Giant 20' x 3' x 2' router build

Just started building 3 aixis cnc aprox 20' x 3' x 2' I am building it from 80mm aluminium extrusion and using a 20mm hywin rail system on all 3 axis with rack and pinion on x and y and a ball screw on the Z. It is a cantilever arm style machine and will mainly be used for cutting foam and other reactively soft materials.

The aluminium i have purchased from KJN and the motion control and linear movement parts are being supplied by CNC4YOU who have been exceedingly helpful and a great bunch of guys.

It is still at very early stages but i do have a few pics of the frame coming together which i will post a little latter. I look forward to any feed back or advice.

heres the first few pics

the frame is almost together although it was pretty expensive i think the ease of construction and accuracy has more than paid for itself plus it would be a lot easier to move if i ever need to.

As no other replies yet just thought I'd say thanks for posting and please keep it up - watching with interest - should be an awesome huge machine when done :-)

no worries, the frame is all locked in solid and started on the gantry. just figuring out bolting the rails down. From what i have read on the hiwin site for the RG20 rail its saying M5 bolts which appear to be a little small in the rail holes but i gees that leaves room for fine adjustment? is bolt them down every 180mm sufficient or do i need to do it more like every 120mm?

Wow Charlie what a build this is going to be!

The rails come pre-drilled so you would want to bolt them down using those fixing distances, I dont want to say to much because I dont know what you have planed but it reads as though you are going to bolt the rail directly to the extrusion?

Edit: Just re-read your first post "soft materials".

.Me

the rails are pre drilled every 60 mm but have been told that i could probably bolt them less hence possibly going less. The hardest thing i will be machining will be ply wood but mainly 90% of the work will be foam. The rail is going directly on the extrusion which is 80X80 ali and using the centring t blocks. More than happy to listen to any other suggestions, i am basically just going by what i have seen on similar machines that have been built with the same intended use.

many thanks charlie

Presumably the rails are drilled at the standard 60mm spacing... guess that could get tedious over 20 feet/6m - approx 100 per rail... but only needs to be done once. Guess you might get away with every other hole (120 mm) but doing them all would add a little to rigidity I would think...

Did you find the Hiwin rail in 6 m lengths BTW or are you going to have to butt two lengths up together? Longest I've seen is 4 m at e.g. WMH Transmissions Ltd - Product List - HGR 20 GUIDE RAIL<BR><BR>PITCH 60<BR><BR>NORMAL GRADE

no the max length is 4 meters so would have to have a joint , the cheapest i could get was a max length of 2.7m so would need 2 joints which i have ended up with however using 3 two meter lengths for each rail. We have spoken with hiwin and they have said it is fine to just but them up together without anything else.

Would definitely bolt every 60 mm then as the last thing you want is anything moving once you've got all the joints perfectly aligned.

cheers thanks for the advice best order a stack load more slide blocks then mounting is no sweat as i can put them on to the rail and then just slide it strait on to the aluminium. Next job i think is to make sure it is all perfectly level on the floor and get a plate made up to bolt the carriages too. If i can get the plates laser or water jet cut do you think i can rely on there accuracy to get the two carriages to aline or do i need to factor in adjustment with a oval hole or something?

many thanks again

been working on modifying the design a little bit to try and simplify the build as its my first. I have just got the parts to build the z axis however annoyingly the mounting holes on the bearing blocks are not spaced the same as the ali extrusion so i am assuming the best way to solve this is by having some plates made up to go between and allow me to bolt one to the other?

This is not unusual. I think you are suggesting adding a plate under the bearing block to bolt to, which is in turn bolted to the extrusion rails. This is fine, but will raise your ballscrew by the thickness of the plate - is that a problem?

If it is a problem, these blocks normally have both vertical holes and horizontal holes for mounting. So instead you can use the horizontal holes to bolt into a thick plate/block which is next to the bearing block, and is in turn then bolted vertically down onto the extrusion. The trade-off this time is a small loss of ballscrew travel equal to the thickness of the block/plate.

If none of these options work, you can drill and tap straight into the face of the extrusion (i.e not using the rails) - but you will need to check if there is enough material to tap into and you avoid the slots.

yes that was what i had in mind.They do have horizontal holes but i am going to have to get the parts made for me so flat plates may be easier. i was wondering if i could bolt it directly but the holes are right on the edge of the slots so it would only be going properly through the very middle of the extrushion.

i have not chosen any steppers yet but was just wondering if i was going to need a reduction with belts and pulley that i plan on the top or wether normally it is just 1 to 1. the plan is to use a 2.2kw spindle and then there will be the weight of what is in the picture and what ever stepper i use.

..............

a few more pics

...........

Charlie,

It's a very unusual machine and I can't quite work out which bit will be which, so the question you've asked about motor sizes is probably proving difficult for anyone to answer.

But be aware that you can move a fair amount of weight with a 3Nm Nema 23 if you give it enough voltage as the ballscrew gives you lots of mechanical advantage. Don't rush out and buy 34's until you posted the final design sketch of the finished moving parts as you'll get a more definitive answer.

Your earlier question about pulley ratio on Z is also tricky since your Z appears to be quite long. Lots of designs are just 1:1 on a 1605 ballscrew for Z. But your travel (if I read the design right!) is quite long so this mught not be your best option. Again, a picture of the overall finished design would get a better answer from the forum on that question.

The wooden template motor bracket looks OK as a concept (assume ~20mm ali in the finished design?). Probably get away without triangulating back to the extrusion. But you'll probably need to recess the underside of the motor mounting area to raise the stepper and leave enough shaft sticking out the top. Also the ballscrew end might also need some pocketing on the top to clear the pulley / grub screw access.

I dont have any 3d drawings at the moment i tried to get sketchup to load on my mac but it didnt like it.

If you take a look at thins video it shows something similar to what i am building however i just have the flat table and i have decided to have the two beams coming out on the gantry rather than one so as to stiffen it up and split the two rails apart, then the z axis will go inside the square frame you see on the top in my picture.

the z is quite long i wanted to get as much as posable to be able to work on large block of poly styrene and by large i mean i buy it in 16' x 4' x 2' blocks so should i have aprox 500mm of z travel when done.

the plan is to get all axis bolted up and rolling before getting the motors which hopefully wont be too long, im hoping to get the plates for all the carriages this week. I wasn't to sure how thick the motor mount for the z was going to need to be, i was thinking only 10mm as i didnt think there would be to much load on it using a ball screw but happy to take advice on that and as you say ether recess the some of the areas to get the pulleys on or extend the shaft if that is posable?

many thanks Charlie

That makes more sense now in terms of the final design. Should have guessed with the O'Neil logo on the wall.

I think 10mm motor plate would also be OK stiffness-wise. What you will need to watch out for is if the bolt heads hit the belt then they will also need to be recessed. M8 cap heads would need 8mm depth of pocket so 10mm plate is then a bit thin.

In terms of recess or pockets for the motor - if you can find the drawings for the motors you might use you can check the length of the shaft. They are not usually that long plus you need access to the grub screw at the base of the pulley so that often means a pocket is required.

The motor plate extends past the extrusion and over the ballscrew - is that bit required? Are you hoping to bolt it to the bearing block as well? If not, you could cut it short and give more space for the pulley on the ballscrew. Just some thoughts . . .

cheers many thanks for the reply, yes one of the things i will be cutting is stand up paddle boards but hope to use it for ply would and other things which is why i wanted the table. I see the problem about the bolts, i will have a look in to that one i would have thought i can come up with something there. I have just been trying to keep all the bits i am having to get fabricated as simple as posable as i only have a basic lathe and no milling capabilities so just getting it all laser cut.

i was wondering about using the block to bolt to as well there probably is no need really and possibly not posable as there is all so a plate having to go under it. If i actually could find a large bit of 90deg angle aluminium i could possibly use that for the motor mount and bolt it straight to the side of the z axis rather than the end.

many thanks for the advice! charlie

something like this but would have to see what size motor i need and if it would fit.

Aluminium Angle from 2.5" to 100mm many thicknesses and lengths cutting service | eBay


sorry the link didnt properly work they do a 9.5mm thick by 76.2mm deep option

Quote:

---

Originally Posted by charlieuk

cheers many thanks for the reply, yes one of the things i will be cutting is stand up paddle boards but hope to use it for ply would and other things

---

If you want to cut plywood at all seriously, then I'd be concerned about the strength of your cantilevered gantry design. The cutting force for polystyrene is practically zero, so you can get away with a very weak gantry, but that's not the case for plywood.

Also, for the axis on the gantry (Y-axis?) your bearings are quite close together on the rails compared to the spacing of the rails, so I would be inclined to drive it with one ballscrew on each side, or at the very least make sure the ballscrew is centred between the two rails. The problem with the former is the ballscrew would intersect the Z-axis, so there may not be space for it. If you put a single ballscrew on one side the bearings on the other side will deflect more (i.e. it will rack like a typical CNC router gantry), so the obvious solution is to use two ballscrews. That's quite expensive though so you may want to just try it and see, but keep in mind the possibility of adding another. You could also increase the bearing spacing. It's hard to say for certain as it's difficult to discern which parts of the machine have the lowest stiffness and thus which parts should be improved if you want to cut plywood efficiently.

Agree with Jonathan about being better to use twin screws for Y axis and to some degree that Z axis will be weak for cutting plywood but with just a simple change it could be made stronger to the point it would manage be ok.
It would cost more due to extra components and ballscrew placement wouldn't be the ideal but would stiffen things up enough. I've quickly thrown together a model to show you what I mean.
Obviously I don't know the spacing etc in side the Y axis frame or how you plan to attach the bearings and what spacing but looking at your pics I think there'd be enough room.? Spindle is just there for affect and the real thing would need extra support.

thanks for that jazz i can see what you are saying and we did wonder about the third set of blocks and there is plenty of room to fit it all in and i have made the y frame so i can slide one side in to be what ever size it needs.

i was wondering about using 60x60 or 80x80 for the z axis but the 40x80 in a relatively short length seams crazy strong as it is.

i guess what i will do it try and get the x and y rolling and see how that feels and with the way it is i can pretty much just add to it if does look like its not going to be up to the job. I was planing on using rack and pinion for y could i just run a second motor on the other side rather than a ball screw? or possibly even one motor with a shaft and pinion on ether end like a car axel, although that maybe a little more tricky to build

if needed i could actually make the hole gantry narrower if needed just buy shortening a few bits if it would help.


I should say that i am not after crazy speeds for cutting ply or accuracy for that matter as it will be just for stuff that i would normally make by hand so 1-2mm tolerance would be fine to start with.

many thanks for all the advice i am taking it all on board

Quote:

---

Originally Posted by charlieuk

i was wondering about using 60x60 or 80x80 for the z axis but the 40x80 in a relatively short length seams crazy strong as it is.

---

Yes it may seem strong Now but promise you when there's a spindle on the end with cutting forces acting on it then at full extension at this length which you would be while cutting ply it will twang like a guitar string.!!
The Z axis is THE single most important area on a CNC machine and seen as your going to so much trouble and expense using Profiled linear rails etc why spoil the broth for a pinch of salt.?

Regards R&P then yes it would work ok and seen as your using it on other axis makes sense but you would still need drive it from both sides, thou it would be easy enough to run a shaft across the back of profile. As would it to connect ballscrews with a timing belt, which woud be my first choice.

Tip for you if you use the Bosch Rexroth 45 x90 or it's equivlent then the BK/BF12 bearing block hole centres match the slots and make fastening very easy.
This stuff Profile 45x90S slot 10

What is the biggest diameter collet you can get in that spindle. For roughing out you probably want some large diameter tools.

all the aluminium i have used so far is ether 40/80 or 80/80 bosh compatible from kjn, i don't think changing the z will be to hard, i have some spare 40/80 i can use however the motor mount may have to go on the top like you have drawn which isnt so great as i have a hight restriction and need to keep it as low as posable so will have to see what i can come up with.

i dont think its going to be posable to do the y with one ball screw any were near the centre as the z will get in the way which mean needing two which would be a nice build but would cost something like a further £350 compared to about £60 ish that rack and pinion would be so will have to have a little think.

I think the max size bit would be 19mm but most i think are just 1/2" shafts so nothing crazy. foam likes to be cut quite quick other wise you get melting issues so you don't need big bits.

Would it be possible to use a timing belt and ballscrew. The ballscrew does the main postioning work and the belt just prevents the gantry from racking?

i think a belt and ball screw may end up getting more complicated than r&p and i would rather try and keep it as simple as posable although i expect it would work.

Just thinking out loud...

If the z axis was mounted to one side of the y gantry and the same side as the rack and pinion and as close to the rails as posable would racking be such a issue?

At present the y rails are 360mm apart, would reducing that help and if so what sort of distance before racking would not be so much of a worry 100mm? 150mm? 200mm? im am just trying to get a good idea in my head of how much of a factor it will be.

many thanks

I've been wondering along similar lines - if the gantry is stiff enough in the axis it's being driven in, is it necessary to drive both ends of it? Thinking of a design where the drive end bearings are wider spaced and the gantry is triangulated...

on the machine in the video they are running the x on 3 carriages, 2 on the top and one on the bottom with a belt drive along the top so i gess they have found a way around it, maybe that is why they only have 3 carriages ?

managed to get the plates made through the forum so i can mount up the x and y carriages and all so some adapters to mount the ball screw blocks to the ali extrusion, its taken a while but now should be able to get the x and y properly rolling.

The next step is going to be best to mount the rack and how i am going to do the motor mounts and keep things relatively simple. I guess i will also need to decide what motors i am going to use so i can get the mounting holes ect right.

Any advice here would be great. From what i see some people are using a pivoting motor mount with a spring to keep it in tension and others are just having a ridged mount that you set the meshing and then leave alone.

I have found this in the usa but i was hoping to use mod 1 rack and it looks like that is made for mod 2.

any advice would be great and many thanks in advance for any info you guys can give

best regards charlie

ps if any one is in the sussex area and can lend any help or advice it would be very much appreciated and repaid any way i can.

X and Y mounted up and rolling finally.

Jeez, that's one helluva build! :smug:

could any one tell me if i was to run two ball screws for the y axis what is the best or most practised way of running them as a pair? im my mind i am thinking there are a few options, i have had a search but not found to much as im not to sure on the terms used for this. Is it better to run two motors one on each or put a motor on one and have a belt going across to the other or have a centrally mounted motor with a belt going to each?

Also After a few people have mentioned it i am wondering if i would be better go for a hybrid closed loop stepper system due to the length of the x? servos are defiantly way out of my budget but is closed loop worth considering? can you run it on just one out of the 3 axis then upgrade the rest later?

cheers charlie

sorry i ment to add this which is me just playing around with the z axis ball screw and seeing how it may work on the y, it looks like it could be quite straight forward to mount it up and just use a angle bracket of the carriage to the ball nut mount.

Call me sad but really getting excited now! Just mocked the z axis up with some wood plates to try and figure a few things out and all looks like it should be relatively simple.

I have decided to go with all your suggestions and do the double ball screw and all so add the 3rd vertical rail to the Z axis. Its going to cost a lot more but it will actually probably be more simple and figured probably give me less headaches in the long run and wont kick myself wishing i had done it that way.

The X axis will have to stay rack and pinion as i cant see any other real way to do it other than belt which people have said is not such a good idea on that length.

I will order parts tomorrow then will need to get a few more plates made up to bolt it all up.

still no real idea on what motors im going to use but hopefully some one will be able to help me with that.

Got the two 1100mm y axis ball screws through in the post and mocked up some plates that i will need to get made, now need to add the 3rd z axis rail and figer how to fit it all in together so it nice and clean.

Sorry I some how missed the post from a week ago.!! . . . . I would definately join the screws together with a single motor and belts. Far more accurate and less hassle.

You really won't regret those 2 screws and extra rail on Z axis.

Looking good and keep it going, shame your so far away I'd gladly help.