BUILD LOG: Steel Framed Router [Archive]

GTJim

31-08-2011, 12:04 PM

luke11cnc

31-08-2011, 12:38 PM

hello first

If you are using 25mm supported rails you will need box of 60mm and over

plus two ballscrews as to stop the bed crabbing

this is just the advice I have received as I'm by no means a expert

you may want to look at Jonathan's and chips CNC machine
James and Luke

GTJim

31-08-2011, 01:08 PM

If you are using 25mm supported rails you will need box of 60mm and over
plus two ballscrews as to stop the bed crabbing
James and Luke

Hi Luke,
I have details of a fully supported rail that has a 21mm dimension across the base. I hope it is still available.
As for the crabbing, I was hoping by spreading the lower bearings apart as far as I did this may be avoided, but I will take the comment on board.

Jim

Ricardoco

31-08-2011, 02:46 PM

Hello. I am hoping to design and make a router for hobby use just like many have done previously on this site, and I am currently putting together a concept and these are my current thoughts.
I am aiming for a working area of 1000mm x 1300mm (approx 3’x4’) and a Z of 130mm (5’’).
The main frame and gantry is currently 50x50x4 steel square hollow section with 25x25x2.5 bracing for the bed. This will be bolted together unless I find a good welder.

4408

The linear rails are 25mm dia for the Y axis and 20mm for the X and Z, and hoping to use a 2510 ballscrew for the Y a 1610 ballscrew for the X and 1404 ballscrew for the Z. (hope I haven’t confused the X and Y).

4409

4410

I would appreciate your comments and constructive criticism as the design is easily changed at this stage before any serious money is spent. This is only a hobby machine and its main use will be engraving hard wood panels and cutting out the odd speaker panel.

Hi Luke,
I have details of a fully supported rail that has a 21mm dimension across the base. I hope it is still available.
As for the crabbing, I was hoping by spreading the lower bearings apart as far as I did this may be avoided, but I will take the comment on board.

Jim

Hi Jim, the design looks cool, I notice that the side bracing for the gantry is biased towards the spindle side of the gantry is this by design, IE is the Z axis and spindle going to exert a greater downwood force than the upward force generated while cutting.

Ive seen the crabbing or twisting of the gantry avoided by framework being added to the opposite side of the gantry to the spindle as in the commercial machines.

These are just observations and are not supposed to imply any knowledge or special wisdom on my part.

Oh and what software did you used to produce the pictures?

Rick

GTJim

31-08-2011, 03:04 PM

I notice that the side bracing for the gantry is biased towards the spindle side of the gantry is this by design, IE is the Z axis and spindle going to exert a greater downwood force than the upward force generated while cutting.

Ive seen the crabbing or twisting of the gantry avoided by framework being added to the opposite side of the gantry to the spindle as in the commercial machines.

Oh and what software did you used to produce the pictures?

Rick

Rick, I am unsure about cutting forces but I was hoping to prevent the gantry flexing back and forth while cutting. I may also add a plate to the back of the gantry between the X axis to further restrict right or left deflection.
I just used Google SketchUp for the drawings as its quick.

Jim

Ricardoco

31-08-2011, 03:13 PM

Rick, I am unsure about cutting forces but I was hoping to prevent the gantry flexing back and forth while cutting. I may also add a plate to the back of the gantry between the X axis to further restrict right or left deflection.
I just used Google SketchUp for the drawings as its quick.

Jim

I know its cheeky but can you send me an editable version of drawing two and i can modify it to show you what i mean..

GTJim

31-08-2011, 03:31 PM

Rick, I've attached a zipped copy and hopefully it has worked.

Ricardoco

31-08-2011, 03:36 PM

Rick, I've attached a zipped copy and hopefully it has worked.

Magic it worked fine, Be back in a mo..

Rick

Ricardoco

31-08-2011, 04:15 PM

Rick, I've attached a zipped copy and hopefully it has worked.

4415

Ok ive just done enough tinkering to give you an idea, of where bracing could do with looking at.

It may also be an idea to widen up underneath the table on the gantry just enough to stop the diagonal flex when the spindle is working at the far reaches of the gantry axis.

I built a table router very similar to this one some time ago and that is the only real problem area i had to re-think..

I would have put a double row but unless your stepper was on top it wont fit.. Good luck with the project i will be interested to follow your progress, its always nice to see projects from concept through to completion..

All the best Rick

JAZZCNC

31-08-2011, 05:31 PM

The linear rails are 25mm dia for the Y axis and 20mm for the X and Z, and hoping to use a 2510 ballscrew for the Y a 1610 ballscrew for the X and 1404 ballscrew for the Z. (hope I haven’t confused the X and Y).

If these are profiled linear rails you dont need 25mm they will be massively over kill. 20mm will be more than enough, even 15mm will be good enough. I'm talking profiled linear thou not round supported.

25mm ballscrew is also not recommended for a machine this size, you've probably gone with 25mm because of the 10mm pitch not being available in 20mm.
25mm screws are large and they produce a lot more inertia which has to be stopped and accelerated this means big strong motors needed amongst over things which all make it not ideal.! . . . Bigger is not better when it comes to some things, screws being one of them.!
At the size your looking to make 16mm is getting on the edge of being whippy thou with carefull aliagnment and sensible speeds you would be ok.
You could go with 20mm 5mm pitch and gearing 1/2 this way you get the speed thou at the cost of a little torque. I can tell you from experience 6nm Nema 34's will handle 1/2 setup without any problem. There is an up side with 1/2 gearing as well to this if you find you ever need the extra resolution of 5mm pitch and higher torque it's a simple pulley change and away you go.

You could go with rotating the nut but it gets a bit more involved, thou does work well if done right.

With the steel box section frame you will also get lots of resonance which has a big affect on quality of cut esp when cutting hard materials so a big tip if you go all steel is cap the ends and fill the frame with kiln dry'd sand it makes a big difference.

Definatly go with twin screws on the X axis it makes a very accurate strong stiff machine. Either with a motor on each screw slaved together or one motor with belt between the 2 screws.
Dont be put off by the belts they work fine and more than good enough for the work you intend to do. I have a machine that use's this 5mm pitch 1/2 twin screw belt setup and I cut every thing from steel to plastic with it no problems and it's super accurate. Even cut cast iron with it last week for the first time without any trouble.!
Not a pritty thing but it gets the job done and very very reliable and super accurate.

Cheers.

Jonathan

31-08-2011, 06:48 PM

25mm ballscrew is also not recommended for a machine this size, you've probably gone with 25mm because of the 10mm pitch not being available in 20mm.

Maybe 16mm screws with 10mm pitch - RM1610.

25mm screws are large and they produce a lot more inertia which has to be stopped and accelerated this means big strong motors needed amongst over things which all make it not ideal.!

I was about to suggest looking into using rotating ballnuts...

You could go with rotating the nut but it gets a bit more involved, thou does work well if done right.

But you beat me to it! You might be interested in how I did the rotating ballnuts on my machine:

http://www.mycncuk.com/forums/showthread.php/3340-Rotating-Ballnut-design-ideas

They work well, so I guess I did it right.

I agree with using belts and two ballscrews on the X-axis. Spacing the bearings further apart will help prevent the gantry racking a bit, however it's still not ideal.

i wouldnt look at my build for any design inspiration,its a bad build and a lot of good materials wasted,im on the verge of scapping most of it and starting again:tup:,

Oh don't say that ... you're doing yourself a disservice. There's still room for improvement on mine and there always will be. The distance parallel to the X-axis from the Y ballscrew and cutter is far too great on my machine for instance.

GTJim

01-09-2011, 07:52 AM

Ok ive just done enough tinkering to give you an idea, of where bracing could do with looking at.
It may also be an idea to widen up underneath the table on the gantry just enough to stop the diagonal flex when the spindle is working at the far reaches of the gantry axis.

All the best Rick

If I was to mount a plate (as shown in pic) to the back of the axis with many fixings would that stiffen up the gantry sufficiently?

4425

The current design supported round rail, but this was down to cost and if available I would use profiled rail.
I also think the 25mm ballscrew is OTT but now that it has been mentioned I'll look at 20mm with 5mm pitch with pulleys and belts. I am concerned about whip in the 16mm screws, but again it could be a saving.
I am hoping to use a single ballscrew on each axis but if I find problems I think it would be easy enough to add an additional one, so I'll look out for a 4 motor and driver package.
Never thought about resonance, but kiln dry'd sand could be a good solution.

Thanks for all your input and please feel free to continue.
Jim

I can see this becoming a bit of a monster and it has to fit in my garage with all my junk.

GTJim

01-09-2011, 03:58 PM

Just a couple of more sketchs showing potential positions for ball screws and motors.

4427
4428

Keep the comments comming as I will be buying materials soon.

Jim

Ricardoco

01-09-2011, 04:09 PM

:confused:

If I was to mount a plate (as shown in pic) to the back of the axis with many fixings would that stiffen up the gantry sufficiently?

4425

The current design supported round rail, but this was down to cost and if available I would use profiled rail.
I also think the 25mm ballscrew is OTT but now that it has been mentioned I'll look at 20mm with 5mm pitch with pulleys and belts. I am concerned about whip in the 16mm screws, but again it could be a saving.
I am hoping to use a single ballscrew on each axis but if I find problems I think it would be easy enough to add an additional one, so I'll look out for a 4 motor and driver package.
Never thought about resonance, but kiln dry'd sand could be a good solution.

Thanks for all your input and please feel free to continue.
Jim

I can see this becoming a bit of a monster and it has to fit in my garage with all my junk.

Ok let me show some diagrams..

4430
With no load and Spindle central this is the look of the machine from above.

4429
Now cutting brings in to play these forces and consequences.

As the others have touched on instead of using 1 screw in the center if you use 1 on each side that will sort it out but at the end of the day it is all down to cost and accuracy. if you take the above examples and apply them to the other two axis you will see the weaknesses in your structure at the time of design not when you have spent the money, most of us on here have been through the building process and made the mistakes for you...:whistling:

BUT. my first attempt at building a machine was flawed and at the end it was not accurate but it was cool looking and i played for hours with it just amazed ide built it and to me for that split second it was PERFECT!!:smile:

then i took it to pieces and joined this forum... the rest is history...

""MOST"" of these guys on here really do know their stuff, but dont just settle for one opinion get lots of opinions and you will soon see the way or be extreamly confused :confused:

Rick

Jonathan

01-09-2011, 06:51 PM

The plate on the gantry to stiffen it is an excellent plan. I think you need to consider more how you are joining the box section, and gantry sides to bearings. Currently you have drawn the box section attached on just one face which is a bit weak.

I have done a drawing for you to illustrate how I think it can be made stronger:

4431
4432

That way the sheet reinforces the joints in addition to the general structure. I would be inclined to weld it for good measure. If not put plates (say 20mm thick) in the ends of the box section and bolt into them.

JAZZCNC

01-09-2011, 09:37 PM

Jonathans suggestion is a good one not just for strength but also it protects the ballscrew from flying debrie.
If you must go with 1 screw done the middle then I suggest you make the cross piece that the ballnut fastens to wide as possible. you need strong bracing and width to counteract racking with a single screw.
Something like the pics below will handle just about most jobs at this width with a single screw. A 16mm 10mm pitch would work nicely.

Edit: The sketchs are rough quick knock up's and I would add some more corner braces to the cross piece but it should give you an idea of what I mean.

It probably looks a bit extreme but honestly it's not and the extra work and slightly more expense would be worth every minute and penny spent belive me.
Dont under estimate the forces's even with steppers.!! . . The forces can be extreme when cutting Ali and hard materials and the torque produced by steppers and ballscrews can twist and mangle a gantry built like this in the blink of an eye when the stick stuff it's the fan. :surprised:

Jonathan

01-09-2011, 10:10 PM

torque produced by steppers and ballscrews can twist and mangle a gantry built like this in the blink of an eye when the stick stuff it's the fan. :surprised:

Yep, mine has done that once or twice. It's quite surprising. Once nice test I can do is if I disengage the motors and rotate one ballnut by hand the gantry is stiff enough to make the other one backdrive and rotate too... but not much.

I wonder how much difference it would make if you have two ballnuts on the single screw? One on each of the two cross pieces on the gantry.

JAZZCNC

01-09-2011, 10:25 PM

I wonder how much difference it would make if you have two ballnuts on the single screw? One on each of the two cross pieces on the gantry.

Dont think it would add much to extra torque but it would certainly add extra precision due to removing backlash and compensating for wear also help with stablizing whip on longer screws.
I have some lovely 25mm Comtop ground Class 3 preload 2 double nut ballscrews for my next project and the difference between the cheap chinese screws and these is night and day. . . . Cant wait to get on with it.

GTJim

02-09-2011, 08:37 AM

Thank you all for your very good comments.

Rick - I will redesign the frame and go with 2 ball screws as suggested, possibly 16mm 10pitch.

Jonathan – In my initial sketches the axis was constructed in a similar manor to your drawings but I found the spindle did not cover the whole of the table. So I either extended the frame of the base or took the rails to the extreme of the axis allowing the spindle to go past the work surface while keeping the foot print of the machine as small as possible. But as you have pointed out it has created a weakness and I will take this on board.

JAZZCNC – I also agree Jonathans suggestion is a good one, but as for the protection of the ball screw I was hoping to use brush seals fixed to the gantry to prevent wood chips causing damage. I’m sure I’ve seen this being used on a machine on this site.
I also like your design for the gantry, so I’m back off to the drawing board. I’ll keep you posted.

What are your thoughts on these Chinese screws as I don’t want to work to microns?

As you are all well aware of, everything is a compromise.
Jim

Jonathan

02-09-2011, 02:37 PM

What are your thoughts on these Chinese screws as I don’t want to work to microns?

The ones I have (C7) are all smooth with next to no backlash. The backlash is less than the resolution I get so it's not significant.

Apparently they do vary...

JAZZCNC

02-09-2011, 07:10 PM

What are your thoughts on these Chinese screws as I don’t want to work to microns?

I dont know about other chinese ballscrews but I have bought and fitted dozens from Chai at linearmotionbearings and they are very good considering the money. . . More than good enough for Diy.
I have some of chai's screws on my big machine and they are just starting to show signs of wear after 3 years of very very hard use cutting mostly Ali plate and in the worst possible conditions due to the fact they are uncovered and partly in the firing line.!! . . It's one of them " I'll do it later" jobs that after 3yrs still aint got done.!!! . . . .

BIG TIP . .DONT Start the bloody thing up untill you've done every thing you planned else they'll never get done.!! . . . So much fun you'll never want the down time.

Jonathan

02-09-2011, 07:17 PM

BIG TIP . .DONT Start the bloody thing up untill you've done every thing you planned else they'll never get done.!! . . . So much fun you'll never want the down time.

:heehee: Yep I agree entirely. That's why I don't have any switches on it, or covers and only crude dust extraction ...

My ballscrews are also from Chai. Only just got round to putting better computer on it that I've had for a while. What a difference...RM1610 screw on Y-axis is now going at 36m/min which is the critical speed and it seems fine. I wonder what their limit is :naughty:

JAZZCNC

02-09-2011, 07:48 PM

:heehee: Yep I agree entirely. That's why I don't have any switches on it, or covers and only crude dust extraction ...

No Switch's.! . . Not even Home switch's.?

Next to touch probe the Home switch's are the best and most time saving thing you could put on your machine. Can't count the number of times they'e got me out of trouble. . . must be in the hundreds.
They must have saved me hundreds of pounds in material that would have been lost or wrecked due to loss of position when an error or tool breaks. . . . . Nothing worse than a tool braking after several hours cutting one job. . . No problem and Simplizz's with home switch's.
Send machine home to reset it's position, rewind the code to just before the tool broke and away you go again. . . Brilliant.

No limit switch's on the large machine, IMO you can live with out them on larger machines, soft limits are more than good enough and in the 3yrs using the big machine I've never hit the hard stops or the soft limits other than me trying to drive the thing into them.:thumbdown:

Jonathan

02-09-2011, 07:59 PM

No Switch's.! . . Not even Home switch's.?

Nope...I did get round to wiring one switch but it's not attached yet. I generally either make a mark at the home position with the tool to reference from if it stalls, or note the position of the ballnuts or screws. Granted that's not as quick as homing assuming it'll home at a reasonable percentage of the rapid feedrate. One day I'll do it...

JAZZCNC

02-09-2011, 08:32 PM

Nope...I did get round to wiring one switch but it's not attached yet. I generally either make a mark at the home position with the tool to reference from if it stalls, or note the position of the ballnuts or screws. Granted that's not as quick as homing assuming it'll home at a reasonable percentage of the rapid feedrate. One day I'll do it...

Jonathan Honestly do it ASAP you wont regret it I must use it 1 in 3 times I use the machine. Infact only to day I was cutting some a parts which had small bearing pockets and followed by an outer profile pass.
The code rough'd the pockets first then was to finish them with 0.2mm pass, now I'd forgot to the set code to continue with out stopping between rough n finish pass's so when it went to x0,Y0 and stopped I thought it wanted a tool change so changed for the outer pass tool.! .
When I hit cycle start it went on it's way at full rapid 10mtr/min. . . But to the wrong place.:rolleyes: . . . . Luckly I was watching and hit the Oh-shite button before it did damage. Now as you know at full 10mtr rapid to dead stop it obviously lost position probaly not too much but it did.
This I probably could have got away with doing it your way for the non critical outer profile pass but my pockets where 0.2 under size and this I couldn't live with so part would have been lost.

With home's it was a 2 minute affair.! Send it home, change tool back and reset tool height, rewind code with "run from here" to finish pass and away I go back on track and in precise position. . . . . Thats why there so good.:tup:

Edit: Actually the part wouldn't have been lost and I could have made use of the next best thing to you can add to a machine "The PROBE" . . Could have Probed to find the centre of the pockets and tracked my way back from there, works good but still not ideal and more time consuming.

Jonathan

02-09-2011, 08:48 PM

Hmm, I guess I'll do it when I attach the profile rails as that'll require a few changes anyway. I can get it plenty accurate enough generally by eye or zeroing from a hole but obviously it can take a while.

Just got my Y-axis going at 51m/min! Seems reliable at that...I don't think limit switches would be much help at that speed!

JAZZCNC

02-09-2011, 09:41 PM

Just got my Y-axis going at 51m/min! Seems reliable at that...I don't think limit switches would be much help at that speed!

!!!! . .51M/Min .. !! . . . . What you using for motors.? Servo's.? Thats 5000rpm with 10mm pitch. . . . What stepper you got to do that speed and still give torque enough to drive gantry etc.?
Also what you using for drives and computer because if Mach you must have a good computer and parallel port that will let you set the Kernel high enough to give them speeds.?
Think even my smooth stepper would struggle to get usable steppers to that speed and that blows the PP away.!!

Please Dont take my asking has me not believing you, I honestly do and I'm genuinely interested how you've done it.! . . . . Does the rotating nut make that big a differance on a small machine.?
Post a video bet it looks amazing. .:toot:

Jonathan

02-09-2011, 11:31 PM

I'm not doing anything special really but yes it's a bit hasty. It's not for long though as the acceleration is only 2m/s^2.

I just got 1m/s! Will post a video soon...

GTJim

06-09-2011, 02:19 PM

I've just started looking into the possibilities for pulleys for the ball screws and stepper motors.

I understand HTD belts have been recommended, but my problem concerns the spindle on the motor I'm considering, it's Dia 8mm x 21mm long approx. I don’t really want to use a belt less than 15mm wide but can't find a pulley to fit the length of the motor shaft.

What has been used before, any ideas?

JAZZCNC

06-09-2011, 03:31 PM

The pulleys I use on steepers are HTD 20T 15mm they measure 25mm wide with 5mm boss. You can either use 4mm set screws thru the boss which leaves the shaft 3-4mm short of being flush.
Or do like I do and drill in the middle of the pulley between the teeth with a larger 5 or 6mm set screws which allows better purchase, just make sure the head of the screw sits below the bottom of the teeth.
This hisn't a problem because of the wider OD of the teeth theres more metal to work with. Then I either machine the boss down less so the pulley sits flush or some times dont bother and just turn the boss to the outside.

Unfortunatly I haven't got any machined to show you but I do have some unmachined pulleys which I've just taken pics of to give an idea of what I mean, also give you some idea of scale against a nema 23 stepper.
The set screws are M5's note the blue tak holding them took me 15 mins before I gave up trying to balance them in those groves.:whistling:

I've never had any problems with slipping doing it this way even with just one set screw, thou I always use at least 2, some times 3 120deg apart.!

Other option if going thru paltes etc is to have some extension shafts machined, thou not too long. Either Jonathan or I could do that for you if no access to lathe.

GTJim

12-09-2011, 11:37 AM

Its been a while but I’ve been updating my design and have made a number of changes to include many of the improvements suggested in this thread. But I have also been persuaded to widen the table to 1250mm (4’) so I can machine a half sheet (seemed a good idea at the time).

So I have now put together a preliminary BoM for the machine along with weights and I am very concerned that it has evolved into a monster. Lol.

Excluding ball screws and misc nuts and bolts, but including spindle these are the approx weights:

X Axis 21Kg
Gantry 76Kg
Base & Table 149Kg

So my next issue is to work out how to drive these masses, What would be the best combination of driver, motor and ball screws. Would the 60BYGH401-03 4Nm steppers work and are 16mm screws to flimsy?

Jim

JAZZCNC

12-09-2011, 04:24 PM

Hi Jim,

I wouldn't be concerned with the weight 76KG for the gantry is not excessive. I would always take a slightly slower heavy gantry over a skinny fast one anyday. Mass can be your friend if matched with right motors and screws.
Regards the 16mm screws.! . . . Depends is the only answer I would give without knowing the full dimensions. I'm pritty sure they will handle the mass/force's etc but length will determine if they are suitable without whipping. . . Give more info if you can.

Don't know those motors.! Are they Nema23 or 34's. . . . Personally I prefer 34's for larger machines with weighty gantry's. They don't quite spin as fast as 23's but the torque remains higher up the power curve, giving strong torque without much drop-off as the rpm's rise then drops off the clifff really fast around 8-900rpm.
23's tend to drop gradual as the rpm's rise but will carry lower torque to higher rpm's 1000-1200rpm + is not unusual.

Again without knowing the full details it's hard to give a definate "Yes" or "no" but @4nm I think you'll be ok for most single ballscrew setup's. Ball screws are very effiecent so allow less powerfull motors than other linear drive system's IE:R&P etc.

Jonathan

12-09-2011, 05:52 PM

We need to know the travel of each axis, and the pitch of the 16mm screws to work out if they are suitable. The main limiting factor is in theory the speed at which the screw starts whipping - critical speed.

The only 4Nm Nema23 motors I know of are these which are the same part number:

http://www.cnc4you.co.uk/index.php?route=product/product&path=20_27&product_id=72

Their inductance is nice and low. The rotor inertia isn't stated...
I'm intending to get a few of them for my router so that I can put the ones from the router back on my mill. Not sure how much difference it'll make over the 3nm. I'll be interested to see what they're like.

GTJim

13-09-2011, 08:10 AM

The X axis ball screw would need to be 1600mm long and the Y axis (gantry) 1500mm long. I was hoping to use 10mm pitch screws, so it seems your limited to either 16mm or 25mm. Whip is a concern.

4526

The motors are the Nema 23’s on ebay from CNC4YOU and appear to have a low inductance and 4.0N.m hold torque, but I don’t know if that’s enough and what I’ll actually need.

4527

Are there any simple rules of thumb to work out your requirements regarding motors and screws?

And one word of advice for anyone designing a machine, make sure you have lists and lists of standard size materials. I have gone through the pain of having to redesign this because I didn’t check on the sizes of steel or MDF. “D’oh”

JAZZCNC

13-09-2011, 05:29 PM

Personally I wouldn't go with 16mm at these length's unless like Jonathan you used a rotating nut setup.
They are just too close to the edge for my liking and would need to be setup and aligned absolutly perfectly and kept this way, even the slightest miss alignment at this length would show up as vibration easily creating whip as the speed rise's.

I wouldn't go with twin 25mm screws either bit OTT and hard on the motors, esp little 23's.

I see you intend using belts and pulleys so you could always go with 20mm x 5mm pitch geared 1:2. If your wanting reassurence this works good then think I've mentioned it before.? . . I have a machine that use's this setup but with a single 6nm nema 34 motor driving 2*1500mm screws joined with belts. It works very very good and cutts every thing from paper to Ali and steel with super repeatabilty and accurecy.

I will guarantee you from experience with my original 2 x 3nm setup that 2 x 4nm 23's geared 1:2 on 2005 screws will handle every thing you can thro at it,. . then some.!

Seen as your using belts and for simplicity sake my first choice from experience would be 2005 geared 1:2 but if must use 16mm or need higher rapids than 10-11mtr/min then go for the more involved rotating nut setup. . . . Both will work good.

OR find 20mm 10mm pitch screws.?

mocha

13-09-2011, 06:47 PM

Rick, A silly question perhaps... most of the utube vids I've looked at seem to position the work at 0,0 giving the forces you describe in the pic 2. Would I be right in thinking that if the work was positioned mid Y (above the ball screws) with x=0 or where ever, it wouldn't be so so much of a problem? And, would that imply that best accuracy / precision would be in the centre of the table rather than at the edges??

JAZZCNC

13-09-2011, 06:55 PM

Jonathan

13-09-2011, 06:57 PM

Critical speeds with BK/BF bearings (approximate):

16mm dia, 10mm pitch, 1600mm long - 800rpm = 8m/min
16mm dia, 10mm pitch, 1500mm long - 910rpm = 9.1m/min
20mm dia, 5mm pitch, 1600mm long - 1000rpm = 5m/min
20mm dia, 5mm pitch, 1500mm long - 1150rpm = 5.75m/min
25mm dia, 10mm pitch, 1600mm long - 1220rpm = 12.2m/min
25mm dia, 10mm pitch, 1500mm long - 1400rpm = 14m/min

So given that in my opinion 20mm diameter 5mm pitch screws are the worst option regardless of the gearing since the critical speed is quite low. Also the moment of inertia of a 20mm screw is 2.44 times greater (since it's proportional to the radius to 4th power) than the 16mm screw. That means the motor has to impart that much more kinetic energy into the screw to accelerate it to a given feedrate. The pitch of the 20mm screw is half that of the 16mm screw it clearly has to spin twice as fast. As the motor torque is proportional to the angular velocity of the screw and the inertia you loose big time since the torque required is now a factor of (10/5)*2.44=4.88 greater between the two options!! This doesn't take into account cutting forces, which would make the figure look not quite as bad ... but still it's a massive difference.

Clearly I can use a similar argument as the above to say why the 25mm screws are too big for those motors...unless you rotate the ballnut. Rotating the ballnut is a win-win situation since screw whipping is no longer an issue and the moment of inertia of my rotating ballnut assembly on the RM2510 screw is about a third of the inertia of my 2m long screw. With that I get 15m/min (well over the critical speed) confortably and I'm limited by the 3nm motors, not the screw. For your size machine I reckon 16mm diameter with rotating ballnuts might be a good option...

If you can find 20mm diameter, 10mm pitch screws then that's potentially good option (though still the inertia is greater as above), but I bet it's going to cost you more than a rotating ballnut setup on the 16mm screw.

This does of course all depend on how fast you want to go. If you're not trying to get the land speed record then the RM1610 screws would be fine...

JAZZCNC

13-09-2011, 08:59 PM

Don't want to get into an argument Jonathan but I know from real world experience all these calculations means nothing in DIY cnc.

All well and good if building cnc machine using high precision machinery built to exact specifications with top notch components all perfectly aligned.!. . . .but in the real world of DIY using a drill press and rudimentry tools trying to align screws etc using primative methods then it don't work like the nice calculations predict. . . . The extra beef of thicker screws takes up the slack of DIY building no matter what the calculations say should happen.!!!

My setup being a prime example that your figures dont work in the DIY world, . . . It happly spins 20mm screws at twice the critical speed you quote.!! . . Those figure say it shouldn't . . . but it does and has done for the last few years without any problems.!

Jonathan

13-09-2011, 09:21 PM

The extra beef of thicker screws takes up the slack of DIY building no matter what the calculations say should happen.!!!

Perhaps, but you can't deny the simple calculations I did in the previous post demonstrate a significant torque disadvantage from using a bigger screw with lower pitch.

My setup being a prime example that your figures dont work in the DIY world, . . . It happly spins 20mm screws at twice the critical speed you quote.!! . . Those figure say it shouldn't . . . but it does and has done for the last few years without any problems.!

I know, my RM1610 ballscrew happily exceeds the critical speed by quite a lot. The thing is the critical speed formula doesn't take into account the position of the ballnut, which acts as another support hence leading to the calculation being an underestimate in a lot of cases. This is partly the reason I concluded that 16mm screws are optimal - from a torque point of view they are far superior to the others and in reality it seems the critical speed is higher than what I worked out, so ultimately you'll end up much better off than with the other options.

The calculations of critical speed may not mean as much as they imply, but I feel they mean more than nothing and give a good basis.

JAZZCNC

13-09-2011, 10:20 PM

Perhaps, but you can't deny the simple calculations I did in the previous post demonstrate a significant torque disadvantage from using a bigger screw with lower pitch.

No regards torque It's undisputable obviously, but I know from experience with actually using them not calculating or theorising.! . . they have more than enough torque even with the disadvantage.

The calculations of critical speed may not mean as much as they imply, but I feel they mean more than nothing and give a good basis.

Yes to a small degree but simply too meny variables to explicitly say that 16mm are the best by far, to the point of dismissing the others completely.

I tend to work and put more value in experience than calculations and I know from my own experiences with helping others that long thin 16mm screws give more grief than the same length 20mm.
If you want to go looking around the zone etc then I think you'll find more unhappy owners of long 16mm screws than you will 20mm.

I know if asked to build a machine with 16mm screws of this length I would re-fuse and recommend 20mm with correct size pitch to suit the intend purpose of the machine.

My proposal to use geared 20-05 was work around to no cheaply available 20-10 and wasn't saying they are the ideal solution just an easy one that I know works thru practicle hands on experience.:wink: . . . The right pitch for the job wins every time but theres always other ways to skin a cat.

I'll take real world experinece over calculated theory any day . . . . IMO it's priceless and feel privilaged when someone shares there often hard learned experineces.

mocha

13-09-2011, 10:39 PM

Oops, sorry gents, dropped the quote somewhere along the line, what I should have posted is:

:confused:

Ok let me show some diagrams..

4430
With no load and Spindle central this is the look of the machine from above.

4429
Now cutting brings in to play these forces and consequences.

Rick

Rick (anyone), A silly question perhaps... most of the utube vids I've looked at seem to position the work at 0,0 giving the forces you describe in the pic 2. Would I be right in thinking that if the work was positioned mid Y (above the ball screws) with x=0 or where ever, it wouldn't be so so much of a problem? And, would that imply that best accuracy / precision would be in the centre of the table rather than at the edges??

Jonathan

13-09-2011, 11:00 PM

No regards torque It's undisputable obviously, but I know from experience with actually using them not calculating or theorising.! . . they have more than enough torque even with the disadvantage.

It's almost impossible to have too much torque since it so directly affects machining and is so useful, and the 20mm screws enable a much greater factor of saftey than the 16mm ones, or higher feeds/acceleration if you choose that way.

When I did the motor tuning on Chip's router we got perfectly good feedrates (can't remember exactly what, long time ago). That's with similar length 16mm screws, but only 5mm pitch so nowhere near as good as 10mm pitch if you're after speed.

If you want to go looking around the zone etc then I think you'll find more unhappy owners of long 16mm screws than you will 20mm.

Surely that is because the vast majority of people use 5mm pitch 16mm screws, which due to my reasons earlier are clearly going to cause grief.

I know if asked to build a machine with 16mm screws of this length I would re-fuse and recommend 20mm with correct size pitch to suit the intend purpose of the machine.

Yes, naturally so would I in an ideal world where it's a reasonable price to get 10mm pitch.

My proposal to use geared 20-05 was work around to no cheaply available 20-10 and wasn't saying they are the ideal solution

Gearing helps but you're still limited by the torque due to the high screw energy and potentially whipping. I will only ever use pulleys anyway as clearly it gets the most optimal torque and has the added bonus of reducing resonance.

I'll take real world experinece over calculated theory any day . . . . IMO it's priceless and feel privilaged when someone shares there often hard learned experineces.

Yes experience is of course great if the situation is similar enough for it to apply. I probably lack in experience to you, but however valuable experience is, experience backed up with mechanics is indisputably a lot more useful.

JAZZCNC

13-09-2011, 11:25 PM

Rick (anyone), A silly question perhaps... most of the utube vids I've looked at seem to position the work at 0,0 giving the forces you describe in the pic 2. Would I be right in thinking that if the work was positioned mid Y (above the ball screws) with x=0 or where ever, it wouldn't be so so much of a problem? And, would that imply that best accuracy / precision would be in the centre of the table rather than at the edges??

Ok understand better now.! If using just a single centrally placed screw then yes the most accurate position is in the centre of the machine and as you approch the outer edges the gantry will flex and twist~(racking) reducing precision.
How much racking depends on several factors but obviously the wider the machine the worse it gets, wide gantry or wide placement of the bearings the gantry sits on helps reduce this to a degree.
Basicly if you have central screw on narrow machine with wide bearing placement then the racking affect can be controlled to acceptable levels. Wide machines are better suited driven from both sides of the gantry using 2 screws.

Regards work positioning I think you'll find when you see a referanceto 0,0 position it will often relate to work coordinates which are often differant to Machine coordinates.

Excuse if you know this.! . . . Machine coordinates define the overall cutting area of the machine. .Work coordinates define a referance point on your material that you want to start work cutting from.
So say you have 400mm * 400mm cutting area with machine coordinate 0,0 defined in the lower left corner (looking down from the top) and place 100mm x 100mm material in the centre with the cutting start point in the lower left corner of the material.
Then the work coordinate is at 0,0 but it's actual position on the machine is at 150,150. Where you position the work 0,0 coordinate will be determined by where abouts you position the 0,0 in you cad program and the part your drawing.
Often Machine coordinates are positioned in the lower left corner so you'll never have negative machine coordinates, so If your work as negative coordinates IE Circle with centre defined as 0,0 work coordinate then you need to offset the 0,0 work coordinate far enough away from 0,0 machine coordinates other wise the machine will crash into the limits.

Hope this helps if i'm teachng dad to suck eggs then i'm sorry.!

JAZZCNC

13-09-2011, 11:41 PM

experience backed up with mechanics is indisputably a lot more useful.

Ok Jonathan we'll have to agree to disagree on some things but I'll agree with you on the above.!
(Provided your experienced in mechanics.! . . I know I am got papers some where that tell me so.:whistling:)

mocha

14-09-2011, 12:13 AM

Thanks for your time with the answer Jazz, most helpful! No, not sucking eggs even slightly, I'm in the happy position of knowing enough to know I don't know enough! :-)

I might have clarified what I meant better by substituting the 0,0 for a "corner of the work area" But, yes, I see what you mean. I'm interested in this thread as the design (Rev 872) I'm playing around with uses steel too, most helpful!

GTJim

14-09-2011, 04:05 PM

Just an idea and probibly silly, but how would this concept work as a rotating ball nut?

4536

Ball screw fixed at each end, pulley machined to fit ball nut (green) and motor mounted to cross brase of gantry under the table with another pulley and belt.

Jim

luke11cnc

14-09-2011, 05:41 PM

I can't see a problem
but your cable and may be vibration on the rail.

maybe box section instead of flat bar

but I'm no expert

James

mocha

14-09-2011, 05:55 PM

if I'm reading the drawing right, is the belt the only thing that will transfer the movement to the gantry, for the gantry to move on the x axis? that would be a problem.

Caveat as above, I'm no expert! :-)

max

Jonathan

14-09-2011, 05:59 PM

I'm afraid that wouldn't work. As soon as you put and tension on the timing belt it will bend the ballscrew which is obviously bad. The ballnut needs to be supported with bearings and it's own mount. The best option is a pair of preloaded angular contact bearings with the pulley in between so that both bearings resist the tensioning force of the belt. This is what I used. Angular contact bearings are best as they accept radial and, importantly, axial loads whereas standard deep groove bearings don't take much axial load. Having said that I've seen one on youtube that uses deep groove bearings, but it's only cutting foam...

if I'm reading the drawing right, is the belt the only thing that will transfer the movement to the gantry, for the gantry to move on the x axis?

Yes!!

Here's mine:

http://www.mycncuk.com/forums/showthread.php/3340-Rotating-Ballnut-design-ideas

I've almost finished designing one similar to that for 16mm and 20mm pitch screws with 7207 bearings.

mocha

14-09-2011, 06:28 PM

yes thanks Jonathan I saw that thread, very interesting and those stepper mounts looked very nice.

Am I right in thinking the advantage of the rotating ballnut is gained on lengths over 100cm?

Jonathan

14-09-2011, 06:36 PM

yes thanks Jonathan I saw that thread, very interesting and those stepper mounts looked very nice.

Thank you... if you've not got the kit to make them yourself then I can make them for you. Let me know if you're interested and I'll work out the price for what you need.

Am I right in thinking the advantage of the rotating ballnut is gained on lengths over 100cm?

The simple answer is it depends. If it's a small diameter screw, or you want a very high feedrate, then yes you'll gain at 1m. See what I put about critical speed in post 39. You do save by not buying the bearing blocks and you can get away without end machining on the screw. This is good since it makes the size/pitch screws that are not on eBay more reasonably priced.
If it's a 25mm screw then the rotating nut setup I made has the same inertia as about 700mm of the screw, so you wont gain anything below that. The critical speed of a 700mm long 25mm diameter screw is very high anyway, so you wouldn't need to use a rotating nut anyway. I've not worked out the moment of inertia of the 16/20mm version yet.