Wobblycogs CNC Mk1

[Wobblycogs]

Yes, I decided to increase the travel. I went and re-measured the workshop and I came to the conclusion that I had to rip out a length of workbench and cupboards if I was going to reasonably fit the CNC in there. I figured I might as well use up all the space I'd created so upped the length to 1800mm (poor excuse I know). Since then I've thought up a couple of projects that would use the extra bit of length. I'd like to get a bearing spacing of at least 250mm on the gantry so I'm guessing a bit over 1600 for the workable length of X.

[Jonathan]

Ok fair enough, if you want 1600mm travel then the discussion is valid :)

Option 3 is to find out how much RM2010 ballscrews cost:
Aliexpress.com : Buy Anti Backlash Ballscrews RM 2010 L1000mm*2pcs with 2pcs SFU2010 ballnut for CNC Machine from Reliable Ball screw suppliers on BST AUTOMATION

If you wait for the seller to be online or just send them a message I'm sure they'll give you a price for RM2010-1800mm (or whatever). They also have RM1616 and RM2020, which would also be good since then the critical speed is no longer an issue (you can gain back the resolution via the pulleys):

http://www.aliexpress.com/store/3147...Text=high+lead

[Wobblycogs]

Right then, after much dithering I've decided to go with the 2510 screws and some nema 34's. I've not picked the motors and drivers yet but looking at what's been recommended I can clearly get something that will work in a 34 package so it's time to press on with the design :-)

Since I'm using 50mm box for the frame I can't directly mount the screw bearing block onto the frame. I'm hoping to get around that by mounting the bearing block and motor on a plate (green) which then bolts onto the frame. The picture below shows the business end of the screw but there's a similar plate at the other end as well. Good idea / bad idea?

The plate is currently drawn at 100x300x5mm as it was going to be steel but I'm now thinking aluminium would be better as it would be easier to work (I want to provide some vertical adjustment holes for example). Will 5mm aluminium be strong enough?

The motor housing is drawn in 20mm aluminium which looks far thicker than I need but I have very little experience to go by. Would 16mm or even 12mm be strong enough? I'm going to drill and tap the edges to join the pieces together so it can't be too narrow :-). Ideally I'd like to use the same thickness plate for parts on the gantry and Z-axis so that I can just buy one large piece of plate which, fingers crossed, will be cheaper.

Finally, what grade of aluminium should I go for? The motor housing and probably a few other bits will need to be milled out; I've got decent variable speed router so I was going to give it a go with that (taking very light cuts and with a template). I seem to remember someone saying they were using C250 but it looks like 6082 is easier to machine from what I've read.

Cheers

[D.C.]

If you are thinking of getting some of the leadshine drives in the next couple of months:

http://www.mycncuk.com/forums/marketplace-discussion/5289-leadshine-digital-drives-bulk-purchase.html

[Jonathan]

I contacted the supplier I linked to in my previous post to find their prices for the RM2010 ballscrews. He replied promptly with the following:


For:
RM 2010 -L1800mm*2pcs with 2pcs SFU2010 ballnut with end machining
Best price for you is USD360.00,include the freight by Fedex/
For:
RM 2010 -L1500mm*2pcs with 2pcs SFU2010 ballnut with end machining
Best price for you is USD325.00.


Clearly that's $121 (£76) more than the RM2510 from Chai, but with 2010 you could still use the 3Nm steppers and drivers and get better acceleration than with the 2510 ballscrews. The 5mm difference in diameter may not seem like much, but remember the torque required to accelerate a cylinder is proportional to the radius to the power 4, so 20mm needs only around 41% of the torque that 2510 requires.
The reason none of us mentioned RM2010 earlier was because Chai doesn't stock them and they are very expensive from England, but these prices change that even though they are a bit more expensive.

Since I'm using 50mm box for the frame I can't directly mount the screw bearing block onto the frame. I'm hoping to get around that by mounting the bearing block and motor on a plate (green) which then bolts onto the frame. The picture below shows the business end of the screw but there's a similar plate at the other end as well. Good idea / bad idea?

Good plan as it makes it easier to align everything since you can have adjustment via slotted holes in the plate.

I'm now thinking aluminium would be better as it would be easier to work (I want to provide some vertical adjustment holes for example). Will 5mm aluminium be strong enough?

No I don't think so, I'd say 10mm based on what looks about right!

The motor housing is drawn in 20mm aluminium which looks far thicker than I need but I have very little experience to go by. Would 16mm or even 12mm be strong enough? I'm going to drill and tap the edges to join the pieces together so it can't be too narrow :-). Ideally I'd like to use the same thickness plate for parts on the gantry and Z-axis so that I can just buy one large piece of plate which, fingers crossed, will be cheaper.

20mm is excessive, but as you say it's sometimes best to keep things the same thickness to make it more efficient when you buy the material. Have you put a pocket in the mount for the motor to rest in? It will make it easier to align the pulleys since you'll have more of the shaft extending through the aluminium, e.g:



6082 aluminium is readily available and cuts easily, so that's fine.

[Wobblycogs]

Cheers, I was reading Iwant1's build thread yesterday realized I was trying to rush this project and that something was going to go wrong if I didn't slow down, do a bit more homework and spend longer on the design. I really wanted to get this done in the time I have off but it looks like that was more than a little optimistic, perhaps it will be ready for the end of next year :-)

As I've now got a lot more time I'll seriously consider using the 2010's instead. Over all I think the cost is going to be about the same but better acceleration would be nice to have. I have to admit I hadn't realized how big the 34 motors were until I imported one into the model. It's a chunky bit of kit compared to the 23 package.

Based on your advice in another thread I've moved the screw up to be level with the rail as well as increased the plate to 10mm.

The motor housing has been dropped to 15mm plate. It already has a cut out to allow better shaft protrusion but I've realized I'll need to slot the cut out and mounting holes to provide tensioning capability.


I think the frame design is pretty sound so I'll aim to get that built over Christmas (I've got to build something or I'll go stir crazy) and then work on the rest of the design in parallel.

[Wobblycogs]

I've done a lot of reading of build logs in the last week and a fair bit of tinkering in sketchup to (hopefully) improve the design. I thought it was about time to do an update and to find out if I'm going in even vaguely the right direction.

First off we have the gantry connection to the rails and x-screws. I noticed a lot of people were advised to make this connection stronger so hopefully this will be strong enough.

The block that attaches to the ball nut is a 50mm lump of aluminium. The green bolts are M8x90 and go most of the way through. The blue bolts are M6x30 and screw in to the block between 15 and 20mm. After seeing someone else's build I've gone for a two plate gantry design as I think it'll be simpler in the long run. There will be slotted screws holding the two plates together so fingers crossed I should be able to get the screw and gantry connections aligned. I'm not keen on the three M6 bolts close together on the top plate but I think they've got to stay. They are there to counter any torque applied by the green bolts and face plate if things aren't 100% square.

I've slightly redesigned the bed to make it simpler to build and to move the legs to be directly under the x-rail. It's shown bolted together but I've decided to weld at least some of the joints so that I get it done this century.

Now here's where I'm going a little off-piste with the gantry I think. I read the recent thread discussing gantry designs and liked the L shaped solution made from 45x90 extrusion. I also liked the C shaped designs so I thought why not try and combine the best of both worlds and use some 90x90 heavy extrusion (this is now in the sketchup library if anyone else wants it). I'm guessing this will give me both vertical and horizontal strength and the extra bit of mass should help damp any vibrations a bit.

I'm currently pondering running the y-rails along the front of the gantry rather than top and bottom as most people do. Although this will cause the Z-axis to stick out about 50mm more than the top and bottom rail design it will also allow me to maximize my y-cutting distance and have a slightly wider y-bearing spacing. The gantry side plate will be braced front and back and reduced in size to only what is required as the design progresses.



One thing that is starting to concern me a little is cutting all this aluminium. I have a good selection of wood working tools but not much in the way of metal working tools. Do you think it would be possible to hand route 20mm aluminium? From what I've read slower bit rotation is better, the slowest my router will go is 10,000rpm which seems to be slower than most so I think I'm in with a shot but some advice would be much appreciated. Cheers :-)

[Wobblycogs]

I've managed to get over my stupidly large gantry design by totalling up the cost of the aluminium plate and extrusion and nearly passing out. My new version 5 design uses KJN 45x90 profile like a lot of builds I see around these parts. What I'm struggling with now though is how to mount the SBR20 to the extrusion.

In the image above I've centred the rail but it doesn't matter where I put it the rail lines up really poorly with the extrusion. Has anyone managed to find a good solution to this? I'm so tempted to just give up and use profiled rails but the extra cost is off putting to say the least.

A couple more shots of the current gantry design, comments welcome.

The gantry over hangs the back axis bearing a little so that I have space to mount the stepper. The over hanging weight is fairly minimal so I think it should be ok. The stepper will be slotted so that it can move up and down to tension the belt.

I went with the taller C-shaped gantry design because I want a good separation between my Z-axis bearings. The Z travel is about 300mm (more than I was aiming for) and I'm hoping that extra bit of height in the C-shape will help stiffen things up a bit.

[WandrinAndy]

What I'm struggling with now though is how to mount the SBR20 to the extrusion. In the image above I've centred the rail but it doesn't matter where I put it the rail lines up really poorly with the extrusion. Has anyone managed to find a good solution to this?

I followed a similar thought process Wobbly... I've seen folk suggest bolting thick flat bar to the extrusion and then in turn attaching the SBR20 to the flat bar, but IMO that partly negates one of the main benefits of using extrusion... relative ease of assembly.
Also just a quick caveat that the SketchUp 45x90 component... certainly if you are using one of mine, sorry... may not necessarily be accurate enough for this kind of accurate line-up exercise.

I'm so tempted to just give up and use profiled rails but the extra cost is off putting to say the least.

Agree, the costs are horrendous. Although it looks like Zapp is now also offering, in addition to the Hiwin range, a value for money range of linear precision profile stuff.... LSK or summat, that I think is about 25% cheaper than Hiwin?

I'm finding that there are sooooo many tradeoffs in any DIY CNC design.... That it's worthwhile setting aside plenty more time than initially anticipated :-)

[Wobblycogs]

I followed a similar thought process Wobbly... I've seen folk suggest bolting thick flat bar to the extrusion and then in turn attaching the SBR20 to the flat bar, but IMO that partly negates one of the main benefits of using extrusion... relative ease of assembly.
Also just a quick caveat that the SketchUp 45x90 component... certainly if you are using one of mine, sorry... may not necessarily be accurate enough for this kind of accurate line-up exercise.

Attaching a plate or bar to the extrusion was all I'd come up with as well so at least I'm thinking along the same lines as others. I agree though it seems to defeat the purpose of using extrusion. I'm using the MacTavish 45x90 component. It seems to be right but I'll double check it.

I'm seriously tempted to swap out the extrusion for some box section aluminium with a good thick wall. I can't help feeling that three pieces of box arranged as I currently have the extrusion would be equally strong. I'd almost certainly then fill the middle one with kiln dried sand to damp vibrations. The problem then moves to joining the box section to the gantry sides but that feels like a fairly easy problem - I feel version 6 coming on...

I was looking at the Zapp rails yesterday. The rails aren't to bad price wise it's the carriages that are the killers. I'm trying to convince myself that I'm only even going to build one CNC so I might as well do it right the first time but I just know I'll end up building another :-)

I've found that there are sooooo many tradeoffs in any DIY CNC design.... That it's worthwhile setting aside plenty more time than you initially anticipated ;-)

Never a truer word spoken.