CNC Having a go

Anything not specifically related to NBTV, but at least of some technical nature that might be of interest to NBTV members. Items for sale and links to retailers do not belong here.

Moderators: Dave Moll, Steve Anderson

CNC Having a go

Postby Harry Dalek » Sun Jan 20, 2013 10:02 pm

I am trying to build a CNC Machine just for the hell of it ,seeing Garys and others on you tube i,d give it a go .

Back to finished my post !

I have used Drawer sliders from an old filing cabinet i had in the shed i changed the design which i will more than likely do as i go if i think some else will work better .

But this will do for now to get it together .

I think the drawer sliders are ok for plastic work ,i am trying to build it best i can with the tools i have ,i have finished the X and Y axis movements ,i have to motorize these with stepper motors which will be connected to a long screw ,i need some bearings and see how best i can connect the steppers to this .

I don't know how long this will take to do but theres lots of help on the net seeing others work so it does much design easy .
Attachments
Picture 101.jpg
My first design which is really drawer like .
Picture 101.jpg (207.24 KiB) Viewed 14351 times
Picture 103.jpg
this is how it would of worked for the x axis
Picture 103.jpg (294.21 KiB) Viewed 14351 times
Picture 111.jpg
Not my idea but i saw some one used the drawer sliders on the flat this stopped any side ways movement
Picture 111.jpg (337.02 KiB) Viewed 14351 times
Picture 114.jpg
last photo and this shows the x and y Axis movement
Picture 114.jpg (331.49 KiB) Viewed 14351 times
Picture 118.jpg
Shows the drawer slider i have used from an old heavy wooden filing cabinet i had handy in the shed
Picture 118.jpg (338.59 KiB) Viewed 14351 times
Last edited by Harry Dalek on Sun Jan 20, 2013 10:38 pm, edited 1 time in total.
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Postby Harry Dalek » Sun Jan 20, 2013 10:33 pm

Forgot to add my video on the Axis movements .

As i am making this i under stand once that Dremel is cutting away any force could bend the Axis off a bit so i am trying to make it as strong as i can with what i am using .
Attachments
Picture 124.avi
Axis movements
(4.18 MiB) Downloaded 604 times
Picture 121.jpg
The Dremel mount
Picture 121.jpg (337.71 KiB) Viewed 14347 times
Picture 126.jpg
Idea to attach the screw to the x Axis movement theres a nut in the aluminium girder
Picture 126.jpg (311.65 KiB) Viewed 14347 times
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Postby Harry Dalek » Thu Feb 07, 2013 3:57 pm

After viewing other people making cnc machines i was thinking i better start again before i go to far on this one ...trying a mostly all steel build apart from some wood for parts i can not drill hole sizes for .

Garys advice on using bearings on the Axis movement really worked well .

I have gone with the v bearing idea using angle steel v point on the horizontal for both the rail and the bearing reason being it holds the bearing movement in place its slotted in with the 4 bearing rollers for each rail..... simple apart from having to weld the rail at that horizontal angle .

My first picture i used a grove type bearing cheap and worked very much so i was not happy so looked for the bearing i am using from some scrapped razer scooters you get 4 bearing via one of these scooters.

I used predrilled angle steel any thing i had to drill holes for for the reason saves me making mistakes ! drilling , all i have to do is cut the stuff and bolt it together ...every thing else is welded ...
Well one axis done and am at last happy on to the next Axis .
Attachments
P1040710.JPG
Here was my second idea was not happy with the rollers and it could of pushed upward without another rail rollers above it to much work scrapped that idea
P1040710.JPG (185.52 KiB) Viewed 14328 times
P1040714.JPG
Much happier with the v roller idea just as i said a bit of welding needed
P1040714.JPG (187.11 KiB) Viewed 14328 times
P1040711.JPG
Roller bearing rails
P1040711.JPG (190.44 KiB) Viewed 14328 times
P1040715.JPG
bearing angles
P1040715.JPG (167.15 KiB) Viewed 14328 times
P1040717.JPG
P1040717.JPG (168.22 KiB) Viewed 14328 times
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Postby gary » Fri Feb 08, 2013 6:37 pm

That's looking great Harry, when I saw your first attempt I was going to advise you of the fact that solid wood (especially Bunning's structural construction timber) can swell, warp, and bow enormously with changes in temperature and humidity - that's why most people use MDF or plywood (the latter is better but, as you know, in Australia plywood has strands of gold running through it which can dull your cutting tools quickly - I should point out the I only know there are strands of gold from the fact it is so expensive here, I actually have never found any - it must be well hidden).

However I really like your new solution (steel) and is the way I intend to go on my next machine. Keep the pictures and videos coming. Well done.
gary
 

Postby Harry Dalek » Fri Feb 08, 2013 8:11 pm

gary wrote:That's looking great Harry, when I saw your first attempt I was going to advise you of the fact that solid wood (especially Bunning's structural construction timber) can swell, warp, and bow enormously with changes in temperature and humidity - that's why most people use MDF or plywood (the latter is better but, as you know, in Australia plywood has strands of gold running through it which can dull your cutting tools quickly - I should point out the I only know there are strands of gold from the fact it is so expensive here, I actually have never found any - it must be well hidden).

However I really like your new solution (steel) and is the way I intend to go on my next machine. Keep the pictures and videos coming. Well done.


Hi Gary

I do listen :wink:

I was thinking the same thing on the wood my shed is an oven with these summer days i was thinking apart from trying to do it stronger it would of been dreadful if i finished it and wood bent it would of been pointless using it .
Lucky it doesn't have any gold init it more than likely end up on the cutting base .

This does have me wondering do you have the wood or plastic you are cutting off the table the base a bit as it would be sort of be cut each time you use it other wise ?

I more than likely will get a cnc controller driver off ebay they are very cheap 30 to 50 dollars but found these kits from Oakley electronics ,they do bipolar and unipolar depending on kit .So theres a few ways to go if need be, i like this pdf information as theres a lot on hooking up a stepper to the printer port but bugger all of hooking up 3 steppers on the net .

These pic's are my plans i liked the design nice and simple ,i need to get some more bearings for my next Axis just taking it slow and trying to do it right .

Looking into the steppers i am thinking all i can really buy is nema 17's i do have a large 6 wire unipolar may work as a bipolar but i think i will ebay these Nemas 17's when the times right.

Least i have a base to work from now build up/
Attachments
K142notesc.pdf
Oakley electronics cnc stepper motor kits circuits
(689.89 KiB) Downloaded 631 times
xaxis1.jpg
this is the one that i liked and copied sort of
xaxis1.jpg (116.79 KiB) Viewed 14305 times
linear2.jpg
this is a good shot of the simple roller and rail idea i am using
linear2.jpg (147.71 KiB) Viewed 14305 times
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Postby gary » Fri Feb 08, 2013 8:31 pm

You look to have done a VERY good job with the linear bearings Harry well done - that will see you in good stead later on.

Understand with the steppers (I assume you really mean they will be low torque jobs) - they will work fine - the main difference is the machine will be slow - like mine - good excuse for a cuppa.

I was going to buy that controller when I building mine Harry. I can't remember why I didn't buy them - I don't think it had anything to do with them not being ok.

I will say this however, since it's been at least 5 years since I built mine Harry, those controllers are quite old (design wise) - I think there are probably cheaper and better solutions available nowadays if your have a good look around. For instance they don't do microstepping.

I think I built effectively the same circuit using components from futurlec.com.au fow about a third of the cost - OTOH it may well not have been worth the extra effort - but I like doing those things ;-)

By the looks of things you should end up with something very worthwhile here Harry.
gary
 

Postby M3DVQ » Sat Feb 09, 2013 1:25 am

If you are using threaded rod and a nut you will probably want to devise some sort of anti-backlash system otherwise your accuracy will suffer. It depends how accurately you need to be able to cut of course!
M3DVQ
Just nod and pretend you understand me
 
Posts: 338
Joined: Sat Feb 10, 2007 10:52 am
Location: Lincolnshire

Postby gary » Sat Feb 09, 2013 9:26 am

Indeed, and there are many different solutions to that from the very simple to the outrageously complex.

I found that the use of an all-thread union nut, being very long in comparison with an ordinary nut, averaged out the backlash to a level less than the resolution of the system as measured with a dial indicator.

The biggest problem, surprisingly, is that I found them very difficult to obtain as zinc plated all-thread in Australia is imperial whereas the union nuts are all metric - they are very clever at Australian hardware stores (NOT!). I eventually found a source at a specialist fixings store.
Harry, you may have better luck in Melbourne.

In addition to this most CNC software allows you to enter a backlash value which is taken into account when moving the position.

It is also important to note that all-thread can vary a bit over it's length - this variation must be measured and entered into the software as a scale value - there is a very good tutorial on how to do that, Harry, on the BuildYourCNC site.
gary
 

Postby gary » Sun Feb 10, 2013 9:48 am

harry dalek wrote:This does have me wondering do you have the wood or plastic you are cutting off the table the base a bit as it would be sort of be cut each time you use it other wise ?


Sorry Harry I don't think I answered this very good question. Actually it covers two areas - the first is trivial - most people use a sacrificial board on top of the bed to protect it. Our American friends tend to use a scrap piece of Mahogany or Walnut, or Cherry, or Birch, and sometimes they even deign to use plywood or MDF. I try to explain to them that we don't have "scrap" in Australia, that every tiny piece of natural or man made wood has to be cherished and wrapped in cotton wool in case it gets scratched - they don't care... (but I digress ;-))

So yes, use your least expensive material as a sacrificial bed - you will get quite a lot of use out of it before it needs to be replaced (or skimmed).

Clamping your work is a bigger question and there are many solutions - none of them, in my mind, ideal.

For acrylic (perspex) cutting I generally use double sided tape from the $2 shop. For wood I use a homemade cam clamp, for PCBs I generally screw them to the bed.
gary
 

Postby Harry Dalek » Sun Feb 10, 2013 1:44 pm

M3DVQ wrote:If you are using threaded rod and a nut you will probably want to devise some sort of anti-backlash system otherwise your accuracy will suffer. It depends how accurately you need to be able to cut of course!


Hi Yes i understand the cutting tool and the movement of the screw have to be rock steady or there will be mistakes...

Well i like it to do nipkow disks at least so might be a bit easier doing plastic records and dvds ? find out ...

I am taking this one slow and easy no rush job .
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Postby gary » Sun Feb 10, 2013 1:58 pm

No harry this has nothing to do with how well you build the machine or what you are cutting - "backlash" refers to the little bit a nut on a threaded rod can move back and forth (without turning it) - what it means is when the axis moves forward and then back to the same place there will be a little bit of difference as the nut has moved from one end of it's "backlash" to the other end - the larger this movement the worse the results because they error can accumulate.

Try it yourself - place the nut on the rod you are using and wiggle it back and forth without turning either the nut or the rod - that's backlash. The longer the nut the less backlash you are likely to have as any errors tend to average out.

For cutting Nipkow disks you need virtually zero backlash.

There are all sorts of error sources in these machines - another source of error is "runout" that is the amount of movement in the spindle as it turns. Another can be "racking" where one side of the axis moves less or more than the other side. Yet another is "flex" - especially when cutting harder materials that puts a strain on the machine.

Of all of them, in my experience, "runout" is the worst.
gary
 

Postby Harry Dalek » Sun Feb 10, 2013 2:11 pm

gary wrote:Indeed, and there are many different solutions to that from the very simple to the outrageously complex.



Yes Gary its really hard to pick the best it depends on your money skill and such ...so i have gone for cheapest best way i can do it with my limited tools and skill.



I found that the use of an all-thread union nut, being very long in comparison with an ordinary nut, averaged out the backlash to a level less than the resolution of the system as measured with a dial indicator.



Oh ok i am just using a butterfly nut at the moment i was thinking of putting 2 in either side of the y axis base ...the middle movement was more for a start test to see it i could move if steady...wonder also on your idea if you could weld a few nuts together while screwed on the long thread would give the same results mmm have to test a few unwelded ones to see if they will move together.

The biggest problem, surprisingly, is that I found them very difficult to obtain as zinc plated all-thread in Australia is imperial whereas the union nuts are all metric - they are very clever at Australian hardware stores (NOT!). I eventually found a source at a specialist fixings store.
Harry, you may have better luck in Melbourne.


OH thats why i got a few butterfly nuts while in bunnings i was testing on the thread loose nuts none were right size but the butterfly were fine ...i think the new store big as bunngings Masters it had 20 bucks box of nuts and screws just right for my project see them in the past photos in my frame. They seem to fit on the thread used them for the threads bearings either side .


In addition to this most CNC software allows you to enter a backlash value which is taken into account when moving the position.


OH thats interesting suppose it takes into account your building skills...hope mine are with in range :wink:

It is also important to note that all-thread can vary a bit over it's length - this variation must be measured and entered into the software as a scale value - there is a very good tutorial on how to do that, Harry, on the BuildYourCNC site.


I was thinking you sort of tell the softtware where the ends of each axis are and middle and such but yes i can see the thread movement would be different for different builds and such mmm...Oh well still a ways off more mechanical work still for a while.
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Postby Harry Dalek » Sun Feb 10, 2013 2:24 pm

gary wrote:No harry this has nothing to do with how well you build the machine or what you are cutting - "backlash" refers to the little bit a nut on a threaded rod can move back and forth (without turning it) - what it means is when the axis moves forward and then back to the same place there will be a little bit of difference as the nut has moved from one end of it's "backlash" to the other end - the larger this movement the worse the results because they error can accumulate.


OH i see thats not good i see why you want a longer nut...OKay i will look into see as i go on with it .


Try it yourself - place the nut on the rod you are using and wiggle it back and forth without turning either the nut or the rod - that's backlash. The longer the nut the less backlash you are likely to have as any errors tend to average out.



Well i have heard and read of this i was thinking it was more movement on the frame and bearings as it wasn't explained ...understand now Gary.

For cutting Nipkow disks you need virtually zero backlash.


OH i have gone for a heard one !

There are all sorts of error sources in these machines - another source of error is "runout" that is the amount of movement in the spindle as it turns. Another can be "racking" where one side of the axis moves less or more than the other side. Yet another is "flex" - especially when cutting harder materials that puts a strain on the machine.


Does some like they need to be well made movement wise ,i can see why the better ones have guides and such another good reason to take this one slow and take my time .

Of all of them, in my experience, "runout" is the worst.


I hope i can build it well enough the tips are very good you have given me
i will have a good think about it and look some move what other builders have done .
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Postby gary » Sun Feb 10, 2013 2:35 pm

I think with these machines, Harry, that you have to build one to find out what all the issues are - then you can build another, better, one - the good news about that is you can reuse most of the expensive bits like steppers electronics, bearings, etc.

They are a lot of fun to build and use, so I think it's a worthwhile effort - when I have finished building all my furniture etc. I will get started on a new, bigger, steel frame machine.
gary
 

Postby Harry Dalek » Sun Feb 10, 2013 2:43 pm

Sorry Harry I don't think I answered this very good question. Actually it covers two areas - the first is trivial - most people use a sacrificial board on top of the bed to protect it. Our American friends tend to use a scrap piece of Mahogany or Walnut, or Cherry, or Birch, and sometimes they even deign to use plywood or MDF. I try to explain to them that we don't have "scrap" in Australia, that every tiny piece of natural or man made wood has to be cherished and wrapped in cotton wool in case it gets scratched - they don't care... (but I digress ;-))


:wink: I was thinking you wouldn't want your bed cut to bits each cut ,i was just going to use the planks from my first try as a bed as its the right size and use those thin wood sheets stuck together i t what ever they are called on top of that don't mind if thats cut up a bit.


So yes, use your least expensive material as a sacrificial bed - you will get quite a lot of use out of it before it needs to be replaced (or skimmed).


Like to keep it looking nice for a while so i will see how it gos with another as i said cheaper sheet on top of that .

Clamping your work is a bigger question and there are many solutions - none of them, in my mind, ideal.


Funny i have not noticed others doing this and it has come to mind i looked what you did on the Nipkow you cut out .


For acrylic (perspex) cutting I generally use double sided tape from the $2 shop. For wood I use a homemade cam clamp, for PCBs I generally screw them to the bed.
:wink: [/quote]

I did look and arr garys used tape ,i did think before i looked how the hell it stayed in one place with the cutting !
The electromagnetic spectrum has no theoretical limit at either end. If all the mass/energy in the Universe is considered a 'limit', then that would be the only real theoretical limit to the maximum frequency attainable.
User avatar
Harry Dalek
"Fester! Don't do that to 'Thing'"
 
Posts: 5364
Joined: Fri Sep 26, 2008 4:58 pm
Location: Australia

Next

Return to Off Topic

Who is online

Users browsing this forum: No registered users and 13 guests