PDA

View Full Version : BUILD LOG: George's 8x4 router-me-bob



georgetheforge
21-03-2012, 06:20 PM
hi all,

Am pushing some ideas around on the back of a fag packet- doing some blue sky thinking whilst pushing my performance envelope- etc etc more generic management speak-!

for my gear reduction for the 8 x 4 router- am looking at probably around a 3Nm stepper, running to a 16t timing pulley- to a timing belt- to a 60t pulley with a 20t pinion running on a rack- if that makes sense-!

so questions are-

1/ is a 3 (ish): 1 ratio the best- or is 2 better?
2/ whats the best belt width?
3/ hows best to keep tension/ the pinion on the rack
3/ has anyone come up with a slick way of mounting all this stuff in a nice dust free casing?! am looking at obviously trying to support an 'axle'/ bearings for these and have seen some on other machines but can't find an off the shelf bolt in solution....
4/ would you do the same thing on the 'Y' axis as well?

so far have decided on the hiwin type linear guides/bearings so one each side on the 'x' with the above rack to drive, and for the 'y' probably 2 hiwins on the top and one on the back maybe? probably mounted on an ali extrusion say 80 x 120 again on rack & pinion?

'z' i'm thinking of a ball screw linear table type thing- are these eay/ cheap to build or is it best to purchase an off the shelf type?- But will the 'z' want reduction?

hope to be in a position to start me build in the next couple of weeks so if anyone knows of a cheap but good source for

1/ the hiwin
2/ the rack
3/ the pulleys
4/ steppers/ drives/ etc etc
5/ decent maybe water cooled spindle with a vfd

have been looking around at the usual places zapp etc etc but just on the off chance wondered if there was a guy in a shed somewhere selling them off cheap-!

Gonna start me Cad drawings off and see how i get on...

George

georgetheforge
21-03-2012, 06:41 PM
it appears i have posted this in the wrong section..... embarrassed face mixed with sad face....... can an admin/ mod correct my hideous mistake and shove it in the router build log section... sorry!!

Lee Roberts
21-03-2012, 06:56 PM
Your wish is my command!

Good luck with the build, look forward to seeing the drawings. Not done a belt system myself so cant comment sorry !

.Me

JAZZCNC
21-03-2012, 10:14 PM
Hi George,

I've built a couple of smaller belt drive machines so have a little experience and they work very well.!! . . . . . That said thou let me try and talk you out of it.!!!

It actually doesn't work out much cheaper than R&P or even cheap chinese ballscrews by the time you have bought the pulleys and good quality belt along with bearings,spindles etc. It's really doesn't lend it's self very well to large format machines due to the difficultys in keeping the belts correctly tensioned.

My experence was that even 3:1 wasn't enough reduction to have a good resolution and keep the motors working in a nice rev range.!! . . . And this was with relatively small pinion so smaller Dia = less pitch or travel per rev. . . . IE: 25mm Dia pinion x Pi = 78.5mm travel(pitch) per rev / 3 for 3:1 reduction = 26.1mm per rev.
So thats 25mm drive pulley with 75mm driven pulley.
This is quite a high pitch, the machine will go like hell and well above 30mtr/min but the resolution will be lower.! . . Not so big an issue . . . BUT . . The bigger problem comes when you want to go really slow like 500mm/min this will mean the motors only turning 20 RPM and at this speed the motors complain and you can get cogging affect.

If your only ever going to cut wood or plastics with a good quality spindle then an high pitch will be ok and will handle the high feeds ok.
But if you want better resolution or have a desire to cut materials that require lower feeds then a pitch more like 10-15mm per rev would probably be a better all round target.!!
This would however require 6:1 reduction and great big pulleys which are not cheap. Plus all the above figures revolves around using a small 25mm Dia pinion and what would more likely be the case given the size of machine and belt requirements is that the pinion would be larger which then means the driven pulley gets very large and expensive. (Unless you do multi stage reduction which gets complicated and more expensive.! smaller pulleys but more of them and extra belts)

IMO the best option regards cost/performance/maintenance/ease of build for large format machines is R&P.! . . OR. . If you want to spend a bit more money and time then go for the ultimate and use ballscrews with a rotating ballnut like Jonathan's machine.
My first choice would be Jonathans way every time because of the bennifits ballscrews give but if budget plays a part then I think you'll find that there's very little difference between R&P and belt drive done correctly.

Regards Z Axis I'd build it my self using ballscrews and belts but no reduction 1:1, this is to help with resonance and also makes build and motor placement easier and that bit neater.!
Best tip I'd give is don't skimp on the Z axis because it takes all the strain of cutting forces and no matter how good rest of machine if the Z axis flexs it means nothing.!!

I can help regards linear rails I'll send you a PM.

Hope this helps.!

georgetheforge
21-03-2012, 10:34 PM
Hi Jazz-

Not sure I explained myself very well-!!

Was defiantly using a rack and pinion on both x and y but a ball screw on z- but wanted To do a reduction from the steppers to the pinion gear-

Interested to hear of your sources-!


G

JAZZCNC
21-03-2012, 11:21 PM
Hi Jazz-

Not sure I explained myself very well-!!

Was defiantly using a rack and pinion on both x and y but a ball screw on z- but wanted To do a reduction from the steppers to the pinion gear-

Interested to hear of your sources-!


G

Doh.!!:redface: . . Yes I did read it wrong ok I under stand now.

Get drawing then give some more detail of motors,drives etc.

Are you planning on using 2 slaved motors to drive long Axis or a single motor driving shaft with pinion on each side.?

georgetheforge
22-03-2012, 10:46 AM
was going to have 2 motors working in opposite on the long axis

here's a shot of the kind of reduction i was looking at-

5533

looks like he's just done a slide to manage the tension....

G

JAZZCNC
22-03-2012, 04:07 PM
Looks to me like it's a fixed pinion with no on the fly adjustment while traveling, this is a bad idea to me because it doesn't account for any deviance over the racks length so will bind in places and wear quicker.
Also if something like a decent size chip gets between rack and pinion it will potentially jam up or if traveling slow with high torque do damage as the pinion will try to continue on it's way eating it in the process, something will have to bend or move and chances are it will be the pinion shaft or belt will snap.??? . . .Either way something bust's.!!

Yes even with on the fly adjustment way if something gets stuck between R&P it's going to jump teeth and lose position but no damge will be done to the machine.!!

Another thing I don't like is the fact the belt or what ever he use's for reduction is on the inside in the firing line.!!. . . . not good Idea IMO.!

Like those massive 35mm rails and flanged carriages thou, Shame he's attached it to what looks like a tall 15mm plate gantry side.!!! . . Would like to see the whole thing.

georgetheforge
26-03-2012, 06:53 PM
having a wee look at this tonite-

for the y axis... was thinking of using the 80 x 120 ali extrusion... with one 20mm hiwin rail mounted on the face

a bit like this-? 55615562

dunno if the 2 rails are over kill? for a router that'll mainly be doing sheet materials like ply and mdf with the odd bit of polycarb etc....

the z'll be hiwin type again with ball screw-

JAZZCNC
26-03-2012, 07:03 PM
a bit like this-? 55615562

dunno if the 2 rails are over kill? for a router that'll mainly be doing sheet materials like ply and mdf with the odd bit of polycarb etc....

These links don't work but if your thinking to use just one rail total then thats a bad idea even for MDF, PLY or plastics.

georgetheforge
26-03-2012, 07:41 PM
the forum seems crazy when i do photos!!oh well try again-!i kinda knew i needed 2 but was wondering if i could save a few quid-!!55655566

JAZZCNC
26-03-2012, 07:54 PM
i kinda knew i needed 2 but was wondering if i could save a few quid-!!

Ye everybody thinks it at first, some try it, but all regret it.

I URGE every body who's eager to build but don't have the funds to buy the right components now.!! . . . WAIT . . SAVE UP. . . I guarentee it will save you lots of money down the line but even more importantly make the build go so much smoother and save you lots & lots of heartache and sleepless nights.

georgetheforge
27-03-2012, 04:31 PM
Think the 2 rails is a definate then-!!

would you put them both on the front or as i have in my crude drawing- one top, one face?

also in terms of the Z axis, whats the best design you've seen/ would recommend?

was thinking of linear rail again, with a ball screw but could do with some pointers-!!

or is it easier/ more cost effectie to actually buy one?!!

G

georgetheforge
27-03-2012, 07:05 PM
ohh a bit of time on google turns out bought ones are hideously expensive-!

so now am looking for some (i assume) zero backlash ballscrew in maybe a 16mm? was gonna do 20mm hiwin type with the carriages fixed and the rail sliding to give more slide,

so i figure i need a motor mount, a flexi couple, the ball screw support, the nut, the nut housing, and the end support?

whats the best pitch?, and will 16mm be ok-

also anyone know of a supplier that has decent drawings/ downloads available?

getting there- only problem is work getting in the way-!!

George-

JAZZCNC
27-03-2012, 09:00 PM
Here you go George have a look at these. This is one way and my prefered way to build them.
The Obviously the gantry is not showing but the Y axis rails could go either way like I have them or both on front or 1 on top 1 on front but this way I prefer because it brings Z axis closer to gantry, increases travel and is very strong. The belts help with resonance and allow gearing up or down if needed.
I use 16mm screws with 5mm pitch, sometime I use BK/BF bearing blocks some times FF/FK just depends on what I'm building and have even been knownt tomachine my own blocks but can't be arsed anymore as the chinese prices are so low it's not worth the trouble.

These are just reference models so not 100% accurate regards alignment etc but all materials and linear rails, motors etc are the correct and actual size and match whats available.

Hope this helps.

georgetheforge
28-03-2012, 08:28 AM
ohh nice drawing Jazz-

see what you mean re the linear rail placement/ getting the spindle as close as poss-

what reduction did you do on the belt-? 2:1?

saw one from marchant dice for £1400....!!

Thanks again for all your help- it's really appreciated-!

now that i basically know what i'm gonna do will do a bill of materials to see what the damage will be-!

Thanks

George

JAZZCNC
28-03-2012, 09:32 AM
Hi George,

There's no reduction it's 1:1. The main reason for the belt is to reduce resonance and also allow easier positioning, the gearing is just a bonus if needed.
The previous drawing shows the motor up high but I also make them with the motor reversed, it just depends on the gantry design really. If possible I use the other position as it keeps the centre of gravity lower(no big deal) and it looks neater. I'll let actual real picks explain they show it better, ignore the bracket that looks like num 1 thats machine specific.

Tip regards material be very carefull when selecting or buying the Ali plate that it's perfectly flat. . . It's often not.!! Often it's cupped across it's width or even twisted in it's length.
If you look close at the pics you'll see I machine all the bearing mounting surfaces. I do this for 2 reasons.
#1 It gives me a register for the master bearings and rail to butt against which makes setting rails parallel easier and allows for easy replacement/setup if stripdown/cleaning required.

#2 Very important.!!. . If you try to bolt rails or bearings to material thats not flat then you get binding in the bearings increasing friction and causing premature wear, if bad it will lock them up soild when tightening.
Linear profiled rails are by far the best for performance and longevity but like all high performance things are tempromental and in tolerent of shoddy inferior material or poor workmanship so therefore require higher accurecy.
If you don't have the capabilty to machine surfaces flat then I strongly urge you to buy ground flat plate like the ECO CAST from Alu warehouse it's more costly but will save you a world of pain and give much less friction and smoother action. This is what I do now as the little extra it cost's is worth the expense as it saves me many hours machining and I just machine insert pins for master registers and machine the edge.

Making one with linear profiled rails, 16mm ballscrew and 20mm ground flat plate along with pulleys etc will cost you around £310 and will handle anything you throw at it for millions of years at DIY level.
Skimping on a Z axis is just plane dumb IMO if you want a accurate machine as it handles all the cutting forces.!!

georgetheforge
29-03-2012, 01:37 PM
Thanks for those Jazz-

am looking into the pros and cons of using ball screw type motion for the x and y as well now- as it looks easier to set up than R&P-


so with that in mind does the panel agree that screws are more accurate/ easier to mount etc- ie less faffing to get the tension/ pressure on the rack/ pinion?

for the 'long' x axis (2500 ish) you'd just have the end block, the ball screw+ mounting block and the machined end bearing- plus the screw obviously-! (one set per side)

would you go straight on with the stepper or do a belt?

whose the best supplier for these- i imagine Zero backlash is what i'm after?

hope everybody's well

G

JAZZCNC
29-03-2012, 03:57 PM
so with that in mind does the panel agree that screws are more accurate/ easier to mount etc- ie less faffing to get the tension/ pressure on the rack/ pinion?

YES


for the 'long' x axis (2500 ish) you'd just have the end block, the ball screw+ mounting block and the machined end bearing- plus the screw obviously-! (one set per side)

Your problem here with a ballscrew is the length, 2.5mtr is too long in my experience if unless you use a rotating ballnut.!! (Ask Jonathan regards rotating nuts)


would you go straight on with the stepper or do a belt?

I prefer using short timing belts to connect motors to help with resonance and allow easy gearing options.

georgetheforge
29-03-2012, 04:15 PM
yeah i was looking at Jonathans post/ thread on this issue- looks like an interesting idea-!

is the problem that it can 'twist' in it's on length- or just it flapping about in the breeze getting its wobble on?!

my slight problem in this build will be the machining of the parts etc, as i'm yet to get a mill and lathe- but want.... hence my initial though of R&P.

i guess it's a toss up between the 2 as the ball screw method may be costly compared to a realtively cheap r & p...

keep going round and round and round with these decisions- sorry to keep asking pretty much the same questions over and over-!!

G

JAZZCNC
29-03-2012, 04:30 PM
George please don't appologise and keeping asking the questions untill your satisfied you fully understand.

The problem with long screws is whip at higher speeds. The only way to avoid it is to use large diameter screws but these bring there own set of problems, like needing BIG motors, drives, PSU Etc or moving over to servo's and gearing etc !!!. . . .all expensive.

Personally I prefer ballscrews over R&P anyday and really it can't live with them in any department but they are not as simple to impliment with a long axis machine.

If it was me then it would be a no brainer and it would be getting rotating ballnuts without a doubt.

Regards parts machining then ask around several here will do them for you I'm sure. I would and will glady make them for you but unfortunatly I'm tied up for the next several weeks but if you can wait then I'd be happy to help.

georgetheforge
29-03-2012, 04:39 PM
i'll get there at some point-!!

came across these badgers- bet they're worth a pretty penny-! http://www.drives.co.uk/fullstory.asp?id=3388

good magazine if you get it? might also go to the trade show in april...

thing is i have the skills to machine them just not the machines-!! i'll have a play with the drawings etc and see where i get too

would you run 2, one at each side? or do you run into sync issues?

G

JAZZCNC
29-03-2012, 04:50 PM
Ye you'll have to sell your house and the wife to buy those mate. . Lol

Yes absolutly run 2 screws and provided you don't over tune them they keep sync ok-ish.!! . . . I have my views on the sync issue and to be honest I won't go into it again in your thread but if you look around at some of the other post's on here I've gave them recently but I can't remember which thread.!!
(This one amongest others I think.?? http://www.mycncuk.com/forums/showthread.php/4327-I-want-to-DIY-some-questions-please!)

Jonathan
30-03-2012, 12:34 AM
When I started out with my machine I did have issues with motors stalling on the X-axis, but really that was just me trying to push it too much. Now I've been a bit more conservative (still 10m/min which is fast by most peoples standards) the machine hasn't stalled for several months so that's not an issue.

I'm going to be making a couple of rotating ballnut mounts soon, so let me know if you're interested and I'll work out the price as it'll be more efficient to make several at the same time. I've updated the design to be fully enclosed with aluminium such that no swarf can get in. Something like this, can of course modify it to suit / aid mounting:

5595

georgetheforge
02-04-2012, 09:38 AM
Hi y'all, spent a bit of the weekend messing with my drawings-

to see what i could come up with-! bit of a mash up of my lurking about here and there-!! to be fair mainly Jazz!

so the plan is as pics below- however.....unsure of the Z as I've allowed for 150mm movement/ gantry clearance, might be over the top, but i get asked for odd things and it might be good to allow that tolerance for poly etc- had half a though of making the 20mm side plate adaptable with another set of holes for the carriages @ say 70mm up so you could drop the gantry-

need to think about the rotating ball screw affair/ speak to Jonathan-(pm coming your way-!) current thinking is to have them below the 120x 80 edge frame connected to the 20 odd mm side plates- and whether the y ball screw is too far back off the spindle- seemed like a nice neat way at the time- feel free to comment-!

also work out belt lengths for the y and z to make sure i can get something to 'fit' but will probably get the bits and 'measure' them before any plates are machined (for which i need a friendly person to quote on as theres some trenching/ holes to accurately cnc..chicken and egg- 'oh if i had a cnc mill i could.....!'

the bed is currently long enough to have the gantry go to a dead space at the end which hopefully will help with loading etc- and is currently 25mm mdf which i hope to make into a vac bed of sorts when budget allows-!

rest of the table frame/ cable management / extraction to draw up but cant see any probs with that but will draw them before i start building to make sure the 'problems' are thought about-!

gantry will probs be a the same 120x 80 x 6mm RHS..... thoughts?Quite exciting really.... now to collate me list of parts/ electronics ( which scares me slightly!)

5623

5624

5625

5626

hope that all makes sense?George

JAZZCNC
02-04-2012, 11:25 AM
hope that all makes sense?George

Makes perfect sense George.!

Some will complain the Ballscrew is too far back from the Y axis carriage, which it is slightly, but because you are using profiled rails they will help absorb some of the effects. Thou be aware this will apply twisting forces to the bearings which over time could cause premature wear.? You may want to consider moving the screw forward and higher sitting directly on top of the upper bearing plate.?

When you say 150mm of clearence between gantry is that from the Bed to the underside of gantry.? If so then again you may want to consider upping this if your wanting flexabilty.
The things that tend to cause the most problems regards clearence are drills and clamping devices.!! Often drills have long length's so by the time you have material, sacrificial base material then clamps etc you don't much left for clearence.
Also if you ever want to work with really small items you may want to add a vice to hold them and with only 150mm of clearence then again it's very tight even with normal endmills.

150mm is a good Z axis travel amount but really I would look to more clearence if your wanting flexabilty. It's very easy to pack the work material up towards the cutter but not so easy to move the gantry up.!!!

Jonathan
02-04-2012, 12:03 PM
Some will complain the Ballscrew is too far back from the Y axis carriage, which it is slightly, but because you are using profiled rails they will help absorb some of the effects.

I was about to complain... it's true that you'll probably get away with it because you're using profile rails, but its unwise to rely on one very strong component to compensate for suboptimal design elsewhere.

20mm for the side plates for a machine this size sounds a bit small. To compensate I would make the 4 triangular brackets extend all the way down the sides as that will really stiffen things up.

The alternative is to eliminate the gantry sides completely, like on mine (or Jazz's) machine:

5627
(Ignore the rotary table madness and yes my Y-ballscrew is badly positioned)

It's close to the strongest format.

Consider carefully what the profile rails on the X-axis are mounted to. Since you have orientated them with the mounting faces pointing outwards (i.e. in the YZ plane) the distance between the rails is critical, so you must have minimum one of the box sections bolted down, not welded, so that it can be adjusted. If you go for the gantry format in the picture above the rails are both facing up which would make this no longer significant.
Same is true for the Y-axis. 120*80*6 box section is a good size, but by putting the rails on opposite faces you're relying on those faces of the box section being parallel. To the tolerances required for profile rails they wont be, so you'll probably end up needing to use shims and stuff to get it to run smoothly. With supported round rails the tolerances are lower, so you get away with it, and you want them in that configuration to even up the loading on each block. However for profile rails (with 4 rows of balls which are the only ones worth getting) the force rating is equal in all directions, so it wont make much difference if you mount them facing or in the same plane. Mounting them in the same plane will make life easier ... lends itself well to getting the Y-axis ballscrew in the correct position - right between the rails.

georgetheforge
02-04-2012, 02:28 PM
Thanks chaps-

What kinda clearance would you recommend?- obviously depends on what I'll be machining-!!

mainly 18/ 25mm ply and mdf, some poly/ foams maybe and maybe some ali if possible....

I wanted the side plates- rather than your solution- so i could keep the top surface clear for loading etc- i do have little triangular brackets which i forgot to make bigger when i was playing with the gantry height-

was tepted to 'split' the gantry so i had 2 sections with the ball screw between them fitted to the back of the front/ z plate. obviously that makes it heigher and was worried about twist/ aligning them correctly.... i did look at getting some 100 x 200 x 10mm RHS and getting the front lasered out to make a so a proper c section rather than an channel- almost like a chunky Unistrut....? then reverting to plan a having one of the rails on the front- bottom and one on the top.....?

georgetheforge
10-04-2012, 07:27 PM
so heres the result of some playing around with the drawing-

main changes were thinking how to integrate the rotating ball nut- the idea is to sling it under the x rail ( which will bolt to the side of the table, so hopefully there will be some adjustment on the profiled rails. Ideally they'd sit on another tube to take the weight rather than just be on the bolts- any suggestions?!

i split the y axis and am planning on using 80 x 80 x 8 with the ball screw between the two, fixed to the front and back of y carriage, but this stops any bracing between the 2 rails, do you think this will be a problem?

i extended the z and moved the y upo to give more clearance- now 230mm travel which has meant massive side plates for the gantry-

jonathan, do you think the roating nuts will sit nicely where i've got them or do you think it's hideous-!

anyway- thoughts on this before i start ordering stuff would be awesome


http://www.youtube.com/watch?v=dDU5Elw923c

Jonathan
11-04-2012, 10:58 PM
Splitting the gantry into two box sections as you have done is a great way to do it. I'd put an aluminium plate on the back to stiffen it further, although it's probably fine as it is.

The gantry sides just look flimsy. Since you don't have an adjustable height bed most of the time the Z-axis will be operating near maximum travel, so the combined tool deflection from Z-axis and gantry bending will significantly limit the feed-rate and finish you can obtain. You should do what I put in the last post - eliminate the gantry sides and have the box sections for Y on the smallest plate, or box section, viable to link them to the X-axis bearings. Add to that an adjustable heigh bed and you'll have a really strong machine with as much clearance as you want for long tools/drill etc.

You need to consider how to accurately align the X-axis profile rails. Currently you can get them both parallel in the X/Z plane, but the spacing of the rails (i.e. parallelism in the X/Y plane) is fixed. No frame will be accurate enough to allow that unless you build in some adjustment, or fix the rails in the same plane as the bed (like on mine).

Correct me if I'm wrong but you've drawn the rotating ballnut mounts fixed to one end of the bed, not fixed to the gantry sides?

georgetheforge
12-04-2012, 01:38 PM
Hey Jonathan,

I did think about plating the back, but currently have the ballnut for the Y carriage elongated to attach to the both the back and front plates, I'll have a look at the gantry side plates, i did want to add in some angled brackets, but will look at your design again,

In terms of alignment, i was planning on bolting the 'x' rails to the 120x 80 and then bolting this through the table so there was some adjustment available with shims/ rizzla papers if required-! but agin may re think.

The rotating ball nuts are currently fixed to the carriage side plates with the 'fixed' bearings at each end...they just look similar to your ball nut design...

heres another rendering-


http://youtu.be/MB7oJDi9rWs

C.AlveSilva
17-04-2012, 09:21 PM
...I'm going to be making a couple of rotating ballnut mounts soon, so let me know if you're interested and I'll work out the price as it'll be more efficient to make several at the same time. I've updated the design to be fully enclosed with aluminium such that no swarf can get in. Something like this, can of course modify it to suit / aid mounting:

5595

Hello,

I'm probably interested ... I am considering using this system instead of the traditional.
In my project I'm planning to use 2005 Ballscrews with Nema 23, (3.1 or 4 Nm).
Already know something about the cost?

georgetheforge
18-04-2012, 04:39 PM
So... an update on this....

think i'm going to have a rethink on the side plates as per Jonathan's points to bulk them up, and maybe look at raising the sides where the x rails/ ball screws live, i know the other thread regarding the loading of materials etc, but mine will go in a place where it wont be that easy to end load, a corner load would be preferable-!

what was the consensus on having a hybrid, so flush on one side and risen on the other- think that would work well for me?

have costed most of the bits- and reckon on a budget of £2k all in should get me something decent?

Jonathan
21-04-2012, 11:11 PM
I'm probably interested ... I am considering using this system instead of the traditional.
In my project I'm planning to use 2005 Ballscrews with Nema 23, (3.1 or 4 Nm).

If the machine is big enough to consider rotating ballnuts you do not want to use less than 10mm pitch screws.
I can make some more, however I'm currently busy making another machine and other parts so it's a matter of how soon you need them?