First steel diy CNC router build

I will start by saying, that this is my first ever CNC design and build, so any critics and suggestions are much appreciated. General idea from here.
Before I start welding, I want to make sure my desired design actually is any good.

So overall description:
Steel frame contains from 100x50mm box sections, 40x40mm intermediate supports; 50x50mm for granty and another 100x50mm in the bottom for locknut. Z axis is built from 20mm aluminium.
Y axis - 900mm SBR20 supported rails (distance between rail blocks outside-to-outside 250mm)
X axis - 700mm SBR20 supported rails (distance between rail blocks outside-to-outside 200mm)
Z axis - 400mm SBR16 supported rails (distance between rail blocks outside-to-outside 200mm)
RM1605 ballscrew for all three axis. Buying complete set from Chai.

For electronics, Nema23 3Nm (425oz.in) should be enough to move this beast around. Buying complete set from here.
Chinese water cooled 2.2kW spindle from ebay as well.

This CNC will be mainly used to mill/cut/drill aluminium. Preferably with air coolant, to get rid of chips.
Like I said above, please fill free to comment on any mistakes in design.

Thank you.

Quote:

---

Originally Posted by ivars211

...
For electronics, Nema23 3Nm (425oz.in) should be enough to move this beast around. Buying complete set from here.
...

---

Don't... They are crap steppers with 9mH inductance... On 36v they'll perform like treacle. Don't buy any unless they are 3.5mH or lower and even those really need 60v supply on 80v drivers

Thank you for your reply, I think I messed up links, the set I was actually talking about was this.

Quote:

---

Originally Posted by ivars211

Thank you for your reply, I think I messed up links, the set I was actually talking about was this.

---

Better, but still a bad choice - go with Irving's advice. e.g these

Whilst it is good to have a large spacing between the linear bearings for better stiffness, you need to consider the 'aspect ratio' (distance between two bearings on rail divided by distance between rails), as if this is too small you'll effectively get lower stiffness due to racking. On what you've designated the Y-axis, this is compensated for by using two ballscrews. However on your X-axis (gantry), the spacing of the rails is large with a small bearing spacing, so when a radial force is applied to the cutter parallel to X, the Z-axis will rotate (or rack/whatever you want to call it).

I'd go with bigger rails if you want to cut aluminium regularly - certainly not less than 20mm. Ideally not round rails, but I realise that's a big increase in cost. Also attach the rails on the Z-axis to the plate which holds the spindle (i.e. the other way round), it's generally stronger due to reducing the overhang.

Thanks Jonathan for your response. The question now is, what's the optimum 'aspect ratio'?
For Y axis - distance between rails 700mm, distance between rail blocks 250mm, so that's aspect ratio of 0.3571
For X axis - distance between rails 320mm, distance between rail blocks 200mm, so that's aspect ratio of 0.625
For Z axis - distance between rails 120mm, distance between rail blocks 200mm, so that's aspect ratio of 1.667

I might consider going double ballscrew design, but that's also increase in cost. I have also read that there is a possibility of racking and damaging machine, if one of the steppers stops while other one is still going. So is the gain really noticeable in relatively small machine? Isn't steel frame rigid enough to prevent racking?

If you want to mainly cut aluminium on this machine then you really need twin ballscrews on your Y axis (X in my world). Even with using steel the gantry will still rack when cutting aluminium or harder materials. Also like Jonathan mentioned the distance between rails on your X axis is too great so either increase clearence or lower the gantry height to bring closer together.

Regards the twin ballscrews then use single motor and join them together with timing belts and pulleys. Use 10mm pitch and gear 2:1 ratio to give same resolution as 5mm pitch but at same time increase torque.
This way it's Win win you get around the racking issues give flexabilty to machine because you can fine tune with differant ratios or increase speed if required with simple pulley change.
Less electronics to go wrong and slight saving on cost's. Motors perform smoother and better due to less resonance affects and with a All steel machine resonance can affect motors quite a lot, esp if using cheap drives.! . . . . This brings me onto Drives.?
Don't buy cheap drives with a steel machine because of resonance, paying that bit more for digital drives will pay back big time in the long run. You'll get much smoother performance from motors but more the point they deal with resonance much better which will drive you crazy and give a very rough running machine if you do encounter it.!

If your dealing with Chai ask him for a price for Profiled linear rails and you may get a pleasant surprise.?

It might be stupid question, but what do you mean with "digital drives"?
I think you convinced me, so I am going to redesign it for twin ballscrew drive then.
Any advice where to buy timing belt and pulleys?

]edited so you did

Yes, I mentioned that in first post as well. Thanks for suggestion, but it seems like he hasn't been logged in for a while.

Quote:

---

Originally Posted by ivars211

Yes, I mentioned that in first post as well. Thanks for suggestion, but it seems like he hasn't been logged in for a while.

---

so you did:distracted: I only remember it because of the cracking job he did on that Harley sign,but that was wood.
like said 2 screws are must for the size of it,i reckon you,d have been buying a second screw inside a week of using it cutting alu ,look forward to seeing how it performs if you get a video up.

ivars211

I'm not knowledgeable enough on the electronics side to recommend anything but this is a popular digital driver for background reading:
AM882 Digital microstepping driver with stall detect

On the mechanical side I would second the twin ballscrew what you are calling the "Y" axis. I used to cut aluminium on a single ballscrew machine with a gantry of around 650mm. You could see the legs of the gantry moving back and forth since the gantry restraint was only in the centre. This effected the cut accuracy which if aluminium is your main material then it will not be sufficient. It is more money, so you can build single now if you like but be prepared to upgrade and build in the holes etc for a twin ballscrew so it is easy to upgrade.

On the "X" axis plate (to which the Z supported rail is bolted to) this needs to be re-arranged at the bottom where it interfaces with the horziontal small plate which connects to the bearing carraige. You have managed to 'trap' the "X" axis plate between the upper and lower bearings which you will have to cut slightly undersize and then shim out etc. Better to grow the "X" axis plate down a bit, and cut back the small plate, then bolt horizontally. Using oversize holes you can tighten these bolts up once everything is set and know that the bearing load is where you want it. Hope this makes sense.

The gantry crossrails are a bit on the small side with 2off 50mm box sections, and no connection between them. I think this will cut OK but surface finish is very sensitive to vibration and I would personally beef up the rails considerably if aluminium is your main cutting material.

Oh, and welcome to the club . . .

Quote:

---

Originally Posted by ivars211

It might be stupid question, but what do you mean with "digital drives"?
I think you convinced me, so I am going to redesign it for twin ballscrew drive then.
Any advice where to buy timing belt and pulleys?

---

No not Stupid question at all.! . . . Digital drives are newer breed of drive that use advanced alogrithams and techniques for working which results in better smoother performance. Older drives used anolgue technolgy and where restricted or lacked advanced features for dealing with issues like Resonance.
Digital drives are much better for this resonance handling alone and the better drives often allow fine tuning thru software if problems do arise but in general they just give smoother running motors and often allow higher speeds.

I use my local bearing supplier but these guys seem ok Timing Pulleys and they will machine them for you at a cost.!

You want HTD 5mm pitch and 15mm. Don't go to small on the pulleys but at same time don't go to large, I use between 18-20T depending if I apply a ratio or not. 2:1 I would use 18/36T.

OK, cheers. I have to re-think quite a lot now. I will post my new design after I have finished.
I cannot tell how thankful I am for all your inputs!

Quote:

---

Originally Posted by ivars211

OK, cheers. I have to re-think quite a lot now. I will post my new design after I have finished.
I cannot tell how thankful I am for all your inputs!

---

Just a mute point when talking about X and Y axis most people on here refer to the X axis as the longer axis (generally) quite often with two screws and the gantry as the Y axis, if you keep to that it will stop a lot of confusion when asking questions. Oh welcome by the way looking forward to see your build log. ..Clive

Ok, I will keep that in mind. The question I am having, if perhaps I am going to buy one pair of profiled linear rails (DFH15), which axis would mostly benefit from them? If I am not wrong it should be Y - 700mm (allowing Z to move).

Quote:

---

Originally Posted by ivars211

Ok, I will keep that in mind. The question I am having, if perhaps I am going to buy one pair of profiled linear rails (DFH15), which axis would mostly benefit from them? If I am not wrong it should be Y - 700mm (allowing Z to move).

---

No all or None is the only way really.

If you must Then Z axis is actually the most important area on the machine because it carry's the sharp bit and if that moves or flexs then doesn't matter how well built the rest is the machine will give rubbish and inaccurate parts.!!.. . . . . . . It's very common for people to under build the Z axis only to instantly regret when the chips start flying, it's also often the most upgraded part of any DIY machine.!

Thanks. The thing is, I cannot really afford going all profile. I wish... someday.

EDIT: I cannot find any information about these DFH15 profiled linear rails Chai is offering me, maybe someone knows?

Quote:

---

Originally Posted by ivars211

EDIT: I cannot find any information about these DFH15 profiled linear rails Chai is offering me, maybe someone knows?

---

Try PDF Linear technology company Ltd. They are actually Taiwanese bearings imported into china. Seems the taiwanese make most of the linear bearings.

I tell you thru experience of using plenty that they are perfectly fine and worth every penny thou the small difference in price between 15mm and 20mm makes the 20mm the better option due to bearing size providing better footprint and giving more clearence for ballscrew on Z axis.

If you want to see a picture of them let me know I have some 20mm here.!

Yes, having some pictures would help. What do you mean by 2:1 ratio, does that mean that pulley diameter(or circumference) that's attached to motor is 2 times the ones attached to x axis?

Quote:

---

Originally Posted by ivars211

Yes, having some pictures would help. What do you mean by 2:1 ratio, does that mean that pulley diameter(or circumference) that's attached to motor is 2 times the ones attached to x axis?

---

Ok I'll take pic for you tomorrow.

The ratio works other way around. The driver is always expressed before the driven. So in this case the motor turns 2 revolutions while the screw turns 1 revolution.
Or put another way the motor moves twice the distance the screw does. So with a 10mm pitch screw 1 revolution of the motor will equal only 5mm movement not 10mm or put another way only half turn of the screw. This gives the same resolution as 5mm pitch at 1:1 ratio but with the added affect of increasing torque but at the cost of speed. Now because the Pitch is higher the same speed is achieved as 5mm pitch at 1:1.

This won't affect you because your not using long screws but the other benifit of using higher pitch screw with a ratio applied is that screw speed is reduced for the same speed as lower pitch screw at 1:1 ratio so reduces chance of whip on long screws. But again at the cost of speed compared to same screw at 1:1 ratio.

Look here for a simple explination of ratios etc. Gears Index Page

Wow, that's a handy website not just for cnc!

Couldn't find many cnc's using timing belts for synchronising axis. So I attached what I came up with.

So the beech colored ones are P32-5M-15F (32T) and the green one is attached to stepper using P18-5M-15F (18T)
If my calculations is right I need around 2380mm of 15mm 5mm HTD timing belt.
A = 940mm
B1 = B2 = R*pi = 50.93*pi = 160mm (assuming half of the belt is on the pulley)
C1 = C2 = 430mm
D1 = D2 = 0.5*R*pi = 0.5*25.46*pi = 40mm (assuming 1/4 of the belt is on the pulley)
E1 = E2 = 50mm
F = R*pi = 25.46*pi = 80mm (assuming half of the belt is on the pulley)
So the total length: A+2*B1+2*C1+2*D1+2*E1+F = 2380mm

The problem is that there are only 2525mm or 2000mm available from here. So I guess I will have to fiddle a bit.
And does anyone knows where to get these rotating pulleys (purple on drawing), at least the proper name for them?

On the side note:
I am about to purchase ballscrews and rails from Chai, just want to make sure I have not forgotten anything:

SBR20-700/900mm(4 supported rails+8 SBR20UU blocks)
DFH15 profiled rails 2x400mm+4blocks
3pcs Anti backlsh ballscrews RM1605-450/750/950mm-C7(3screws+3ballnuts+3end machining)
+1 anti backlsh ballscrews RM1605 950mm - C7(ballnut + end machining)
4PCS NUT HOUSINGS/BRACKESTS(Made of Steel) FOR THESE 4 BALLNUTS
+ 4 set of BK/BF12 with locknuts and circlips
+ 2pc 6.35mm*10mm flexible couplings.

Thanks again.

sorry, delete this.

Quote:

---

Originally Posted by ivars211

Sorry for double post. As I couldn't sleep, I finished x and y axis. I need some approval from clever people, so that I can continue. I also realised that I have no proper bench to put my machine anyway, so I will most likely end up welding also that.
Thanks in advance.

---

Have you thought about how you are going to mount the rail blocks to the frame, for instance the X rail blocks appear to be mounted to the gantry without access to the bolts. I would also wait until the clever people check your frame out and lack of table:cower: I noticed that you are ordering 5mm screws all round!! just bringing this to your attention ..Clive

EDIT: ok I realised I want the frame to be as simple as possible, so 100x50 steel box sections for Y axis makes more sense. Those white plates are supposed to be aluminium 20mm thick. Ignore the missing underneath structure.
Anyway, this is what I came up with after ~14h of browsing through hell lot of designs. Feel free to comment on any mistakes.
P.S. Never thought that design itself would take up so much time.

Hello, I just wanted to ask quick question, is ~35kg gantry too heavy for two nema23 435oz? Does weight even matter so much? Tried to look up on google, but without success.

Quote:

---

Originally Posted by ivars211

Hello, I just wanted to ask quick question, is ~35kg gantry too heavy for two nema23 435oz? Does weight even matter so much? Tried to look up on google, but without success.

---

You're looking at things the wrong way round. Design the gantry to be as strong as you need, then get motors to suit. If the gantry ends up too weak due to lightening it to suit particularly motors, then the machine is arguably no longer any use to you.

The 35kg isn't a problem - the inertia of all the pulleys will be more than the gantry. See these threads:

http://www.mycncuk.com/forums/machin...do-i-need.html
http://www.mycncuk.com/forums/machin...do-i-need.html

The latter was intended for servos, but the formulas to find the required torque obviously don't change so you can use that to calculate the torque required for each axis and compare this to the motor torque vs speed curve.

P.S: Listen to Clive regarding the ballscrews.

Ok, just received steel boxes cut to size. Question is, what's the optimum bolt size to use? I have 3mm thick steel to play with, M6 sounds about right, or maybe I need M8? I tried googling it, but either my keywords were wrong or there is no information about it.

the optimum bolt size is the one that gives sufficient thread. generally you need at least 3 threads. so M6 x 1 will just give you 3, M8 x 1.25 will only give 2, so M6 but you'll need more of them to withstand shear & twisting forces. What many people do is use a 3mm plate inside the box section to add thickness so a bigger dia bolt can be used, ideally welded in place..
.
Also if you're bolting across the box section you'll need some tubing to stop the box being crushed...

Quote:

---

Originally Posted by irving2008

What many people do is use a 3mm plate inside the box section to add thickness so a bigger dia bolt can be used, ideally welded in place..

---

I used Gorilla glue to fasten a piece of 6x30 flat bar inside a 3mm thick box section and it worked great, Gorilla Glue UK | GLUES | GLUE LIST | Gorilla Glue
While I agree with Irving's post about thread sizes, I find a good method is to look at nut thickness dimensions for any bolt size and use that as a guide to how much thread depth is required.

Eddy, that's a good rule of thumb and one I should have thought of!

Here's a good site for all such things:

http://www.roymech.co.uk/

Just remember to compensate when the material you're making the thread in is different to the nut - e.g. for aluminium you want (very roughly) twice as long thread depth.

I have got some 0.5mm^2 4 core shielded cable at home, would the wire thickness be enough for ~3m 36V supply and 4A stepper? Stepper should get around 35.3V at the end. I know 1mm^2 would be much better, but I would like to use what I have on hand already, perhaps I can run two cables in parallel?

Thanks

Personally, I wouldn't bother. I doubt that your power supply is accurate to that level anyway, and it wouldn't make much difference if it were. It takes a bit of mind-adjustment but with these kinds of drivers and motors, the actual voltage as such isn't that important. The driver limits the motor current to whatever you have set, whatever the supply voltage, and the reason for wanting a higher voltage is that it allows the current to rise more quickly to the set value, taking into account motor inductance (which is why the experts on this forum bang on so much about low-inductance motors - it's just so that the current rise is quicker for a given supply voltage and it's the current that makes the motor go faster). Having a slightly lower voltage across the terminals means that the current rise will be very slightly slower but you ain't going to notice the difference in practice; for most of the stepper pulse the driver will be regulating the terminal voltage to less than this anyway.

Quote:

---

Originally Posted by Neale

It takes a bit of mind-adjustment but with these kinds of drivers and motors, the actual voltage as such isn't that important.

---

Actually voltage to a stepper is very important but I agree the difference 35 and 36V won't be noticed. That cable will easily handle the amps you'll be pulling so yes get it used.

I am at a point, where pretty much everything mechanic is finished.
I am building my CNC in garage, and I only just realised that there is no earth wire coming from electric box, there are only phase and neutral. Now I am quite stuck, do I have to dig in some peace of metal (quite impossible in that location), or do I just leave earth wires from spindle and PSUs untouched? Would it cause safety issues? The thing I am struggling the most, is where to connect common earth from all of the shields to? Any ideas how to approach this?
On the side note - how big fuses are you using for VFD+2.2kw Spindle and motors/drivers? I am planning to go 15A for spindle and 10A for drivers/steppers?

Quote:

---

Originally Posted by ivars211

I am at a point, where pretty much everything mechanic is finished.
I am building my CNC in garage, and I only just realised that there is no earth wire coming from electric box, there are only phase and neutral. Now I am quite stuck, do I have to dig in some peace of metal (quite impossible in that location), or do I just leave earth wires from spindle and PSUs untouched? Would it cause safety issues? The thing I am struggling the most, is where to connect common earth from all of the shields to? Any ideas how to approach this?
On the side note - how big fuses are you using for VFD+2.2kw Spindle and motors/drivers? I am planning to go 15A for spindle and 10A for drivers/steppers?

---

You do need an earth in your garage you will need to hammer an earth rod in the ground and connect it to and RCD type consumer unit this will then trip at 30mA and give you protection. An electrician would do this for you. The spindle will run from a 13A plug and the drivers from say 10A.

All the shields and earths in the control box should be connected to one point called a star point and then that point should be connected to the earth. The shields are generally connected at one end only, at the control box. Hope this helps ..Clive

OK guys, just finished my electric installation, got all the wires connected, everything seems to work, except FVD, nothing happens when power is applied. So my question is, am I doing something wrong or do I have faulty FVD?
I connected mains to FVD like this: http://www.mycncuk.com/attachment.ph...5&d=1381658518
but did not connect spindle yet. I would think that it should power up (at least LED showing power or something) even if spindle is not connected.
My FVD and spindle: http://www.ebay.com/itm/250952872638...84.m1439.l2649

Any comments will be much appreciated!

Have you checked to see if you have mains voltage at the terminal in the VFD? With a meter or test lamp. (be careful) ..Clive

I could be wrong but isn't there a system on these that stop it coming on if there is no spindle attached, i'm sure I read somewhere that you shouldn't power them on with no spindle connected as it can damage them, I could have my wires crossed however.

.Me

Quote:

---

Originally Posted by Lee Roberts

I could be wrong but isn't there a system on these that stop it coming on if there is no spindle attached

---

I'm not aware of such a system - I would recommend putting the settings in the VFD with no spindle attached, in case you accidently press 'run' before it's ready.