[dfranks]

As I am new to CNC I have been doing tons of research before I start building. I came across a German made CNC router table on Ebay and thought it interesting. As I have examined the pics a little closer I began to wonder if the design was good or not. Here are some of the things that I noticed (as a novice of course).
1.) Very thick uprights on the gantry, maybe 1" or better? The width is not much larger than the width of the stepper motors. The height seem to be minimal as well. The overall working area is 15.75"x11.8"x4.3" though a 39.37"x23.6"x4.3" is available and is more of the size that I am looking to build.
2.) There is no bottom to the gantry. It looks like the X axis rails (rods?) may be pressed in? Anyway I was wondering how stable this would be. The horizontal brace(?) connecting both left and right sides of the gantry appear to be thin, almost more decorative than structural.
3.) There are 2 Y axis ballscrews. I don't know why the idea tends to 'freak me out' but for some reason it does. (I realize that you have to use a controller that supports 2 Y steppers.) I am assuming that if you just use 1 ballscrew on the Y axis that binding is going to occur. Also notice the box mounted under the X axis stepper motor. What's up with that?

So any opinions or other observations? I can see where you might save a few bucks on materials with this design. Oh by the way here is the link to this item on Ebay http://cgi.ebay.com/ws/eBayISAPI.dll...522534852&rd=1
You may want to check out their site which is listed within the Ebay ad as it has more pics and a few videos. Hopefully I pasted the pics below correctly.

[ger21]

Thats a pretty light duty machine. Keep in mind that unsupported rails will always have a bit of sag and flex.

3) I like to call that the X-axis, with the Y-axis being the axis on the gantry. I think you're right, without beefing up that gantry quite a bit, it would bind if not using 2 screws. Those little black boxes appear to be the stepper drives.

That's a nice machine for doing engraving or balsa work, but not really designed for anything heavier.

[Dave's_Not_Here]

... looks to me to be tied together with wide Aluminum plate across the back where the logo is placed...

A channel for the the cable runs across the top... does not appear to be intended for any serious structural support...

Looked at their site... they seem to be able to mill Composit, MDF and Billet Aluminum in their videos...

Hummmm....

[Jayson]

I would agree that the x axis being suspended just on the rods is not ideal. This is a similar design to my machine however my x axis rails are supported. I run 2 screws on the x axis of my current machine both briven off the same stepper but when number 2 comes around I will be going to 1. If you are planning to build a machine based aroung this type of design all I can say is modify it a little then go for it...

Hopefully there will be an image of my machine attached to the bottom of this post. Its only a small image but I can take more if required.

Regards,

Jayson Wallis
(Horsham, Victoria, Australia)

http://www.cnczone.com/gallery/showp...00/ppuser/6402

[dfranks]

Jayson - That is very ingenious on how you are driving both X axis ballscrews off of 1 stepper! What is the actual size that you can cut on your table? On your next design did you mean that you are going to only use 1 ballscrew for the X axis or still use 2 ballscrews and 1 motor for each? Any suggestions on what type of drive (controller?) that 2 X axis steppers can be driven by? I'm still researching as to what desigin I am actually going to use. I have been looking at Beezer's design pretty close. It's very close to what I have been envisioning. Now I will have to look your design over as well. Thanks for the comments. Post more pics if you have them.

[spalm]

Maybe the little black boxes are just cheap "project" boxes used to contain the splices of the motor wires to the cables. Not a bad idea if so.

[Jayson]

Hi All,

In the next design I plan on using 1 motor and 1 thread for the x axis, mounted under the table in the centre. The work area of my current machine is 280mm(Y) x 500mm(X) x 100mm(Z). I can not suggest a controller that will allow you to run 2 motors as I have not done anything of that nature. Good luck with your researching and forget ever sleeping especially after the machine is built... Geez I wish I could stay home and play and still get paid :-)

By the way, I uploaded sone more pics to my gallery for you.

Regards,

Jayson.
(Horshan, Victoria, Australia)

[mad.sculpture]

Hi,whilst looking around at different designs i came upon this also.
the bit i thought was good was those 1" blocks you mentioned on y slides supporting x rails because if you as a hobbyist had these cnc'd for you there wouldn't be much to cock up as far as alignment goes,you'd know x&y pretty good,flat sheet(poss thick) accross back between these and also on top and the x would be quite solid for low end machine.
thing is other than the z axis most everything else could come from marchant dice type people and be simple to align.plus if someone started selling 2 blocks like these which could i think be common to each other if you wanted a machine different sizes it'd be so easy to change as determined only by rod lengths really,you couldn't easily screw it up could you?
too much sag in rails, make 20 mm rail all round supported y rails actualy would be easy to do anyway.
wouldn't it be a great alternative to mdf for us newbies.
Never been able to see why marchant dice not do something like this,you can buy all the saddle blocks end end supports but why can't they publish more complete info as to how they all go together,you can by a complete twin round rail axis set but you still have to get the bits machined to take your leadscrew fittings etc and this could introduce errors.the set looks like cool vesatile axis fot 140 quid,but if you look what they charge for that axis with that bit of machining done it's like 500 quid.
as machining in volume brings prices down it makes sense for marchant dice type suppliers to get it done at source and charge a bit for it rather than us lot getting it done piecemeal.
Ross.

[Regnar]

Did anyone notice that the z axis is only supported by the bottom of the rails. Seems it would flex under load.

[marchantdice]

Hi,whilst looking around at different designs i came upon this also.
the bit i thought was good was those 1" blocks you mentioned on y slides supporting x rails because if you as a hobbyist had these cnc'd for you there wouldn't be much to cock up as far as alignment goes,you'd know x&y pretty good,flat sheet(poss thick) accross back between these and also on top and the x would be quite solid for low end machine.
thing is other than the z axis most everything else could come from marchant dice type people and be simple to align.plus if someone started selling 2 blocks like these which could i think be common to each other if you wanted a machine different sizes it'd be so easy to change as determined only by rod lengths really,you couldn't easily screw it up could you?
too much sag in rails, make 20 mm rail all round supported y rails actualy would be easy to do anyway.
wouldn't it be a great alternative to mdf for us newbies.
Never been able to see why marchant dice not do something like this,you can buy all the saddle blocks end end supports but why can't they publish more complete info as to how they all go together,you can by a complete twin round rail axis set but you still have to get the bits machined to take your leadscrew fittings etc and this could introduce errors.the set looks like cool vesatile axis fot 140 quid,but if you look what they charge for that axis with that bit of machining done it's like 500 quid.
as machining in volume brings prices down it makes sense for marchant dice type suppliers to get it done at source and charge a bit for it rather than us lot getting it done piecemeal.
Ross.

Ross, I think you answered your own question...
We offer quality with a warrantee
as for the price
£500 - 30% for our master distributors = £350
3.5 hours labour £157.50 = £192.50 less the price of the stock!
Regards / Kevin

[mad.sculpture]

Hi kevin,
ok, fairy nuff, i'm out of touch with hourly rates having done too many duff jobs lately.I'm used to what the local engineer used to charge my last company i worked for for one offs,but they were to become rolling orders in future so machinist prob. did them as lost leader.

At the time I wrote I couldn't imagine anybody ever using the components without having the holes for leadscrew through the saddle and endframes,specifications may vary but thought there would be a kind of 'common denominator' smallest size that at least could be opened up by customer if necessary.I've got a cheap ass drill press but wouldn't like to go through that size lump of alloy on it.
It's frustrating as a non mechanical engineering person who mainly want's to do cnc art on a low budget in a house i don't own(ie. can't stick workshop in garden)that you need to have a load of machining tools
and the wherewithall/wokshop to use them to be able to make the poduct that will then render many of them obsolete and make it a (relative)doddle to cut accurately.
I was thinking at the time to buy rockiffe plan and try to make it in alloy.
I then saw the rail assembly and thought i'd be able to make just about any machine i'd need for 3 * £140 ish for sliding elements exluding screws
then when i saw cost of machined ready to use item it burst my bubble because I realised it's going to be 3 times more expensive.
sometimes seems the only way doing something as a hobby bit by bit is cheaper is because you don't see the whole bill at one time.
Again it'll be down to my lack of knowledge about machining but I had no idea from what I perceived to be the necessary work that the unit would take 3.5 hours for a professional.

Sorry if I annoyed you,thankyou for reply

Regards Ross

[sansbury]

You could always look for a fellow hobbyist with a more developed shop setup to help out. What you're looking for is not complicated and doesn't even need CNC to get it right-just a good mill and tooling and a steady hand with a ruler. A hobbyist also doesn't have to cover rent, employee overhead, investments in tooling and equipment, advertising, etc. There might very well be somebody out there who'd be delighted to help for $100 or even a good bottle of Scotch.

[mad.sculpture]

yup, that's a good idea (wasn't sure if you were offering)
only just getting into this so not ready to ask ANYONE yet as no drawings.
wouldn't know at the moment how to specify the right tolerance that the bushes were snug fit and rail holes tight enough, although i have seen a few people slit the metal so a bolt can clamp like a bike seatpost,still leaves the y rail fitting though.

would the best way be to send one with the Scotch ?


i thought the Y end design on this was great as not got to buy 4 pillow blocks (X & Y) each end and align everything and was just sort of thing i was working on (if i don't just cop out and buy complete axis 2'nd hand on ebay)
The sort of simplified design where you think blimey where have all the other bits gone)
But lot's of people don't seem too keen on it so i'm out of spending mode and back to reading a bit more.
As far as x axis looking flimsy i thought that if some holes were tapped on top and back side of the uprights any muppet (ie.me) could just have some alloy lengths cut and drill/bolt them on.If the machine's not been seriously abused those slide rails could be 20mm would hold it all square as you bolted and that could be quite well braced then couldn't it? the alloy could be as thick as was necessary and you'd have a unit not dissimilar to a commercial robot cartesian axis by thk or whoever.

I realise these could be done without cnc but (naively i expect) that if they were done cnc everything would be square by default? and machine innacuracy would be down to tolerances/flex rather than builders lack of skill

to be honest there's a voice in the back of my head starting to advocate rapid prototyping for my needs (as in bureau) but it would be nice to work in house.

p.s always wary of asking someone on a DIY site to do something for $
since i saw someone on a computer graphics forum ask a favour and get the reply."yea ok mate, and after would you like to borrow my girlfriend"

Cheers ross

[sansbury]

The thing about this site that drives me a little crazy sometimes is the amount of overbuilding that is encouraged. I don't mean to single out ger21, but his comment that 'this machine is only good for cutting balsa and engraving' strikes me as being a bit of an overreach, at least without more context.

This isn't unique to the 'Zone by any means--I build audio equipment too, and over on diyaudio.com you'll find guys recommending that you use a power supply with 2-3x the capacity that you could possibly ever draw because "it provides better regulation" or some other sweeping generality they heard someone with an engineering degree say somewhere.

You'll see guys here working on wood routers talking about .005" or less precision. Go ahead--cut a piece of wood exactly 12" square and wait a few weeks, and measure it again. Bet you a bottle of malt it's off by more than .005"--if it's not plywood or MDF it could be off by a lot more! Not saying that higher precision isn't useful, but it's completely irrelevant in a lot of cases.

The design at the top of this thread does come with certain caveats. I suspect that sagging of the Y-axis rails is a real possibility, but how much? Say it's a whopping .05"--that would still work fine for cutting clean through a sheet of material. I'm not saying this is a good thing, just pointing out that this doesn't render the machine useless.

There are a lot of guys here who will tell you that any machine that isn't able to mill a 2" aluminum billet in a single pass to .001" at >100ipm might as well be a stone chisel lashed to a stick with a piece of rawhide. Fact is you've probably worked most of your life with lots of stuff that's .01" or worse and never realized how bad it was, because it DIDN'T MATTER for what you were using it for.

I really want to start a Sloppy CNC Anonymous thread so that all of us with our hacked-up machines can show off the things we've done with the piles of cheap junk we slapped together with duct tape.

[mad.sculpture]

i know what you mean..
but the y rails could be fully supported anyway if the uprights had the open bearing type without the core element(verticals) i liked being much different (just an extra slot across bottom isn't it),nobodys really posted an opinion as to wether this design really offers the apparent advantage of being adjustment free if those core bits are good.
it was the combination of x & y ends apparently requiring no real adjustment i wanted peoples tech. point of view on.


i can't comment too much really on the overbuilding as i've just wasted much ebay browsing and probably some dosh on ideas to harmonic drive a 4th axis before coming to the conclusion i'll probably buy a sherline one leadscrew one.

so far all my browsing of pdf data sheets has actualy got me in real terms is a pc virus !
don't panick - dealt with

rotary axis are mentioned all over place on these forums because there's a
4th axis? thread
5th axis? thread etc. so the info's a bit dispersed

s' all good though

my first mill if i persue it will probably be a mutant combination of ebay linear actuators...

i plan to call it Fugly pig - because it will be !

[sansbury]

i know what you mean..
but the y rails could be fully supported anyway if the uprights had the open bearing type without the core element(verticals) i liked being much different (just an extra slot across bottom isn't it),nobodys really posted an opinion as to wether this design really offers the apparent advantage of being adjustment free if those core bits are good.
it was the combination of x & y ends apparently requiring no real adjustment i wanted peoples tech. point of view on.

Here's the catch: no-adjustment designs are theoretically easier to set up, but you essentially get only one chance to get it right, and if anything changes down the road, you're probably screwed. Adjustments aren't a get-out-of-jail-free card in all cases, but they are in many, so you see them on pretty much every DIY design. This is how you get better than .01" accuracy on a machine where most parts were built to tolerances of 1/16-1/32". Adjustability does add some cost and complexity to the build, but the benefits are worthwhile.

i can't comment too much really on the overbuilding as i've just wasted much ebay browsing and probably some dosh on ideas to harmonic drive a 4th axis before coming to the conclusion i'll probably buy a sherline one leadscrew one.

If you can afford to buy ready-made and your main goal is to build things with the machine, then by all means do so. Building these things is fun and educational, but no more than that.

rotary axis are mentioned all over place on these forums because there's a
4th axis? thread
5th axis? thread etc. so the info's a bit dispersed

The challenge with 4+ axis machines is primarily in the software to generate the toolpaths. Mach and EMC will happily drives 6 axes, but creating G-code that really makes use of the additional axes is still largely the domain of higher-cost professional CAM tools, or hobbyists with excellent math and programming skills. As a result, relatively few people build ABC axes because they can't drive them, and since no one has ABC axes, no one writes free/cheap software to use them. Hobby CNC is still young though, and I imagine in 5 or so years there will be a lot more stuff out there.

[mad.sculpture]

thanx for replys,

with regard to the 1st comment about adjustments giving the advantages of high accuracy ,
cheers for that as makes a bit more sense now
But - if the parts were cnc'd properly, so i do everything else as best i can,would it still be necessary to have adjustment ? - because i haven't built the key axis parts to 1/64'th they'd be cnc spec.
also being able to compensate for stuff down the line,you mean wear on parts or something else i can't forsee me yet?

i know nothing yet of G code but hoping to have a chance of wrapping my head around it as many years ago i was a software bod.(programmer)
on the Rainnea Graphics website there's some info of how they generate the toolpaths from some packages they use and i'm pretty sure it interfaces with silo cad package i use so ever hopefull of driving like them.
they point out that as they're shareware versions they are limited to certain size of Gcode export(so many 1000 lines) but even then they tell you it's possible to export in sections and paste together after so workaround available.

also there's a guy doing cool stuff on the art forum(old postings)who also gives methods, he's using lathe as 4 th axis,presumeably with stepper as indexer but supose he couold have just continuous very slow synchronised rotation.

if it seems over the top to go 4 axis as soon as possible, it's because i want to machine busts of creatures/dinosaurs organic shapes.event if i have to do every tooth and horn separately it'll help when i try to do bigger stuff.
as the two previous people i mentioned have and it will be a lot more practical to do in one main piece.only got (hopefully it arrives)6" z axis at moment but hoping the pig on a spit method will give me parts up to 12" dia

I'd like to be able to rough a plasticene model out,
scan it in with (laser davidlaserscanner) cheap, apparently works well
mirror it in software,modify and detail it
mill it in foam (or whatever) and hardcoat it or if millings not so good then mill a slightly smaller shape so i can add small clay layer to detail but hope to avoid this.
mould and sell to creature fans as kits

so far i can do 1st stage and 3rd or just build direct in software so laser'n not crucial but so far only got
z axis 6" coming in post with servo + possible y option(13" robust slide) awaiting reply ebay + spacecraft spec.(really ovebuild) servo/harmonic drive (that's prob not going to work) but no amp yet for it (so it's just a posh paperweight at moment)

+ i've done lots of Radio Control type servo pc interfacing from scratch in the past.

it's a lot of varying disciplines & i find it totally compelling but the motivation varys like the weather just because it's taking so long.
it wastes a lot of time scouring ebay for the bits,then you need to track the data sheet for what is often a legacy part to find the ratings/bearing type etc.
plus as i'm in the uk postage is a big issue in getting bits,not just money but lots of people only ship internal america.
or sell 3ft perfect Y axis for $61......$250 to england by their carrier of choice to blighty!
Someone's just said actualy they'll forward stuff for me but not sure she's talking about a bridgeport.

at least as far as software goes it's only getting easier as the tools improve and if you know where to look you can usually get your puds on them.

rapid prototyping is getting affordable for mortals now (action figure size) but they're not going to be producing any trex parts any time soon for my money.

I'm going to have to write less on the forums actually, or train more typing fingers.

Cheers ross

[mad.sculpture]

p.s. i couldn't possibly condone using it but there's a website called gfxworld that has lots of good cad links insofar as drawing and modelling,possibly not machine control
- just so you can see whats available