operation cnc

Quote:

---

Originally Posted by kingcreaky

...[stuff]

1,)What is the maximum length I can have, with a conventional 'spun-ballscrew' setup (keeping the nut static)?

I intend to have two screws for the x axis, (one on either side)

The ballscrews descibed in the ebay auctions (linearmotionbearings2008) detail as being 'end machined'

4 pcs of anti backlash ballscrews RM1605-350/800/1450/1450mm-C7(4screws+4ballnuts+4endmachined)
4PCS NUT HOUSINGS/BRACKETS FOR THESE 3 BALLNUTS
4 sets of BK/BF12 with locknuts and circlips+ 4pcs 6.35mm*10mm flexible couplers
Linear slide rails SBR20-600mm(2 supported rails+4 SBR20UU bearing blocks)
Linear slide rails SBR20-300/1400mm(4 supported rails+8 SBR20UU bearing blocks)

so with the 'flexible couplers' also supplied in the kit, I assume will fit my steppers ?
Nema23 Stepper Motor 3.1Nm x 3

---

Given the BK/BF mountings you have a fixed/supported arrangement (supported BF at the far end, fixed double BK near the stepper) the critical speeds are:

350mm:10000rpm (lots of m/min)

800mm: 1965rpm (9.8m/min)

1450mm: 598rpm (3m/min)

To get closer to 6m/min on X you either need 10mm pitch screws, or you need to increase diameter to 25mm but then the inertia of the screws is going to be limiting on acceleration.

Another option is to upgrade the bearings to be double angular bearings (BK) at both ends. This will increase the critical speed to 880rpm (4.4m/min), better but still not great.

Here is a good PDF that shows how these are assembled: https://tech.thk.com/en/products/pdf/en_a15_314.pdf

And the critical speed calculator is here: http://www.nookindustries.com/engine....x=0&reset.y=0


Couplers are fine: 6mm stepper shaft to 10mm end machined screw.

now I am confused! I think im going to have to read your post another 100 times slowly to understand, then study the pdf. perhaps im not asking the right questions? or perhaps relying too much on telepathy
Essentially you have succeeded in making me aware there are different 'types' of ball screw housings. I think what your saying is the ones ive detailed above (which incidently I havent yet bought anything) are for a fixed screw arrangment, and if I want a spinning screw I need to re-think my choice of screw housing? once again, one question results in ten more :D sorry :ambivalence:

As irving has said, you need to use 10mm pitch (RM1610) ballscrews on the X and Y axis to get respectable feedrates without using a larger diameter ballscrew.

Clearly it depends on what the minimum feedrate you can accept is, but I wouldn't go above a 1500mm long RM1610 ballscrew without using a larger diameter screw or rotating the nut. If you're interested I can make the rotating ballnut mounts for you, let me know if you wanted the details. I wont be able to make them any time soon, but it looks like it will be a while until you're ready anyway.

The 3Nm stepper motor shafts are 8mm, so you would need 8mm to 10mm couplers. However I would strongly advise using timing belts and pulleys since this enables adjusting the drive ratio to get better speed or resolution with the added bonus of adding damping to the system to suppress resonance.

Quote:

---

Originally Posted by kingcreaky

now I am confused! I think im going to have to read your post another 100 times slowly to understand, then study the pdf. perhaps im not asking the right questions? or perhaps relying too much on telepathy
Essentially you have succeeded in making me aware there are different 'types' of ball screw housings. I think what your saying is the ones ive detailed above (which incidently I havent yet bought anything) are for a fixed screw arrangment, and if I want a spinning screw I need to re-think my choice of screw housing? once again, one question results in ten more :D sorry :ambivalence:

---

Nooo... Imagine the screw is a long bendy pole... the faster you twirl it the more it flaps around. To stop it flapping you need to constrain it. The way to do that is to put bearings at each end. The bearings are fitted into bearing housings. The one at the far end has a single bearing in it, the one nearer the drive end has two bearings. The single bearing just supports the screw, it doesnt stop it moving end to end (axially). The double ones stop the screw moving end-to-end, hence 'fixed'. Both types allow the screw to turn! By putting a 'fixed' one at the far end as well controls the whip a bit more but it needs a different end-machining of course.

this really is like a mind boggler. ive spent pretty much every waking moment, reading posts, writing notes but mostly thinking. Oh, and getting frustrated with google sketchup.
I have made a final, firm decisions on a few things.
*It will be around 1.5m by 1m
*after reading the pros and cons, and weighing them up in my own head, Im using chai's supported bearings SKB16, maybe 20's for the gantry to slide along on(is that the X some people call it X some people call it Y)
*The first version, Il use the 1610 ballscrews I think that is what Irving is trying to suggest in his above posts) , and although maybe not recommended il spin the ballscrew not the nut. I may eventually re-think this once ive learnt from my mistakes. As Il be driving the ballscrews with timing belts I can gear it so they for optimal reliability not speed.

I am very confident with steel. I can weld well, (although I understand the issues over warping) so please see my gantry design.

Ive read thread after thread explaining the importance of keeping the gantry supported to stop flexing. I also appreciate the ultimate goal is keeping the cutting tip as close to the center of the gantry as possible through organising the Y and Z axis /ballscrew configurations.

With the comment over confidence in steel and no experience in extrusion or ali whatsoever, please see my current gantry plan (attached?)

as you can see, ive used two peaces of 50x200x6 I intend to triangulate the back for extra support and place the Y bearings one at top and one at bottom (top marked in red)

In terms of fixing the material together, I think im going to carefully weld it.

I need your thoughts now, Ive heard Jonathon banging on about steel box not being 'straight' and I should probably be using ecocast ali. or even extrusion like some of the builds here. Without entering into yet another full on discussion over gantry design. is carefully welding the steel going to distort it that much the z unit wont slide the length of the gantry without binding?

should I use extrusion?

With extrusion... how do you know the t-slots or slots in the extrusion are going to line up with the pre-drilled holes int he supported rail?

where do I buy extrusion?

totally consumed in this, and seeking some support!

m@

I'm also new to this KingCreaky, so can only try help a wee bit...

For this size machine which is similar to the size I'm designing, choosing 1610 ballscrews seems spot on, but I think the gurus are likely to suggest 1605 only for the short Z-axis. And spinning the ballscrew rather than the nut is the default norm.

If you're gonna use supported round rail I think the gurus would suggest Chai's SBR20 rail.

Where peeps do use extrusion, a version of 45x90 extrusion is popular on this forum that hails from your neck of the woods... KJN in Leicester... Aluminium Profile with 10mm Slot

The KJN 45x90 extrusion has slots that are well suited to the precision-profile-rail that is terribly pricey, but not particularly suited to supported round rail.

Hope this helps.

I am working on a similar sized machine and have been through the same questions on ball screws - seemed to me there was a lot of different opinions floating around. I too decided that spinning ballnuts was a bit out of my league! I went for 2005 in the end, the larger dia will hopefully help with whipping and I am going to gear it 2:1 so as to get faster feed rates, with the option of gearing 1:1 for high res work...

I am coming round to the idea of using steel (ill have to learn to weld tho!) for the frame and x axis but Ali extrusion does seem the best option for the gantry. The L shape arrangement used in a few builds allows you to keep the gantry squat and rigid. Why does your gantry need to be so tall and thin...?

If you need any help with sketchup just ask I have been using it for years and have just about got through the pain in the ass stage !

Quote:

---

Originally Posted by JoeHarris

I went for 2005 in the end, the larger dia will hopefully help with whipping and I am going to gear it 2:1 so as to get faster feed rates

---

Not really, you'll end up worse than with 1610 regardless of the drive. Critical speed is proportional to the diameter, so going from 16 to 20 means the ballscrew can rotate 25% faster. However to get the same feedrate with 5mm pitch (2005), you need to spin the screw 100% faster, but can only go 25% faster... so you end up with 62.5% of the feedrate attainable with 1610.

If the L-shape gantry is reinforced by putting some box section on the back at 45° to make triangles it will be very strong.

OK but surely it depends on the speed you need. I remember reading on another post that Jazz gets no whip running his 1500mm 2005 up to 12meter/min... Plus you get greater mechanical advantage with a smaller lead? Well I have bought mine now so when I have built the machine I will be able to report - Just hope I will get enough speed to cut ply without burning it up!!

Quote:

---

Originally Posted by JoeHarris

OK but surely it depends on the speed you need.

---

Yes, if it's sufficient and the motors can drive it at that speed you're fine. But that's the other issue, you will need significantly more torque from the motors to obtain the same feed-rate/acceleration as with the 16mm screw since it is not a linear relationship, power 4 in fact, so for a given acceleration you need (20/16)^4=2.44 times the torque. So if at the speed/acceleration you would have run the machine on with the 16mm screw you have that much spare torque then it would be fine with 20mm, but it's a shame because you would definitely have got better feedrates with the 1610 screw and it's added an unnecessary risk.

For a 1500mm 2005 screw the critical speed is about 1200 rpm with the standard bearing arrangement, so 1200*5=6m/min feedrate.
For 1610, same length etc, it's about 1000rpm, so 1000*10=10m/min feedrate.

From experience the formula used here seems to be conservative, so that is probably enough assuming the motors you have can actually output enough power to achieve that feedrate.

Quote:

---

Originally Posted by JoeHarris

Plus you get greater mechanical advantage with a smaller lead?

---

True, but you can use the pulleys to get whatever mechanical advantage (i.e. ratio) you want, so it's a bit of a moot point.