Thanks... it's not going to rust where it is so I don't think I'll paint it. It's there to make things not look at :smile: Paint would also get annoying if I need to add bits to it.
Printable View
Jonathan
Reading more and more post I find my self thinking o's##T what have I let my self in for
I really thought 20 mm rails for the bed would be more than ample to carry the weight of the whole cnc machine
two 20 mm rails (made from 304 stainless steel) and four bearings mounted on the 90 degree side of 40/80 profile aluminium extrusion
and a centre ball screw (centre of the bed) 0f 16 to 20 threaded bar
where on earth do you all get your aluminium from as I have ordered of ebay, it arrived today so bent it is unusable
James & Luke
They will be ample to carry the weight since each of the 20mm bearings is rated for about 880N. That's not as excessive as it might seems though, see what I posted here:
http://www.mycncuk.com/forums/showth...near-bearings-...
The reason I advised using the bigger bearings is not directly due to the higher load ratings. It it because they have 5 rows of ball bearings, not 4. This seems to make them have a lot less play than the <=20mm bearings. So I'm not saying you have to get the 25mm bearings, just that it will probably be better if you do.
I now use http://www.aluminiumwarehouse.co.uk/
Only issue with them is you have to buy 2000mm lengths. If there's something smaller you need let me know as I might have a bit...
What size did you buy out of interest, and who from so I don't make the same mistake?
1000th post :eek:
So here's a video showing my router doing 1000mm/s (2362ipm) on the Y-axis:
http://www.youtube.com/watch?v=k3MG7...=youtube_gdata
All I did was change the pulley to 42T so the pitch of the screw is effectively 35mm. It's well above the critical speed of the screw, but interestingly it seems fine.
The video also shows the rotating ballnuts going happily at 15000mm/min.
I had a request for some more photos, so here they are in no particular order:
Attachment 4572
Attachment 4573
Attachment 4574
Attachment 4575
Attachment 4577
Attachment 4578
Hi,
Jonathan what have you used on your Y axis, I can't quite make it out
Yep:
It's mounted with 20mm aluminium plate milled to fit in both ends. It's adequate, but I would advise using something a bit stronger. Since the rails are on both sides of the box section it had to be aluminium since the width of steel box section is not accurate enough. Extrusion is a good option, but quite expensive.
The first of many upgrades to come:
Attachment 6825Attachment 6826Attachment 6827
I'll be fixing 20 meters of 2"x1/2" aluminium strip under the bed in a grid to make it a bit stronger, since it's only 20mm thick aluminium. I've got 6m of the aluminium angle to go round the perimeter to enclose flood coolant, which will be the next upgrade. I'll have the router drill an array of holes to tap M8 for clamping. I elected not to go for strips (to imitate T-slots) as the difference in price is not that great in the whole scheme of things and this should be stronger although either would have been sufficient. Need to get it at the right height too...
That looks like an expensive piece of Ali make sure you don't Fubar .!! At least with strips if you do then it's a cheap repair.!
After fitting the strips are you going to fly cut it.?
Ascmetals are to blame for the good price. Not that bothered about marks to be honest - if I do mess up it's unlikely to cause more than aesthetic damage, which luckily I don't care about!
Not sure yet since the issue is I can't reach the full area with the spindle in one go. Once they're fitted I'll height map it and see how close it is. I can always put some aluminium 'sacrificial' strips on top (which could even make T-slots), but I think I've spent enough.
90% sure you will.? . . . When I did mine I height mapped it with 45mm grid and was surprised at the differences.!! Thou mine was an easy surface has I only had to do the strips.
I had the same problem regards cutting the whole area but actually it works out good because the shoulder it leaves acts has a parallel reference edge for X & Y.
I tend to a agree with the 90%...
I reckon if the spindle on there, or the one I made, can support a big enough fly cutter I can easily get over both sides by moving the spindle to either side of the Z-axis especially since 2" wide around the border will be occupied by the aluminium angle unless I mill that down. The problem then is maintaining an accurate height between passes, but that shouldn't be too bad. On the previous bed it was very handy having the ridge left from surfacing for a parallel reference
Thanks, but it's not that great really...quite a few things I'll be changing as the gantry is not strong enough for what I want to do next.
Yes, I mounted it and it worked well with low backlash:
http://www.mycncuk.com/forums/router...html#post16838
It's clearly not as good as a real ball-screw, but as something to get started it's excellent due to the low backlash and very low friction.
I cut the first piece with the new bed last night - just a pocket in the back of an aluminium motor mount and the finish was much better than with the [edit] MDF bed and I didn't need to do a finishing pass to get that.
I used it for quite a long time, until changing to an RM1610 ballscrew. It was working well, although had caused some wear to the M16 rod (although not enough to be a problem) due to bearing inner ring being significantly harder than the screw, however I think if you used springs to push the angled bearing against the screw and/or used a trapezoidal lead-screw instead then it would last a long time.
Yes, thanks for pointing that out - fixed.
What size fly cutter do you use for surfacing yours?
I'm cutting the aluminium angle now to fit round the sides for the coolant barrier, then once I've got the bed aligned I'll get the router to cut a suitable hole in the bed at one end for the coolant to flow through and make a filter etc...
I seem to be able to cut 3mm per pass with the 6mm single flute cutter now :)
Not really, anyone can put a big pulley on the stepper motor and lower the acceleration to get a high feedrate at the expense of resolution. Such a high feedrate would only make a negligible difference to machining time on a router used mainly for cutting metals. It's just for show...
Jonathan,
A quick question about your "box" type Y-carriage in this pic please... Particularly about the position in which the horizontal plates are bolted to the vertical plates....
The top plate is bolted slightly above the top of the vertical plates and the bottom plate is bolted slightly below the bottom of the vertical plates... If that makes sense.
Did you do this for a specific reason such as making the initial DIY build/squaring easier?
Cheers,
Andy
Yeah I ordered the wrong size aluminium plate!
To make it easier to square you should mill a shallow slot into the back of the Y-carriage plate which holds the Z-axis bearings. This can then align to one of the smaller plates on the Y-axis to get Z perpendicular to Y. You can just about see what I mean in this picture, circled in red:
Attachment 7064
Which reminds me I never posted the pictures from when I changed the Z round rails to linear guides:
Attachment 7065Attachment 7061Attachment 7062Attachment 7063
Using the indicator made aligning the rails properly and getting it to run smoothly trivial. Fitting everything in the small available space was less trivial. Hard to say how much difference it made to the performance of the machine, since at the time the MDF bed was the limiting factor and I think my spindle might be on it's last elbows...
I finally had chance to drill the array of M8 holes for clamping in my router bed. Here's the result:
Attachment 7457Attachment 7458Attachment 7459
I used a carbide drill since the spindle speed required for a 6.8mm HSS drill is too low for the spindle to have enough torque to cut efficiently, if at all. I ran the drill at 9500rpm and 840mm/min with flood coolant, which according to the manufacturer's recommendations is quite conservative, but still plenty fast enough. I didn't see any point pushing it harder as that would only increase wear on the spindle bearings.
Here's a video showing the process. Possibly the most repetitive video I've uploaded to youtube so far...
http://www.youtube.com/watch?v=gFLwSYARcUU&feature=plcp
The Z-axis motor, not surprisingly, got pretty toasty but other than that it went well. I had to move the bed out to reach the whole length, so it's currently 360mm from the normal position. Just need to make the router cut a drain in it, then I'll put it back.
Forgot to say in the previous post...
Around the edge of the aluminium bed I've put 3x2", 1/4" thick aluminium angle. This acts as a barrier for the coolant, so to ensure it seals I put a strip of 1.5mm thick neoprene under each piece. I used my metal cutting bandsaw to cut the 45° angles, but the finish from this although not bad for a bandsaw, is still quite rough.
Attachment 7469Attachment 7470Attachment 7471
Bit of an unusual set-up on the lathe so I videoed it:
http://www.youtube.com/watch?v=Xj1AYb41dQs&feature=plcp
(Something wrong with the sound on the past two videos)
Hello, looks like you've got a fine machine there Jonathan! Good work, It inspired me to go the way with the bed-box-gantry type mill, if I get started with one on my own.
Usage: pcb, wood, plastic engraving, possibly some easy metals, maybe 3d printing?, 3d scanning? Bigger polystyrene works. I like the idea of the adjustable bed height for different jobs. I already have three industrial dc motors+encoders and two ballscrews I was planning to use. I was thinking of a machine capable of moving 800*700*300mm.
A couple of questions came to mind:
1. Where do you get all parts? Is there a standard online place with good prices/service? Guides, ballscrew, bearings. Saw that Hong kong was mentioned.
2. I was thinking of moving the gantry with two rotating ballscrews and one motor. Long
gearbelt to sync the screws and have the motor directly on one screw. Whats your thought about this? I think I'm ok with a 1/10 or 1/20 mm milling accuracy. (I don't think I can build the complete works with greater tolerance)
Cheers
/Ola
So just about everything!
That size is very manageable, although you will have to make it quite strong to make good use of the large Z-travel. Are you sure you need that much? I guess it's for the 'bigger polystyrene works', in which case since you will use an adjustable height bed, the required rigidity is lower.
The nearest you'll find to a cheap standard online place is the seller linearmotionbearings2008 on eBay. Many people on this forum have used him, myself included, and been pleased with the products and service. It works out cheaper to get the spindle from other eBay sellers, or even better ali-express.
6 of one half a dozen..
If you use two motors the available torque is greater, so you should get higher acceleration and top speed, however it will work out a bit more expensive than one motor and a long belt. I don't consider one motor stalling and the gantry racking a valid concern since if your machine is tuned properly, and you shouldn't be running it if it isn't, then that should never happen. On my machine I had little choice since with the current gantry design it's impractical to route a timing belt to link both rotating nuts.
Not difficult.
In other news, I have made and installed the drain for the router bed so I've been cutting with flood coolant. I fount that one corner of the machine is ever so slightly lower than the rest, so I put the drain in that corner so the coolant flows. A week later I discovered the reason for this corner being lower - there's a crack in the floor and it's sunk down!
The drain is fairly basic, just two parts I made on the lathe to hold some fine mesh, which is easily replaced:
Attachment 7571Attachment 7572Attachment 7573Attachment 7574Attachment 7575
After running for some time:
Attachment 7576
Video of cutting with flood coolant:
http://www.youtube.com/watch?v=gkGf5c9VQyk&feature=youtube_gdata
Did you source or mill the aluminum support for the guide rail?
Thanks, Jeromy