AndyUK's Build - 1.2x1.0m Gantry

[AndyUK]

I promised an update on the frame - here she is in all her badly painted glory!!

Since the last photos, I've

• Faired the edge of the epoxy into the frame and repainted the box sections.
• Filed off the meniscus along the edges and at the ends.
• Marked drilled and tapped the epoxy for the rails.
• Roughly aligned the rails for parallelism
• Checked the levelling with a precision straight edge - more details below.
• Thrown the gantry on the rails - it glides super smooth (and made me very happy!)
• After getting happy about that, I then got very sad. Turns out my steel frame is a bit of a rhombus! When I bring the gantry to one edge, it hits the end of one rail about 3-5mm before the end of the other, and the opposite occurs at the other end. I don' t think this will actually affect anything, except perhaps my pride...
• I then did some more milling, making my Z and Y nema 23 motor mounts. That was fun. :)
• Did a rough assembly - All I need now to make the Z and Y axis move under their own power are the belts which are on order. It was nice to be able to measure the distances between pulley centres physically rather than just in CAD.

The rails both show the same trend, they're high at the ends and low in the middle with an error of about 0.05mm - I'm sure this could be better, the epoxy still set relatively quickly, so I'm sure there are improvements to make. It could be that I just haven't filed down the ends of the epoxy enough and they're slightly propped up at the ends, but its 'close enough' for me at this stage. The other consideration is that this value is the stated accuracy of my straight edge, so optimising it now seems counter productive. I will probably come back and shim the rails, but I'm keen to do a 'proof of concept' and am aiming to have the machine cut something by Christmas.

I've noticed that the Z axis isn't as smooth as the others; I think it might be the alignment of the ballscrew vs the rails, I've discovered that my design is a bit of a nightmare to try and measure the alignment between the two, because the rails and the screw are attached to different parts. The other thing it might be is that the Z axis is the only one thats had a load of EP2 applied - does that typically increase the resistance on the screw? I can turn the pully by hand no problem, but you can feel that it has more and less resistance at different ends of the travel, which is something I don't feel on the Y. Any hints or tips?

The other problem I'm now considering is how to check the axis are perpendicular... can anyone point me in the right direction? I think Boyan mentioned needing two precision squares to align properly... I can see why you'd need one, but two?

Thanks!

[Nickhofen]

If the diagonals of the x (long axis ) are equal, then the axis is orthogonal,start from this !
Then bring the Y axis at a point and measure with a square, that you know it is at true angle.
How you know that you square is really a square?
Check this video:
https://m.youtube.com/watch?v=gDh-E5xhm0o
If your square is not as it should ,do not through it away,check on YouTube how you can make it square again.
Good luck!
And remember any precision tool that is fall on the ground or it it is banged on something hard,stops to be a precise any more,it needs re calibration.

[NordicCnc]

Hello. Looks really good! Is there any benefit if looking a how rigid the machine is by having the gantry rails on top and at the bottom, compared to mounting both on the same side?

For example if we compare these 5 options, which one is the best for rigidity?

1. Both rails on the spindle motor side of the gantry, ballscrew in between rails.
2. Both rails on one spindle motor side of the gantry, ballscrew in between on the opposite side.
3. One rail on the spindle motor side side, ballscrew on the same side and the other rail on the top of the gantry.
4. One rail on the top, one at the bottom and the ballscrew on the opposite side of the spindle motor side.
5. One rail on the top, one at the bottom and the ballscrew on the same side as the spindle motor side.

[AndyUK]

1. Both rails on the spindle motor side of the gantry, ballscrew in between rails.
2. Both rails on one spindle motor side of the gantry, ballscrew in between on the opposite side.
3. One rail on the spindle motor side side, ballscrew on the same side and the other rail on the top of the gantry.
4. One rail on the top, one at the bottom and the ballscrew on the opposite side of the spindle motor side.
5. One rail on the top, one at the bottom and the ballscrew on the same side as the spindle motor side.

See the Beginner Moving Gantry Design Thread

[AndyUK]

Had a weekend playing with the machine and finishing up a few bits.



I've used a 75mm cable tray to hold the 62.5mm wide X cable chain and fitted it with 3D printed supports. Looks great, and I think that cable tray size must be made for the cable chain, smoothly fits right in and glides in and out when going back and forth (gosh this post is starting to get a little suggestive...).

I've tidied up the wiring around the machine generally, although I still need to make a control box for all those switches!

All the limit / home switches are in place and working nicely, again 3D printed mounts. Thanks for the advice on placement Jazz.

I've got the coolant circuit filled and burped - the flow rate is absolutely tiny though, just a trickle, serves me right for running such a long loop with just the crappy little pump that turned up with the kit. Its 15m of loop in all due to the X axis stepper cooling, and 8mm piping, so I might need to upgrade the pump.

UCCNC is now controlling the spindle speed and direction which is great, and I'm using an old Xbox 360 controller to move the machine around which works great. However, I'm seeing the digital display on the VFD fluctuate and can hear the spindle isn't at a steady speed. This is an issue I've seen described with the analogue control methods with the UB1 and UCCNC, so I suspect I'll need to change over to a digital control signal in the future.

Regards spoil board, I plan to use trespa, but am using some cheap MDF to start with in an attempt to get some mistakes out of the way on something that really doesn't matter.

• Roughly aligned the rails for parallelism
• Checked the levelling with a precision straight edge - more details below.
• Thrown the gantry on the rails - it glides super smooth (and made me very happy!)
• After getting happy about that, I then got very sad. Turns out my steel frame is a bit of a rhombus! When I bring the gantry to one edge, it hits the end of one rail about 3-5mm before the end of the other, and the opposite occurs at the other end. I don' t think this will actually affect anything, except perhaps my pride...

Spent last night doing spot drilling on a waste board at all four corners and roughly measuring the squareness of the machine. I should have paid more attention to the above at the time - Its not that the steel frame is a rhombus. The X and Y axis are not quite 90 degrees, which leads to an error of 4mm across the two diagonals. I have a precision angle, but haven't got a way to nicely attach a dial gauge to the gantry yet - Its on my list though!

Initially I'd forgotten the above, and tried correcting by adjusting the home position switches, assuming the gantry was racking. I got the error to 1.5mm (thats 90.1 degrees) on the diagonal, and the next correction stalled the X axis, so clearly I was pushing the ballscrews out of alignment with one another and was twisting the gantry beyond it's limits. That was a big warning sign that I was doing something stupid; obviously the gantry should be nicely aligned to evenly wear the screws and not try to twist against the rails.

Now, I'm not too concerned. Ultimately this just means I've screwed up somewhere along the way getting the squaring on the gantry. Looking at the gantry carefully, I can see the gap between the side plates and the X carriages is slightly larger at the front than the back on one side, and the opposite on the other, which tallies with the rhombus I'm seeing. So I think the main problem is that the carriages are bolted onto the gantry slightly askew. My hope is that addressing this will solve my issue, and mean that without any screws attached my bearings will hit the same point on my rails and remove that earlier error I spotted.

Ultimately, worse case I'll need to remake the bearing plates which those carriages bolt onto (which I already planned to do anyway, I broke a drill bit in one of them so one carriage only has two bolts in - I was kinda hoping I could remake them with the machine itself...!).

I might however do a few cuts with this error in place, just because I'm at the stage where I can start to use the machine for it's purpose, and if I don't get some output soon I think I may get lynched by SWMBO....

[Tom J]

Had a weekend playing with the machine and finishing up a few bits.

Like the idea of hiding steppers inside box. I designing similar machine and will pour granite into top frame and use heavy duty dampers between top frame and stand.

[AndyUK]

Like the idea of hiding steppers inside box. I designing similar machine and will pour granite into top frame and use heavy duty dampers between top frame and stand.

It seems to have worked well for me, but I did have concerns about the temperatures, so they're on the same watercooling loop as the spindle using some aluminium blocks wedged tightly up against them and a bit of thermal paste.

[CNCMEBABY]

Hi Andy,

I'm converting an X3 mill to CNC and after reading your build log amongst others I've been 'borrowing' heavily from your router electronics and circuit design. Could you tell me what software you used to create your circuit diagram please? Also, I saw your capacitor holders and noted that you printed yours after buying the wrong ones - inspired me to get a 3d printer, they were my first job. Just made protective caps for the top too.

I'll put photos of the bits I've made for the mill over on my build log when I've got a little further.

many thanks and stay safe

Geoff

[AndyUK]

Hi Geoff,

Pleased the build log is helping :)

I used MS Visio 2016 to create the diagram - I tried a wide range of CAD and electronic design softwares, but in the end Visio was just easier to use and create a colour coded diagram that made sense in my head. It took a long time to get it all down, but really helped organise my thoughts and work out a load of problems on paper first. I've attached the Visio file if you want to see how its done.

Yeah the cap holders were a bit of a "d'oh!" moment - the 3D printer did a great job with those and a number of other little bits around the machine towards the end - I love the cable tidies and the proximty sensor mounts. Unfortunately with house rennovations the poor thing is currently sitting under a mountain of dust.... The cap lids sound like a great idea though - far tidyer than my hot-glue blobs!

Looking forward to seeing your progress photos! :)

Andy

AndyUK Circuit Diagram.zip