View Full Version : CNC Frame Design Help

14-11-2015, 07:05 AM

I am planning to design my first cnc that is mainly for PCB and aluminium. The dimension of is about 700mm x 700mm x 700mm. The design is to have the spindle to move in x and y axis. The bed will be moving in the z axis. Would be glad if anyone willing to give suggestions.

The current specifications are:

1. 8080 aluminium profiles for the frame (with angle brackets ?)
2. 1605 ballscew bearings (2 x-axis, 2 y-axis and 2 z axis)
3. Nema 23 stepper motor

Attached are the design from various perspectives.


14-11-2015, 12:47 PM
Hello and Welcome.

Ok well for both these Tasks then this design isn't very good.? The moving bed is too weak for aluminium cutting and the level of accuracy you'll need for PCB this setup will make it very difficult and cause more trouble than it's worth.

If you need high level of Z clearence then I would look to a conventional Gantry setup but have the Gantry raise up/down in place of conventional Z axis.
This will give you high strength and great accuracy while having large Z axis clearence.
Your idea of the bed just being supported by unsupported linear guides and just one ballscrew at each side will fail badly i'm afraid. To work any where near good enough you'll need the rails supported and ball screw on each corner. Then you have the problem of keeping the ballscrews in perfect sync so the table is always perpendicular to the cutting tool.
In practice the outcome will be you'll always be chasing the bed and needing to surface the area you want to work in if accuracy is required. I build machine that use adjustable height tables and while they work great in allowing nice range of materials to be used it's not something you move up/down often because of the work involved.
Unless you build into the machine a very accurate way of keeping the table perpendicular to the cutter, Ie using Encoders or sensors etc then your design will be more trouble than it's worth believe me.! . . . . . Even then it needs a complete redesign from what it is now.!

Sorry to be so Negative but it's what I see and with the foresight of experience I see huge potential for failure or at best dissapointment.

14-11-2015, 06:17 PM
I'll echo Jazz's comments. Plus it looks like there is no gantry as such, just a pair of rails which support the spindle?

This might be OK on a plotter or laser cutter but if you want to cut aluminium this won't be anywhere near stiff enough.

Combined with the unsupported rails on the sides of the bed and you will do more rubbing and vibrating rather than cutting aluminium.

14-11-2015, 11:57 PM
Stick an extruder on it and you have yourself a wooping good 3D printer!

17-11-2015, 03:14 PM

I'm looking at making a machine similar to yours with a moving bed only abit bigger. What your going to need is HIWIN linear rails or similar http://motioncontrolsystems.hiwin.com/Asset/HG-Series-Catalog.pdf these are available from several sources such as https://www.cnc4you.co.uk/. One of the main reasons this type of linear guide system is superior is that it is screwed down along its length to a (hopefully) solid base, meaning its basically as strong as whatever its mounted to. However they do have a rating, the lowest rating on HG25 is 26kn. A newton is basically 1kg so thats 26,000kg.

So aslong as you screw it down to something solid your good. Which brings me to my next point. Unless you really absolutely need it to be portable then it doesn't make sense to construct a machine with aluminium. Steel is actually stiffer for its weight, its cheaper and you can easily weld it, welding is significantly better than bolting it together. Per weight steel hollow box section is the stiffest strongest thing around for its cost. I don't know if you've ever played with it but if you can imagine a machine built with 100x100mm steel box section welded together, its going to be incredibly strong and stiff.

Why all the talk about stiffness? Well it comes down to deflection, you can get a good approximation of how much by calculating your cutting forces, then use an FEA program or there are some online calculators for simple beams.

Using some feed/speed numbers for aluminium cutters here http://www.endmill.com/pages/training/spdfeed.html you can calculate your cutting forces here http://www.kennametal.com/en/resources/engineering-calculators/end-milling-calculators/force-torque-and-power.html I was surprised to see the forces for a small aluminium cutter are actually alot higher than a bigger cutter in steel, I guess its because you can run a much higher spindle speed in aluminium. I'm getting over 300n Tangential cutting force which means how hard you have to push the cutter sideways through the material. 300n is 300kg, so then you take that number and calculate the deflection in your steel box section and you might get a very low number. Lets say its 0.1mm, you might think thats OK but what you've got to think about is vibration. Vibration is a bad thing for a number of reasons and its really why you need to make the machine as stiff as possible.

Another thing you will want to factor in is the force your stepper motors can apply. Steppers are rated for holding torque or 0rpm. Often you can find a torque curve for your motor, you might be surprised to find how much power a stepper will loose once it's up to speed. You might not ever be able to make that 300kg of force at the feed rate the cutter wants. Search for something like ballscrew force calculator to find how to calculate it. I figure you'd need about 1.3nm torque to get your 300kg force assuming your using 2 ballscrews. Thats alot for a stepper motor turning at a decent rpm, plenty have 1.3nm at 0rpm but not many at speed.

This brings me to my final point you need to use the cutter manufacturers recommendations as a guide, reduce those numbers and you'll find you can build a machine strong enough.

20-12-2015, 09:23 AM
Some updates on v2. Any feedback is welcome. By the way, any consideration in sizing the gantry and linear profiled rails? Currently, I modeled it with HGH35.


20-12-2015, 10:16 AM
Some updates on v2. Any feedback is welcome. By the way, any consideration in sizing the gantry and linear profiled rails? Currently, I modeled it with HGH35.

Kinda hard to see whats going on but you definitely need a wider gantry with 2 carriges per rail like this

20-12-2015, 06:21 PM
Don't need HGH35 that's massive overkill. 20mm will be fine 25mm Max. After that your just wasting money and increasing weight for little to no benifit.

Also Like Rufe0 mentions you need 4 bearings per axis on all axis.

26-12-2015, 03:31 PM
So, I have changed the rails from HGH35 to HGH25. Added 4 bearings on the rails. The gantry consist of 3 8080 profile and hold by a 20mm plate that sits on the profiled linear guideway on both ends. For the endstops, i am thinking of some capacitive sensors and the motors, i guess a Nema23 2N.m. would do?


26-12-2015, 04:44 PM
You could extend these rails to get alot more travel.

Not sure these bits are really providing any extra suppprt
I think these are the motors for the bed, I'm thinking that you've just not drawn the supports for them. If you moved them to the bottom facing up you could get more gantry travel.

If your gantry went all the way forward it would give you two advantages, you could fix work to the front of the machine, so that you can work on all 6 sides of things and you could possibly mount a 4th axis on the front in the future.