuk.com/threads/10231-Hobby-milling-machines-now-in-stock-Perfect-for-conversions) and upgrade that to CNC .

Add later to that high speed water cooled spindle and you will have all covered. I highly doubt you could design something more rigid for the money.

Look at that build also DIY CNC Milling Machine From Scratch (http://www.mycncuk.com/threads/7662-DIY-CNC-Milling-Machine-From-Scratch) and eventually minimize some details, though i wouldnt.

I don't doubt that if you put your mind to it you could develope your design, but why not follow sth which is already done and checked?

But honestly from what i see on your initial design - savings on material for such a small machine? I could only explain that to myself if you want this machine to be extremely portable. If that's not the case and you decide to continue, better thicken all plates, and thing of steel, not aluminum.

02-10-2016, 10:39 PM
-The worksize area i have in mind is mini-Mill range I draw 170x120x50mm, the portability/weight is not a main priority I thought it would be nice since its very small, but yeah better increase rigidity and plate dimensions.

I have an aproximate budget of 2.5k€ for this little machine . Some estimation of my part list :
3x ballscrews C3 1605 P2
3x Nema 23 steepers plus Leadshine drivers and a cardboard, Power supply
3x BK & Bf fixing, couplings and Nut brackets.
Thats arround 1k and I still have to find a supplier for the spindle. The profiled rails arround 0.3k .

I thought the router could work since the cutting forces required would be small and seems more easy to assembly and buy.
I agree the mill desing with dovetails instead a router with profiled would be more ideal (like Tormach machine has), the problem it seems more hard the CNC convert process than the alignement of the router .

Wal, your machine looks great. Indeed I took your design as a reference when I draw my sketch (I use CATIA V5 CAD).

It seemed to me more dificult to do the soldering frame almost squared and latter to put shims in many points not only 3 to square it. So i thought a Z column with plates squares and clamped to a flat surface more easy, i will have to keep re-thinking.

I found here some series 5083 milled that is close to flat, i didnīt know about Eco-cast i guess that is more than enough flat

Boyan Silyavski
02-10-2016, 11:59 PM
Broncesval have that aluminum, i buy from them.

I meant not that you can not design it and do the job. if you saw some of my videos in other threads, i am milling steel with 6mm bit without any problem on my 2600x1300x200 router, not mill.

What i am saying is that routers are made for wood and plastic and the mill is made for metal in mind:

-first of all its ready, you have to add ball screws and motors. Not mess with all details till you get there
-its made for metal, so solid slides are stronger than roller bearings
-it has the hold down table ready
-you don't need big cutting area
-any small mill is capable of 1mm deep slot cuts. For such a cut with a router you need very serious Z axis. Reinforced in all directions. And made from steel, not aluminum. steel is 3 times stronger, remember that.
-screws and guides are protected from swarf already

You should understand that when you see similar machines like the one you desire / the router/, people make them for fast cutting of aluminum, not steel.

Though using HSM trochoidal toolpaths you can mill steel on any machine/ but for how long the bit will hold that's another question/.

I also suggest you forget about the videos of people milling steel with the routers. Most will try to mislead you to say how great their machine is. even if with unsupported rails, i have seen people/ companies/ insist their machines could do that occasionally. Which means never well. My router could machine steel occasionally, cause i know it can make 6mm deep slots, i tried it the other day. HSM as i said, which is different/ and misleading/

If i were you, i would start thinking backwards when i am in doubt. Where to spend my money wisely:
-what resolution i will need
-what bits i will use mostly, size and speeds and feeds
-what limits that above- the spindle speed, the motor speed, the weight of gantry / not your case/ , the motion controller frequency,
-what type of motors/ i would use servo motors if i were you, i know i will get blamed for that but no more steppers for me/

When i have clear all the above then i will decide on what machine exactly.

And when you are speaking of less than 0.05mm, make sure you could actually measure that. You will need so much additional stuff even to put the machine together.

16-10-2016, 08:11 PM
My first idea to assembly the machine with less than 0.05mm is mostly the steps of my first post.

-Using a straight edge or the surface edge or a small granite surface plate to align 1 rail of x stage and clamp it.
-Later I put the other x rail on the table without clamp it and the X table over the carriages and bolt them. Having in mind that the bolt holes of the X table are precision machined I just move the X table and at the same time I bolt the other X rail. Later I check with a DTI over one carriage the paralelism of the 2 X rails.

-To put the Y rail square to the X rail, I fix a surface plate to the X rail, then I use the inner face of a precision square DIN875 fixed to that surface and the outer face of the square fixed on the Y rail and bolt it to the X table.
-Then I repeat the same steps for the another Y rail and Y table.

-Later i check with the squareness of the Z plate to the Y table with the precision square DIN875 and if needed put some shims, Then I use the precision square as the drawing to align and clamp the Z rail, and later repeat the other process for the another Z rail and Z table
-Finally i put the DTI on the spindle to check squareness to the Y table and put some shims on the Z table if needed.

Thats what i thought to achieve the 0.01 squareness

I will change the steppers for the servos , as loosing steps is not an option for me.

I donīt think i would need bigger drills than 4-5mm , most of the time <1mm and engraving soft metals and ALU but with rigidity and precision. Resolution I estimate arround 0.02mm.

Boyan Silyavski
17-10-2016, 08:14 PM
OK, but still remains the question : From aluminum or steel? Do you need it portable? Steel is 3 times stronger than aluminum which is 3 times lighter and 3 times more expensive as stock