450x300 Aluminium extrusion build from scrap/used parts

[Nealieboyee]

Hello boys and girls,
After a long long time of promising myself to build a machine, I'm finally going to do it. I hope to mill/cut plastic, wood, and hopefully some aluminium up to 6mm thick. Accuracy I'm aiming for is around 0.1mm if possible.
Here is my machine plan. I would love some constructive criticism.

Specs:
Y-axis supports are 40x80 extrusion.
X-axis gantry is 80x120mm heavy duty extrusion.
Gantry side plates are 10mm aluminium plate.
machine base is a tooling plate of 10mm aluminium. It has a few holes here and there, which I've modelled in sketchup, so excuse those please.
Final working area is around 440x280mm, but if I can find a bigger base plate, that would be great.
Motors: I'm not sure, but probably around Nema 23 size, perhaps 4Nm?
Rails will be 20mm Hiwin, or Chinese, depending on price.
Ballscrews will likely be 1605.

Would someone be so kind as to critique my build please? :) Note that I haven't included the ballscrews for the x and z axes yet. Still working on positioning those. Ideas?

24576
24577
24578
24579

Thanks all :)

[driftspin]

Hello boys and girls,
After a long long time of promising myself to build a machine, I'm finally going to do it. I hope to mill/cut plastic, wood, and hopefully some aluminium up to 6mm thick. Accuracy I'm aiming for is around 0.1mm if possible.
Here is my machine plan. I would love some constructive criticism.

Specs:
Y-axis supports are 40x80 extrusion.
X-axis gantry is 80x120mm heavy duty extrusion.
Gantry side plates are 10mm aluminium plate.
machine base is a tooling plate of 10mm aluminium. It has a few holes here and there, which I've modelled in sketchup, so excuse those please.
Final working area is around 440x280mm, but if I can find a bigger base plate, that would be great.
Motors: I'm not sure, but probably around Nema 23 size, perhaps 4Nm?
Rails will be 20mm Hiwin, or Chinese, depending on price.
Ballscrews will likely be 1605.

Would someone be so kind as to critique my build please? :) Note that I haven't included the ballscrews for the x and z axes yet. Still working on positioning those. Ideas?

24576
24577
24578
24579

Thanks all :)Hi Nealieboyee,

Good to see you going for a build.
Like me i think you will like building a machine your self.

Things i see that a lot of people do different.

Move the horizontal rails on gantry to top and bottom of the profile instead of front side.

The downside is that it will raise the height of the machine some 5cm


Move the vertical rails to the moving part of the Zaxis.

These changes will be more rigid and the spindle moves closer to the gantry


Please take a look at the routercnc mk3 and the Davek build.


Would end plates connecting the base plate to the Xaxis, to get the Xaxis profiles a little more rigid, be a good idea?





https://r.tapatalk.com/shareLink?url=http%3A%2F%2Fwww%2Emycncuk%2Ecom%2Fs howthread%2Ephp%3Ft%3D10066&share_tid=10066&share_fid=13482&share_type=t

Grtz Bert.

Verstuurd vanaf mijn SM-A320FL met Tapatalk

[Nealieboyee]

Hi Nealieboyee,

Good to see you going for a build.
Like me i think you will like building a machine your self.

Things i see that a lot of people do different.

Move the horizontal rails on gantry to top and bottom of the profile instead of front side.

The downside is that it will raise the height of the machine some 5cm


Move the vertical rails to the moving part of the Zaxis.

These changes will be more rigid and the spindle moves closer to the gantry


Please take a look at the routercnc mk3 and the Davek build.


Would end plates connecting the base plate to the Xaxis, to get the Xaxis profiles a little more rigid, be a good idea?





https://r.tapatalk.com/shareLink?url=http%3A%2F%2Fwww%2Emycncuk%2Ecom%2Fs howthread%2Ephp%3Ft%3D10066&share_tid=10066&share_fid=13482&share_type=t

Grtz Bert.

Verstuurd vanaf mijn SM-A320FL met Tapatalk

Thank you for your kind words :) I've moved the x axis linear rails to the top of the gantry as suggested. I've also rotated the base plate to give me more usable space in the x direction. If I can find a bigger base plate I may use that, or even welding a steel frame....We'll see soon.
I've also added bracing from the y axis rails to the base plate. Do you think it would really help on the x axis too? Those braces also serve as the new plates for the bearing supports for the ballscrews.

As for the z axis, I've moved the linear rails to the moving part of the axis and added side supports to give a bit more strength.
Anything else that jumps out at you?

24589
24586
24587
24588

[driftspin]

Thank you for your kind words :) I've moved the x axis linear rails to the top of the gantry as suggested. I've also rotated the base plate to give me more usable space in the x direction. If I can find a bigger base plate I may use that, or even welding a steel frame....We'll see soon.
I've also added bracing from the y axis rails to the base plate. Do you think it would really help on the x axis too? Those braces also serve as the new plates for the bearing supports for the ballscrews.

As for the z axis, I've moved the linear rails to the moving part of the axis and added side supports to give a bit more strength.
Anything else that jumps out at you?

24589
24586
24587
24588Hi Nealieboyee,

Yes like that. maybe make a full box around the gantry... just make sure no play can occure... i think you have draw side plates for that.

So what size materials would you like to machine? Z is kinda low..
Clamping solution and spoilerboard accounted for?


So is there a butget reason to go stepper direct mount on the bascrews?


If not, 1610 with belt pully reduction 2:1 is popular for several reasons.
15 teeth on stepper, 30 on ballscrew.


BTW what are the cutouts for in the base plate? now they reduce rigidity and make the endplates redundant.

So is the size set or still looking for the right parts to come by?


The best advice ever offered by this forum:

Dont buy any electronics kit / balscrews / rails , before plan is completely done and everything fits in cad.

More experienced forum memeber can help you pick steppers psu and so on.


I only bought stuff when i ran out of work on every step.
This saved a lot of money for me.


Grtz Bert.

Verstuurd vanaf mijn SM-A320FL met Tapatalk

[Nealieboyee]

Hi Nealieboyee,

Yes like that. maybe make a full box around the gantry... just make sure no play can occure... i think you have draw side plates for that.

Done and done :)



So what size materials would you like to machine? Z is kinda low..
Clamping solution and spoilerboard accounted for?
I'm not too fussy on size of material. Obviously the larger the better, but space will become an issue. I've resized the machine, as shown in the images. Final footprint is now 1000x700mm..ish. Thats about as big as I can go. The spoilboard...I was thinking of 25mm MDF? Maybe incorporate that tooling plate full of holes into there somewhere. If, IF I can get my hands on enough free 40x40mm extrusion for a surface, then I may make bolt-through hold downs using inserts.



So is there a butget reason to go stepper direct mount on the bascrews?
Not really. I didn't even think of using belt reduction. What would be the reason to go for that over direct coupling? Resolution? As for budget, I'm putting most of it into decent linear rails and bearings, as well as the electronics. I'm trying to build the frame and mechanical side of the machine as cheaply as possible, using mostly offcut 10mm and 12mm aluminium plate.


BTW what are the cutouts for in the base plate? now they reduce rigidity and make the endplates redundant.
Yeah I've gotten rid of the base plate now. It was just a bit too small to give me enough cutting area for the size of the machine.


So is the size set or still looking for the right parts to come by?
As of now, the size is set at 1000x700. I'll be using 80x40mm extrusion that I may be able to get my hands on for free. If not, I'll weld the frame from 50x50x3 steel box section.

2459324594

[Clive S]

I'll be using 80x40mm extrusion that I may be able to get my hands on for free. If not, I'll weld the frame from 50x50x3 steel box section.

If you can get 90x45 extrusion you will find that the spacing holes on the BK12 and BF12 bearing blocks will line up with the slots in the extrusion

[Neale]

Even if you use 1-1 belt drive, it can let you tuck the motors almost entirely within the machine envelope - reduce space and remove vulnerable sticky-out bits. There is also a feeling that it can reduce resonance effects although I have no experience of this myself. I went belt drive for space reduction, plus thought that it would allow me to change stepper-ballscrew ratio if needed. I found that this was one design decision I did get right, though, and I have stayed with 1-1. Works well for me (similar size machine).

[driftspin]

Hi Guys,

When i was asking for advice for my build on 1605 vs 1610 with belt reduction the motorcalc.xls helped me a lot to understand.

If i remembered correctly, it was about max needed speed vs max ballscrew rotation speed.
1605 direct drive vs 1610 and belt reduction 2:1 opens up possibilities for higher speeds and no loss in resolution.


I found the motorcal.xls helpfull to play with the numbers and start understanding the physics.


https://r.tapatalk.com/shareLink?url=http%3A%2F%2Fwww%2Emycncuk%2Ecom%2Fs howthread%2Ephp%3Ft%3D7355&share_tid=7355&share_fid=13482&share_type=t


PS i did alter the xls for belt reduction.

Hope this helps,


Grtz Bert.

Verstuurd vanaf mijn SM-A320FL met Tapatalk

[A_Camera]

Don't make the Z clearance and maximum Z movement too small because you are going to regret that. Anything less than 15cm is too small in my opinion.

[Nealieboyee]

Hello all,
Been a while, but this is still being built. My linear rail, bearings, ballscrews, supports etc from Fred at BST should be here in a few days. Can't wait. I've changed my design a little bit. It will now be bench mounted, so the base is a lot shorter. See below. I've also swapped out the aluminium extrusion in favour of 50x50mm steel box section. Space isn't a major problem, so I don't have any problem with the motors sticking outside the envelope of the machine. Other than that, is there any other reason I should go with belt driven ballscrews?

Any thoughts?

24814

[driftspin]

Hello all,
Been a while, but this is still being built. My linear rail, bearings, ballscrews, supports etc from Fred at BST should be here in a few days. Can't wait. I've changed my design a little bit. It will now be bench mounted, so the base is a lot shorter. See below. I've also swapped out the aluminium extrusion in favour of 50x50mm steel box section. Space isn't a major problem, so I don't have any problem with the motors sticking outside the envelope of the machine. Other than that, is there any other reason I should go with belt driven ballscrews?

Any thoughts?

24814

Hi Neely Boy,

What is your free Z height ?


Looks like there will be very little room left for clamping solution spoilerboard and length of the cutter...


Maybe you could lower the bed in your design a little to free up some space.


What material thickness are you looking to cut?


Beltreduction has some advantages.

-Foot print
-Option to change steps per mm later on.
-Noise reduction.
-Less stepper vibration transfer to cutter.


Also you can gain speed vs ballscrew rpm
when you raise the pitch of the ballscrew
and "belt pulley reduce" the ballscrew rpm vs stepper without losing resolution.
This is usefull when you hit the critical rpm for a given ballscrew/pitch vs speed.

Also some high pitch screws tend to be more efficient.

But since you already ordered this is a non issue..

Did you order ballscrews prepped for pulleys? I ordered mine with a slightly longer F lenght shaft... 30mm of pulley shaft.

Hope this helps.
Did i forget anything?

Grtz Bert


Verstuurd vanaf mijn SM-A320FL met Tapatalk

[Nealieboyee]

Hi Neely Boy,

What is your free Z height ?


Looks like there will be very little room left for clamping solution spoilerboard and length of the cutter...


Maybe you could lower the bed in your design a little to free up some space.


Whoops....I just placed a sketchup model of the 2.2kw spindle into my model and the tip of the collet was touching box section. I'll need to add....how much below that? 100mm, 150mm?



What material thickness are you looking to cut?

With that z height, 0mm lol. I'm hoping go from 0 up to 30-40mm.




Beltreduction has some advantages.

-Foot print
-Option to change steps per mm later on.
-Noise reduction.
-Less stepper vibration transfer to cutter.


Also you can gain speed vs ballscrew rpm
when you raise the pitch of the ballscrew
and "belt pulley reduce" the ballscrew rpm vs stepper without losing resolution.
This is usefull when you hit the critical rpm for a given ballscrew/pitch vs speed.

Also some high pitch screws tend to be more efficient.

But since you already ordered this is a non issue..

Did you order ballscrews prepped for pulleys? I ordered mine with a slightly longer F lenght shaft... 30mm of pulley shaft.
Nope. I ordered plain finished ballscrews which couple directly to the steppers.
Would 3.1Nm steppers be enough here or should I go to 4Nm? Just realised I only ordered three ball screws, and I'll need a second one for the Y axis...I'll have to get one locally or add it to my order for the spindle and electronics from Fred.



Did i forget anything?

Does it look rigid enough with 50x50mm square tube?

Thanks very much for taking the time to answer my questions. Much appreciated.

[driftspin]

Well 50x50 does not sound like a lot.
webbed in the bed it could be enough.

If you decide to go 50x50 at least make sure it is 4 or better 5mm wall thickness, much easier with tapping threads.

For the size i would go 80x80x4

The main thing is your total footprint.
It is 1000x700 right?

When you can run FEA on your design you know.

I did some basic checks on my design.
Quickly went for 80x80x4 bed and 80x120x4 gantry parts.

Get end caps on the box section.


I think you need to make the gantry as rigid as reasonably possible.




You can easily make the gantry 80x120...
Massive impact on rigidity.
Only 3 cm extra in the design.


Grtz Bert.


Verstuurd vanaf mijn SM-A320FL met Tapatalk

[Nealieboyee]

Well 50x50 does not sound like a lot.
webbed in the bed it could be enough.

If you decide to go 50x50 at least make sure it is 4 or better 5mm wall thickness, much easier with tapping threads.

For the size i would go 80x80x4

The main thing is your total footprint.
It is 1000x700 right?

When you can run FEA on your design you know.

I did some basic checks on my design.
Quickly went for 80x80x4 bed and 80x120x4 gantry parts.

Get end caps on the box section.


I think you need to make the gantry as rigid as reasonably possible.


I went with the 50x50mm because it was free for 7.5m length :). If I go to 80x80mm, I will have to shrink the height due to the cost. The new 80x80mm design would look like the one attached. What do you think? I've also increased the free space between cutting bed and spindle to 130mm. My gantry is already 120x80mm.
https://farm2.staticflickr.com/1849/29725483907_d01e2676db_b.jpg

[driftspin]

I went with the 50x50mm because it was free for 7.5m length :). If I go to 80x80mm, I will have to shrink the height due to the cost. The new 80x80mm design would look like the one attached. What do you think? I've also increased the free space between cutting bed and spindle to 130mm. My gantry is already 120x80mm.
https://farm2.staticflickr.com/1849/29725483907_d01e2676db_b.jpgWell free is the magic word :-)

Just make sure ther is enough meat to tap threads in for the linear rail bolts.

Also you need at least 4cm of width flatsurface to pour epoxy on the steel.
Make sure you have that.

Grtz Bert



Verstuurd vanaf mijn SM-A320FL met Tapatalk

[driftspin]

Well free is the magic word :-)

Just make sure ther is enough meat to tap threads in for the linear rail bolts.

Also you need at least 4cm of width flatsurface to pour epoxy on the steel.
Make sure you have that.

Grtz Bert



Verstuurd vanaf mijn SM-A320FL met TapatalkBtw,

Why are the the z carriages up that high?
And close to gether?

Move those to the top and lower horizontal plates.

So top Z and Y and bottom Z and Y carriage are really close to gether, assuming X is the bed.

Grtz Bert.



Verstuurd vanaf mijn SM-A320FL met Tapatalk

[Nealieboyee]

Btw,

Why are the the z carriages up that high?
And close to gether?

Move those to the top and lower horizontal plates.

So top Z and Y and bottom Z and Y carriage are really close to gether, assuming X is the bed.



Thanks for your comments.

I did that because the further I move them apart, the less z height I have, because I have only 300mm rails for the Z axis.

Here is the machine at the moment. The z axis is at the lowest point of its travel. I've modelled the ballscrews and rails etc exactly as they are. They arrived a few days ago. Still waiting for the second ballscrew for the x axis.

My only problem was trying to keep the z axis as close to the Y as possible, so that there is less overhang or twist forward on the gantry. So I had to move the ballscrew to the back of the gantry. Not ideal, but with everything being 16mm and 20mm alu plate, I figured it wouldn't pose too much of a problem. Thoughts?

I have now sourced 80x80x5mm steel box section, and I got it through a friend at work for £76 for a 7.5m length. Not too bad!

I've also ordered a 2.2kw water cooled spindle, 425oz motors and drivers. I'll be buying the transformers and caps soon for the power supply. Any idea what control board I should use?

Note I haven't modelled the second ballscrew for the x axis. Its basically a copy of the right hand side one.

Please let me know your thoughts?

https://farm2.staticflickr.com/1969/29904457367_df4d946365_b.jpg

https://farm2.staticflickr.com/1897/44791186582_73ef4c4b33_b.jpg

https://farm2.staticflickr.com/1959/44791186322_992ff6535e_b.jpg

https://farm2.staticflickr.com/1854/29904457187_4863617607_b.jpg

https://farm2.staticflickr.com/1863/44791186462_8da1373762_b.jpg

Cheers
Neal

[driftspin]

For controls... well it all depends.


I have UCCNC / 300eth icw two cheap 5 axis boards. trying to delay buying UB1.



Will you be running 3 or 4 steppers?

A single 5 axis board icw 4 steppers and auto homing 4 axis X-A Y and Z only leaves you with 1 estop input left and no Zero probe input.

So plan ahead.

You will also need outputs for
spindle
spindle cooling
air blast
fluid spray
Vacuum

I think UCCNC does everything you need.
Check out what controller fits your needs.
the 100, 400, 300.

I have no experience with something else so please ask around for opinions.

Grtz Bert.






Verstuurd vanaf mijn SM-A320FL met Tapatalk