PDA

View Full Version : Open source group project, anybody interested?



Wotsit
08-12-2009, 11:31 PM
Are any of the forum members interested in designing an open source cnc project?

I have some initial ideas for design criteria:

The project should be for owner / users of a basic mdf machine wanting to upgrade to a mid level machine

The build cost should be less than ??? (not including steppers or drivers) and as little as possible

The machine should be capable of a tolerance of +/- 0.MM

The final design should be available as GCode DXF and PDF and open source
:feedback:

Tom
09-12-2009, 12:51 AM
As someone nearing the end of my first build - that's a good idea.

What's the aim though? An MDF machine more solidly constructed, or something more fancy (alu sections, etc)? For 200 it's probably either sticking to MDF, or possibly alu with leadscrews rather than ballscrews? But alu would need decent rails to get any benefit - is that achievable with 200?
What are you thinking?

Anyway, I'm happy to contribute my learnings so far...
1) Leadscrew whip is bad.
2) My next router will be more rigid (although the secret aluminium skinning stage for this router isn't complete yet).
3) Ballscrews would be great, but would also be ruined (and pointless) on an MDF machine.
4) Cutting alu takes a properly sorted machine.

aventgps
09-12-2009, 02:40 AM
this is my arduino mega controlled cnc machine....

http://2.bp.blogspot.com/_LPubRWRpa08/Sxf79AM8o7I/AAAAAAAAAGk/sXKjYyf9XyE/s1600/worked1.JPG

http://3.bp.blogspot.com/_LPubRWRpa08/Sxf77loX46I/AAAAAAAAAGc/-j0C8aYQELc/s1600/zaxis.JPG

http://2.bp.blogspot.com/_LPubRWRpa08/Sxf7-h8WumI/AAAAAAAAAGs/dvAVU_0Zof8/s1600/electronics.JPG

and a video of my machine working...

YouTube- CNC TESTING GIRL using arduino mega

it is all open source........ if you want i can post codes and stuff

Wotsit
09-12-2009, 06:40 PM
Ok, I think i think the budget should be left open but lets try to keep costs down.

What constitutes a mid level machine? I'm thinking ball screws, linear bearings a rigid frame and sold / flat bed.

What are our material options considering it will need to be cut with a home brewed MDF machine?

Wotsit
09-12-2009, 06:53 PM
Might also be an idea to choose a supplier that people are happy with so we can use off the shelf parts that are available in the UK, they might even help us along the way.

How about ZAPP automation? (I have no affiliation with this company)

Ross77
09-12-2009, 10:20 PM
Im on board. :beer:

If this is mid level then its a good idea to go with linear rails or at least supported round rail. Ive not had a good experience with leadscrews and delrin but many have, and espesially if the rest of the machine is well designed. :whistling: might be a way to compromies on the cost......

My machining skills are limited and cnc knowledge is growing but I do understand structures and forces so could help here if required.

I'm currently looking at a composite beam design that uses mateials to their best and combine to form an effecient structure.

As I understand it Cast iron is the choice of a professional machine as it is rigid, strong and dense enought to damp vibration. and it is the small movements (deflection)from the vibration that cause poor surface finish or in worst cases tool breakage.

But the high density has the problem of being heavy so requires large motors to move. I propose to use an aluminum skin as a former, and for asthetic reasons, but filled with concrete and steel of the correct proportions and place to make a cheap beam that is much stonger than ali extrusions, able to provide moderate damping but not be to heavy.

I know this sounds similar to the epoxy granite idea but will be much cheaper and hopfully stronger. Its still only an idea and I havent done any calcs yet but the principle has been used in buildings for many years and yeilds a 50% saving in materials for the same strength. :smile:

Sorry if this is off topic but seems a logical approch to an affordable mid range machine.

Any one have any data on cutting forces? I seem to remember 20N for Ali for some reason :whistling:Probably wrong tho.

Ok wake up. I've finnished :heehee:
Later

Wotsit
09-12-2009, 11:10 PM
I like the epoxy granite idea and have followed the thread for a long time but it seems to me a bit to involved for a mid level machine, and has some complicated problems. Do you agree? - Would make a great bed though.

I think the composite idea is a good one, alu box filled and reinforced? Sand in polyester resin / concrete with rebar. Thoughts? Calcs? I wonder what 5kg of resin's exothermic reaction is like?

I agree that linear rails or supported round rail are the minimum for a project of this type

What is a good enevlope? For me it would be 2' x 4' ish

Wotsit
09-12-2009, 11:40 PM
Deleted by wotsit to keep the thread on track

Wotsit
09-12-2009, 11:41 PM
Deleted by wotsit to keep the thread on track

Wotsit
09-12-2009, 11:42 PM
Deleted by wotsit to keep the thread on track

Ross77
10-12-2009, 12:43 AM
Im sorry but I didnt read all that because it is a subject in its self that I'm not interested in it and it sounded like salesmen speak.....:whistling:

My initial thougths were a cheap workable solution to DIY cnc. Stuctural epoxy in the Uk is expensive and as far as I can tell has a poor modulas of elasticity( it still bends) and dosnt solve the high density required to reduce vibration. (Im talking about the machine frame and not the bed)

Yes it has excellent sticking qualities but so dose a better known natural product:heehee:

Srinkage of concrete is a problem but is controllable and the comments on strength laughable. Yes concrete dose get stronger with age but its characteristic strenght is at 28 days. after that it get stronger.

The shrinkage is due to its water content, but I propose to use minimal water and lots plasitiser and as the concrete is only used for the compression then any skrinkage can be zeroed out with shims ect.

For a stucture to work you have to undersand all the forces and all the reactions......................................... .............................................and if you want cheap you have to make compromises:heehee:


As a final note. When I started using my lathe I soon realised that the quality of the finish was proportional to the type(quality ,shape) of tool and speed of the spindle

So dose this mean we should be focusing our efforts on the spindle and operators capabitiy?

Ross77
10-12-2009, 01:21 AM
Not to put you under any pressure Gary but would you concider a group order? may be for X#linear rails and X# ballscrews? What would the min quantiy be?

Just thinking the option of a cheap ish reliable cnc machine would ulltimatly require the support of a willing suppller?????

Gary
10-12-2009, 08:42 AM
Not a problem, you just need to tell me what and the quantity.


Not to put you under any pressure Gary but would you concider a group order? may be for X#linear rails and X# ballscrews? What would the min quantiy be?

Just thinking the option of a cheap ish reliable cnc machine would ulltimatly require the support of a willing suppller?????

CraftyGeek
10-12-2009, 10:43 AM
I'm interested in this as an idea. I'm still refining my first mdf machine to get it to a point that I'm happy to start playing with it to cut 'proper' projects on. I can already see that I would like to use this one to build another machine in the future.

I think an important point here has to be that it uses easily available materials that are still cheap, without using any complex construction processes - I'm sure that epoxy granite is great...not sure how many people would actually make that though.

I think that a combination of mdf & aluminium angle/box/U-channel can be assembled as an I-beam structure that would create a suitably strong & rigid framework.

I also think that it is a good idea to have a ballpark materials costing in mind - 200-250 is a good figure I think. The cutting area, I feel should be no smaller than 2'x4' - & it would possibly be a good idea to make it scalable/adjustable to a point to suit different people.

I'll be watching this thread to see where it goes...

irving2008
10-12-2009, 01:22 PM
...
Any one have any data on cutting forces? I seem to remember 20N for Ali for some reason :whistling:Probably wrong tho.
....

Are we talking high-speed machining or conventional cutting? With high-speed machining you need special cutters and high feed rates if you dont want them to burn out... If we are designing this machine to be 'aluminium capable' we need to work that through and be realistic about our goals. Here is an example working as a starter for 10... dont take it as gospel, there are people who will say " I can cut a 5mm deep 10mm wide slot with my 30k rpm spindle on an MDF machine" maybe they are right, maybe not... but I go by my own experience...

The cutting rate for ali is ~100m/min. With a cutter of 10mm diameter you need a spindle speed of 320 * 100 / 10 = 3200rpm. (the 320 is a contant) (10mm assumes a 1/2" router based spindle)

Aluminium needs about 0.3mm/rev feed rate per tooth. If we assume a 2-flute cutter (1 flutes are hard to find) then we need a feed rate of .3 * 2 * 3200 mm/min = ~1800mm/min. With a 5mm pitch lead screw thats 1800/5 = 360rpm, 6rps = 1200 step/sec, thats going to need top end drivers... or we'll need to gear up 2:1 or 3:1. 1mm or 2mm pitch trapezoidal is probably not going to work. TR12x3 might just.

Aluminium is 17W per cc/min removal rate. So taking a 2mm deep cut 5mm wide with a 10mm cutter at 1800mm/min = 0.2 * 0.5 * 180 * 17 = 306W, say 600W input on the spindle.

Power = torque * revs -> torque = power/revs = 306/3200 = ~0.1Nm, which at a 5mm radius = 20N cutting force... (and thats where the number came from Ross). Remeber the torque must be available at that spindle speed, for many variable speed routers the torque falls off quickly as the speed is reduced.

Dont want this to be a blocker... just so people are aware of some of the design issues and recognise the limitations.. There are some good examples of router-style machines cutting ali.. they are all capable of high traverse rates and have adjustable spindle speeds. A high spindle speed isnt always a good thing, it can burn out cutters if the feed rates are too low... rubbing isnt cutting!

The question I suppose is... do you want the ability to do the occasional bit of small ali work or a machine truely capable of doing it as a matter of course?

Tom
10-12-2009, 05:30 PM
Irving, that's a most useful post - I'm going to bookmark those calculations for future use..., thankyou.

I recon Irving's hit the nail on the head with
The question I suppose is...

The meaning of mid-level could be very different depending on who you're talking to. Although both of these chaps has built/converted a machine to run with EMC2, you'd probably get a different answer about "mid-level" from each... :)

YouTube- Fresa CNC MDF
YouTube- EMC2 5 axis cinci at MPM

So how about a list of variables, so that we can discuss pros and cons?

(low-spec -> high-spec):
* Working size (min A4 -> max 4ft x 8ft flat sheet)(Z depth?)
* Linear bearings (unsupported rails -> skate bearings -> v bearings -> supported round rails -> recirculating ball carriages)
* Drive method (chain/allthread -> belt/trap screw -> ballscrew)
* Target cutting speed vs. target resolution (trade-off)
* Rigidity & strength (just cut alu, or rip through it in a breeze?)
* Spindle type and power (dremel -> 43mm router -> handheld router -> VFD spindle)

Probably loads i've missed, so copy paste and add/change bits...

On my wishlist would be the ability to add a 4th axis (parallel with x) at some future time, without hacking the machine to bits. Designed in expansion capability if you will...

tribbles
10-12-2009, 06:49 PM
If you need PCBs designing, I can do that.

Wotsit
10-12-2009, 07:05 PM
So how about a list of variables, so that we can discuss pros and cons?

(low-spec -> high-spec):
* Working size (min A4 -> max 4ft x 8ft flat sheet)(Z depth?)

For me 2 x 4 is sutable



* Linear bearings (unsupported rails -> skate bearings -> v bearings -> supported round rails -> recirculating ball carriages)

Supported round rails or carriages


* Drive method (chain/allthread -> belt/trap screw -> ballscrew))

Ball screw


* Target cutting speed vs. target resolution (trade-off)
* Rigidity & strength (just cut alu, or rip through it in a breeze?)
* Spindle type and power (dremel -> 43mm router -> handheld router -> VFD spindle)))

Cut alu at a resonable feed rate, spindle down to end user but designed for VFD


Probably loads i've missed, so copy paste and add/change bits...)))

How about bed construction?



On my wishlist would be the ability to add a 4th axis (parallel with x) at some future time, without hacking the machine to bits. Designed in expansion capability if you will......)))

This is important to me

Ross77
10-12-2009, 10:29 PM
Plenty of interest then.:beer:

Defining mid range could be a problem. linear rails, ballscrews and vfd spindle seems quite high spec to me....

theres enough flexibility for people to use what ever they need to suit their requirements. Although I dont think it wouldnt be very economical to mount a dremel on a machine designed to take a heavy vfd spindle?

Thanks Irving. thats perfect, knew I'd seen that somewhere. Is that the rotational force of the motor or the force required to push the axis?

If high speed is needed then would it be better to use belt drive? or would that cause problems with higher powered motors?

For the drivers has any one else seen the "CPLD tutorial" on cnc zone? from memory :whistling:

80V @ 7A with 5 micro steps and works out at 10 to 15 an axis. The only drawback is the chip is a tiny SMD so not very diy friendly. It was proposed to mount a preprogramed chip on a separate daugther board that could plug in to a 40 pin wide DIL and be sold cheaply so people could then build their own but it was never realised. any takers? :smile:

irving2008
10-12-2009, 11:46 PM
Plenty of interest then.:beer:

....

Thanks Irving. thats perfect, knew I'd seen that somewhere. Is that the rotational force of the motor or the force required to push the axis?

No thats the effective linear force pushing back on the cutter. For deflection calculation purposes assume its both perpendicular and parallel to the direction of cut . For leadscrew/stepper torque calcs assume it adds to the load opposing the motion...

Also note in the previous calcs, a screw rpm of 360 means a minmum of a 12mm dia leadscrew for a 1m screw supported drive end only, or 16mm for a 1.5m screw assuming supported both ends (class B support). For bigger 2m screws you're into 20mm+ ballscrew class C support territory and big bucks.

Ross77
11-12-2009, 07:17 PM
Thanks Irving thats what I meant but I pharsed the question wrong. I seem to have a problem getting the thoughts in my head to paper lol

is that 20N a worst case or would it be better to up it a little? Have you got any info for other materials or a table maybe?

I'm starting to see what you mean, this aint going to easy........efficient design isnt easy to build and requires good machining, sounds like the chicken and egg situation I'm in at the moment :heehee:

With regard to the design is space an Issue? ie. overall size. Because I cant help but think that this needs to be a fixed gantry design. The frame could then be as heavy as required and just move the table underneath. This would mean its twice as long tho. :eek:

Kip always said "router for routing and a Mill for Milling" or something along those lines.

Ross77
14-12-2009, 08:10 PM
Ok I been having thinking about this and was wondering if Aliminium U channel is the answer? The only problem is that it would need welding. Steel would be easier but is quite a bit heavier and seems to cost more at the mo.

The idea is use 2 u channels vertical and the fix one top and bottom. This would provide a channel for the ball screw and should be plenty strong enough for dremel, trim router. but if more strength is required then the top and bottom cavity could be filled with concrete and rebar(could be threaded bar that then bolts the uprights to the beam)

Done a few rough shetches to get an idea, I know theres a few mistakes and will probably need some bracing in the webs of the channel (more welding) and the x axis is probably overkill........

Any imediate or obvious problems before I go any further? 100x 50 ali is about 10 a meter so with the 4 bars on an 800mm y axis it would be 30 ish

ptjw7uk
14-12-2009, 08:16 PM
At first glance the design seems to ba all wrong, there is no lateral support for the x-axis and when the head moves it will move in the direction of the movement.
I think some form of twin track stacked one above the other would be better to control lateral movement.
Thats my 2p.

peter

Ross77
14-12-2009, 08:48 PM
Sorry not sure I understand, do you mean because to X axis rails arnt connected? I did say it wasnt finnished yet :smile: mainly concentrating on the Z and y axis :whistling:

Any sketches or pics on the twin track Idea?

ptjw7uk
14-12-2009, 09:20 PM
In the drawing shown 2 bars each side and and 8 bearings each side the bar at top pulls the 2 sides together and will give lateral support.

Not shown attachement of bearings but this arrangement will allow spacers to be used between rods.

Peter

I would like to have a router using this method were the base fixed to a wall as per the saws in B&Q as I am running out of floor space!

Ross77
14-12-2009, 09:55 PM
Still not sure where the problem is? The bearing blocks can take loads in all directions, (slightly reduced where the rail is supported maybe, but thats why I put them vertical rather than on their side)

I see what you mean In your diagram but not sure how that relates to the one I've drawn, dose that use skate bearings and unsupported rail?

I could put another rail underneath like in your sketch but this will add to the cost, And in that case Id rather use linear rails as they would be the same cost as 2 supported rails but are 4 way loading and much much stronger.

If Ive made a mistake I want to known, but at the moment I dont understand......:redface:

I have thought about the wall mount option but not sure how easy it would be to use.... Im sure I remember a design that could be stored on a wall tho.

ptjw7uk
14-12-2009, 10:28 PM
Its just that I dont think that the sitting of the moving part on 2 rails just like a train then there would appear to be very little stopping it rising up unless you intend to use expensive ball slides which grips on their hollowed sides.
I thought the skate bearing system is a cheaper option plus use of unsuported rails although I am unsure myself how to fix them to supports without drilling right through them.

Peter

Tom
14-12-2009, 11:46 PM
Ross, that's an interesting concept using off-the shelf alu sections. What about using one of the slotted alu sections like kev is doing at the moment (ITEM, or similar).

I was buying some router bits the other day from trend, and saw this video. If in doubt, copy the professionals! :) (have you seen the price of these machines from trend!!)

One interesting part of this machine (well, to me anyway!) is the relatively small Z travel, allowing a simple frame.

The vid may also contribute to the bearing discussion. I can't make out the bearing system on this machine (about 1min in is where the machining starts). Are they v bearings?

YouTube- Trend CNC Machining Centres

Ross77
15-12-2009, 01:05 AM
Hi Tom, Yes I have looked at the Al sections and i do have some of the item ones, It just seems an expensive route (espessially if you go for the heavy range) once you get all the connectors fixings ect. If budget allows then use it. It sure is much easier and dose look very proffessional.

Cheers for the link that looks good, not sure on the bearings tho, they look similar to pnematic actuators but could just be a low profile rail. with that design the forces will be small so could be similar to the igus stuff.

The small z axis definetley helps as it reduces the loadings. I think the main point here is that it is being used as a router for fairly low precision work. which is much easier than the breif of this project to cut Al.

I only thought of the u channel as it can be accuratley drilled by an MDF machine and then bolted/welded together to make a strong frame.

I was pricing up the large uprights in 15mm Al and realised how expensive they are to buy and get machined. I figured that they only need to be that thick because they are 2 dimensional, if there was some bracing in the 3rd dimension then they could be a lot thinner, hence the 3.175mm U channel. I maybe wrong and someone else has already tried and failled.

Is it worth Re-clarifiing the goal, As I understand its a 2"x4" (cutting area?) mid range with Al frame(that can be made on an MDF machine if required) and uses supported round rail or linear bearings with ballscrews to cut Al.????? Probably 500 for frame and rails (ballscrews vs lead screw and motors/drivers optional)