My DIY CNC machine build mk4

[routercnc]

I've been documenting the build of my latest CNC machine on 'mycncuk' as I'm a long time member there and I'm based in the UK. I've been creating videos of the build and putting them on YouTube and noticed that the analytics show a fair percentage of viewers from the US so I thought I'd post a link here too in case more were interested.

It is an unusual machine with twin ballscrews on X, Y, and Z axis, a fully supported Z axis running on 4 rails and afull perimeter gantry. This is to try and approach a mill style machine for stiffness (to machine aluminium), but still have the large footprint to cut wooden sheets.

Latest video is Episode 8 (earlier videos in the series are also on YouTube):


Similar Threads:

• Build Thread- Router home build Australia. Solid steel build
• Need some direction to build a small machine to allow me to build a larger one
• Mint's Build Aluminum/Steel Build thread.
• Newbie - To build or not to build Router/Plasma Table

[1Jumper10]

Yes, I subscribe to your thread on the UK forum. Incredible job!

[routercnc]

Yes, I subscribe to your thread on the UK forum. Incredible job!

OK cheers 1Jumper10 !

[fretman_2]

Wow...nice job!

[routercnc]

Wow...nice job!

Thanks Fretman

[Andreiir]

Hi dear friend .
As i see u are using round rail for long axis , as i konw they are not that rigid , and i am wonderind why u destroy ur nice build with this crappy rails .
Just my thought .

[routercnc]

Hi dear friend .
As i see u are using round rail for long axis , as i konw they are not that rigid , and i am wonderind why u destroy ur nice build with this crappy rails .
Just my thought .

I agree the round rails are not very good

I'm using them initially because I have them but when funds allow (these are expensive ) it is easy to upgrade. I have spot drilled the holes in the bed frame to take profile linear rail so it will be quick to swap over.

[NIC 77]

Hi dear friend .
As i see u are using round rail for long axis , as i konw they are not that rigid , and i am wonderind why u destroy ur nice build with this crappy rails .
Just my thought .

I agree the round rails are not very good

I'm using them initially because I have them but when funds allow (these are expensive ) it is easy to upgrade. I have spot drilled the holes in the bed frame to take profile linear rail so it will be quick to swap over.

What a minute, I think there's some confusion. You guys are talking about the machine that was used to cut the parts for the new machine, NOT the new machine? The rails I saw in your vids about the new machine look square to me. LOL.

I've watched several of your videos about this build RC, (yes, I have given you a better nic name), really fantastic work. Really awesome work in aluminum. It's not the way I would have done it.....but......how can you criticize such beautiful workmanship?

Are you going to fill your gantry tubes with epoxy granite? They are smallish, so, it shouldn't take too much or add too much weight.

What size lead are you using on your ballscrews?

[routercnc]

What a minute, I think there's some confusion. You guys are talking about the machine that was used to cut the parts for the new machine, NOT the new machine? The rails I saw in your vids about the new machine look square to me. LOL.

I've watched several of your videos about this build RC, (yes, I have given you a better nic name), really fantastic work. Really awesome work in aluminum. It's not the way I would have done it.....but......how can you criticize such beautiful workmanship?

Are you going to fill your gantry tubes with epoxy granite? They are smallish, so, it shouldn't take too much or add too much weight.

What size lead are you using on your ballscrews?

Hi Nic, I can understand the confusion so to clarify:

The MK3 machine is the current working machine you see in the videos. I'm using that to make the parts for MK4, which is the subject of the video build log.
The MK3 machine does use supported round profile rails on the long axis (X axis in the UK), and square profile rails on the gantry (Y axis) and Z axis.

As for the new MK4 machine - if you pause the video on the introduction you may be able to see the CAD image in the top LH corner which is the MK4 machine under construction. As drawn this does indeed also have supported round rails on the X axis, and that is probably what Andreiir was referring to. That is because it will re-use the rails from the MK3 machine for now. Square profile rails of just over 1m are about £220 ($300 US) and I've spent a fair bit so far on parts so it will have to wait.

Without doubt this will be the weakest part of the machine as they are not as stiff as profile rails, are not as accurate, plus they need a little bit more maintenance. You won't have seen this yet as the bed pieces are being worked on and won't be on video for a while, but there is the option to swap them out for profile square rails in the design. The bed supports have been spot drilled to where the profile rails would need to be, plus the gantry ends have been designed to accommodate mounting the different carriages. So it is a quick and simple upgrade.

Hope that clears things up.

I've not seen many people make a machine like this and there is no strong reason to recommend it for most people. The reason I went down this route was that I wanted something stiff enough to machine aluminium on a regular basis, potentially steel (with different spindle arrangement), but still needed a large cutting area to machine wooden panels.

Filling the gantry sections with epoxy, sand, etc. have all been considered many times and the answer is I've not decided yet. There is also the option to fill the fixed raised sections on the bed.

Ballscrews are mostly re-used from the old / current MK3 machine. These are 1605 all round (dia 16 mm, pitch 5 mm) with direct drive. If I was starting again I would go for 1610 on X and Y through a pulley drive. This allows 1:1 for wood (faster), and 2:1 step down for aluminium (slower, more torque). Z would still be 1605 at 1:1

[NIC 77]

I've not seen many people make a machine like this and there is no strong reason to recommend it for most people. The reason I went down this route was that I wanted something stiff enough to machine aluminium on a regular basis, potentially steel (with different spindle arrangement), but still needed a large cutting area to machine wooden panels.

Most people can not build something so precisely aligned or well made, in particular I'm talking about your 4 rail Z axis.

There are two builds I have found that do a super great job on aluminum, and can also do steel:

https://www.usinages.com/threads/cnc....63531/page-54

http://www.cnczone.com/forums/uncate...uminium-5.html

Linux_fan mentioned his machine's performance improved with the epoxy granite, and both of those machines have filling of some kind, so that is why I mention it. Personally, the machine I'm building will already be too heavy for it, but I started back a while ago and didn't realize all the benefits of it at that time.

Ballscrews are mostly re-used from the old / current MK3 machine. These are 1605 all round (dia 16 mm, pitch 5 mm) with direct drive. If I was starting again I would go for 1610 on X and Y through a pulley drive. This allows 1:1 for wood (faster), and 2:1 step down for aluminium (slower, more torque). Z would still be 1605 at 1:1

This is a Nema23 driven machine?

Personally, I think the 10mm lead is better for all stepper driven axis, with a pneumatic counterbalance for the Z (if it's really heavy) and direct drives on all. Actually, I think 15mm lead is the sweet spot, but that typically isn't available. For servos, I would guess 10mm still but a belt reduction for the drive. I don't think your spindle will have the HP for 100's of pounds of cutting force, what kind is it? So I don't think gearing would be an issue or necessary, but then again I haven't looked at your specific setup in detail.

During the past year I have gotten into the math, including inertia effects.

http://www.cnczone.com/forums/linear...118-cnc-3.html

Graph in post #5 and a sample calculation in post #21

And another example here:

http://www.cnczone.com/forums/diy-cn...45126-cnc.html

Post #27.

With the 5mm lead, you should be able to cut as fast as you would want to in aluminum I think, but the speed for wood and rapids could be faster with 10mm. Alot depends on the specific motors and power supply you have obviously. Really, I'm not trying to bust your chops over anything, just giving an opinion. Your workmanship is fantastic, I really hope that you are able to cut aluminum like a boss with this thing when it is all said and done. Some of the best workmanship I've ever seen posted here.

[routercnc]

Most people can not build something so precisely aligned or well made, in particular I'm talking about your 4 rail Z axis.

There are two builds I have found that do a super great job on aluminum, and can also do steel:

https://www.usinages.com/threads/cnc....63531/page-54

http://www.cnczone.com/forums/uncate...uminium-5.html

Linux_fan mentioned his machine's performance improved with the epoxy granite, and both of those machines have filling of some kind, so that is why I mention it. Personally, the machine I'm building will already be too heavy for it, but I started back a while ago and didn't realize all the benefits of it at that time.



This is a Nema23 driven machine?

Personally, I think the 10mm lead is better for all stepper driven axis, with a pneumatic counterbalance for the Z (if it's really heavy) and direct drives on all. Actually, I think 15mm lead is the sweet spot, but that typically isn't available. For servos, I would guess 10mm still but a belt reduction for the drive. I don't think your spindle will have the HP for 100's of pounds of cutting force, what kind is it? So I don't think gearing would be an issue or necessary, but then again I haven't looked at your specific setup in detail.

During the past year I have gotten into the math, including inertia effects.

http://www.cnczone.com/forums/linear...118-cnc-3.html

Graph in post #5 and a sample calculation in post #21

And another example here:

http://www.cnczone.com/forums/diy-cn...45126-cnc.html

Post #27.

With the 5mm lead, you should be able to cut as fast as you would want to in aluminum I think, but the speed for wood and rapids could be faster with 10mm. Alot depends on the specific motors and power supply you have obviously. Really, I'm not trying to bust your chops over anything, just giving an opinion. Your workmanship is fantastic, I really hope that you are able to cut aluminum like a boss with this thing when it is all said and done. Some of the best workmanship I've ever seen posted here.

Wow, thanks for the encouragement! Building a CNC machine can be a long road.

The gantry machine looks great, thanks for the link. But I have promised the wife this is the last one . . . . so must stop looking at other machines.

My motors are all Nema 23, driven by home made linear power supply giving around 70V. There are 2 on the X (long axis), 1 on the Y (driving twin ballscrews), and either 1 or 2 on the Z depending on how it goes. Objective was to drive the spindle equally from both sides in all axes.

The motor drives are not that special, had them a while. They are analogue "Stepper Digital motor driver 80V/7.8A/256 Micro Wantai DQ860MA". Work fine but AM882 digitals would be the better option if starting again as they are reportedly smoother and have other features such as fault alarms signal.

Agree 5mm is fine for aluminium and 10mm for wood, but I already have 1605 screws so will use that. I'll experiment with 1:1 and 1:2 (step up) ratios on X and Y as they are both pulley driven. I'll check what the max speed is before stalling on the 1:2 and if ballscrew whipping is a problem and play around with fixed/fixed bearings etc.

For background I've been building these for a good few years now and this is my 4th machine, with iterations on each along the way. In the UK we use the expression 'Trigger's broom' which is the exactly the same broom, just with new handles and new heads fitted over the years !

[ericks]

Great job.... you are correct in that it's a long road, long and quite expensive

[routercnc]

Episode 9 complete. Finishing off the lower cover plates, including a bit of '3D printing . . . .'

[pippin88]

Very nice build!

[routercnc]

Very nice build!

Thanks Pippin. One step closer . . .

[routercnc]

Here is episode 10 - machining parts for the bed:

[Biggs427]

Great build!

[routercnc]

Thanks. Just need to check I've machined everything before dismantling the current machine and then it's on with the assembly.

[fretman_2]

You should be proud of that machining! I would be!!!

[routercnc]

You should be proud of that machining! I would be!!!

Thanks Fretman. I like trying to make everything as good as it could be - just takes much longer !