PDA

View Full Version : dual drive x-axis?



fasteddy
07-06-2009, 11:27 AM
Hi all..

Was hoping to drive my x-axis with a dual drive. Anyone know if it is possible to run two motors from a single controller point? I had originally thought of a dual pulley system but it is rigidity and accuracy I am after and not sure that is the way to go..

Really would like to drive twin ball/lead screws on my x-axis (2 meter x-axis x 1.3 y-axis meter work area, floating gantry)..

Any suggestions will be very welcome as this is my first cnc project...

ta..!

John S
07-06-2009, 11:39 AM
Two motors, two drivers and slave the axis.

Two motors on one driver is asking for problems, a belt and pulleys over 1.3M is also asking for problems.

Your size basically restricts your choices.

Using two motors / one drive makes the driver the weakest link which they normally are, any problems will take the drive out so you will have to buy a new drive, that's the cost of doing it correctly in the first place.

.

fasteddy
07-06-2009, 11:43 AM
Thanks John..

Does that mean I will need a 4 axis controller? 1 for the z, 1 for the y and 2 for the x-axis? If so, could you recommend one or do I have to decide on which motors first..? This is my weakest area of non-expertise so any advice would be v helpful..

fasteddy
08-06-2009, 07:12 AM
Sorry everyone, just noticed the thread by Hiltonsteve, 3 below mine, asking exactly the same thing..! There are more than enough answers there to sort me out... Sigh.. Looks like I was wide asleep again!!!

:joker:

fasteddy
09-06-2009, 07:57 AM
Have finally settled on a single motor, single leadscrew, but with extra bearings and rails under the gantry. I know I know... Pictures..! I will try to put a couple of sketches together tonight and post them..

Have listened carefully to everyone's if, buts and maybe's about dual driving and have to agree that a single drive is much less prone to errors, and belts are, in my experience, another added layer of mischief..

But I do think it important to stabilize the gantry as any yaw will not be acceptable. To this end I have come up with a way of using additional track and bearings (as these are so cheap) to give extra guides to the gantry...

Pics later..!

BillTodd
09-06-2009, 01:49 PM
But I do think it important to stabilize the gantry as any yaw will not be acceptable.
What about a cable (or two) wrapped around pulleys (like a drawing board) ?

Smiler
09-06-2009, 06:36 PM
What about a cable (or two) wrapped around pulleys (like a drawing board) ?

Apparently works incredibly well too!

Jeff.

fasteddy
10-06-2009, 08:09 AM
Any pics of that? Will do some sketches of my thoughts later, gotta 3 year old with tonsillitis to sort out today..!

BillTodd
10-06-2009, 12:34 PM
One method:
(pinched from http://www.cnczone.com/forums/showthread.php?t=57030&highlight=cable+rack&page=4)

fasteddy
10-06-2009, 09:01 PM
The cable design looks funky! But I am still unconvinced by using an active system, I would be concerned about slippage, stretching etc.. Not sure if I have uploaded the image properly in this reply, but if I have you will see that I have added two tracks under, and attached to the bed. There are bearing runners attached to the inside of the gantry base that locate onto the tracks.

Have not drawn it yet (have not finished designing it) but the base of the unit (the whole CNC will sit on top of it), will also have another pair of tracks that will fit to the bottom of unit, that will locate to a further pair of bearing runners attached to the underside of the gantry for further stabilization...

Hope this is clear, if not ask away and will try to further clarify it. I like this principle as it is passive and cheap..

Have to apologize for offering the link on my first post of the build videos that first wowed me into all of this... This is the proper link http://buildyourcnc.com/step1.aspx (cant seem to make it a link, having one of those days...) which shows every step of the build in videos (best watched at 2am when web traffic is light, yawn...)

fasteddy
10-06-2009, 09:32 PM
Oops.. Forgot to set my render settings... This is a bigger image...

:redface:

BillTodd
10-06-2009, 09:53 PM
I would be concerned about slippage,
No they can't slip; theres two cables, one from left/rear to right/front the other from right/rear to left/front.

Basically, if say, the left side moves forward, it pulls on the cable which pull the right side forward it the right side moves it pulls the left side.

As long as the cable tension is maintained (e.g. by a spring) the gantry cannot twist or 'rack'.

Ross77
10-06-2009, 10:06 PM
The cable design looks funky! But I am still unconvinced by using an active system, I would be concerned about slippage, stretching etc..

Dont know if this will help but here goes........
The cable system shown looks like it is one continuous cable. I recently had the joy of repairing a Freelander electric window:thumbdown: How ever the way it worked looked to be adaptable for CNC. (now the hard bit)

It used 3 cables. 1 connected the two slides together (via pullys) so they always stay the correct distance apart(i.e. anti-racking) each slide then had a cable going back to the motor. These were attached in opposing directions,(as one wound on the other wound off) In effect the sytem is only ever pulling and cant slip because each end is locked to the drive pully.

The only dificulty is that the distance of travel is governed by the diameter of the pulley and the width (no, of turns). If you increase the width to much then it will cause an arc and introduce an error.

If circumference = pi x D then a 100mm dia. pulley with only 5 turns on it the travel would be 1570mm (if Ive got that correct :rolleyes:) would probably need a belt reduction to the motor to get any reasonable power.

For a bit of stainless cable and a few pulleys its got to be worth a try

Ross77
10-06-2009, 10:09 PM
Bill sorry to repeat what you said, you posted whilst I was writing

fasteddy
10-06-2009, 10:24 PM
Hmm.. Still sounds like a lot of cable, surely stretching is a potential issue? Also additional belt reduction is another layer of complexity?

Bearing in mind this is an ultra cheap mdf jobby (will make the next one on the mdf machine), I was looking for simplicity and ideally a passive system..

Still, if you have actually used this system and got accurate results from it, I will investigate it as another option...

BillTodd
10-06-2009, 10:34 PM
I was looking for simplicity and ideally a passive system..
Well, that's what I was talking about. I.e. the gantry is still controlled by your lead-screw, the two cables just stabilise the gantry.

Ross77
10-06-2009, 10:43 PM
Hmm.. Still sounds like a lot of cable, surely stretching is a potential issue? Also additional belt reduction is another layer of complexity?

I would have fought that the wire would be able to rip the machine apart before streching becomes an issue, I'll have to check the tensile strengh of stainless (got an equation somewhere for calculating stretch)

Belt reduction isnt that much of an issue, some people use it on ballscrews anyway to reduce alignment problems with direct drive.

Sorry didnt realise it was a first machine, Im at the same stage, but converting a pillar drill instead so it can make the parts for the proper one:yahoo:Chicken and egg srings to mind....

fasteddy
11-06-2009, 12:08 AM
Alrighty... Gotta make a start on this project or will end up theorizing forever... Gonna try it my way, if there is any measurable yaw whatsoever, I will add the cable and pulley principle to it as well...! Belt & braces!!! (plus safety pins and a staple gun!).. Also a fan of over-redundancy..!

With that last comment in mind, and a quick reminder that it has to be cheap, I am thinking of a Nema 23 for the z-axis, something a smidge more torquey for the Y, and even more torque for the x-axis.. All this on the principle that the weight driven increases with each axis.. Does that sound right?

If so, would love low cost suggestions for the x and y motors...

HiltonSteve
11-06-2009, 12:43 AM
Just read through the last few posts about pulleys and cable's and must admit after using those old fashioned drawing boards a while a go that it does seem like it could work.

But...

Instead of all the pissing about with pulleys why not make each of your your gantry sides a little wider so that you can space your linear bearings further apart to prevent racking, yes you would loose travel distance but it may be easier, this is what I have done with mine for exactly the same reason. I have also upgraded my X rails to THK 20mm and put two bearings on each side, slightly overkill for what I am doing but better safe than sorry.

How accurate do you want your machine to be?

How much do you not want to spend on your X and Y motors?? and what screw are you think of using??

I know how you feel threorizing which is why I went out and bought a load of parts to build my machine this week before I thought about it anymore, whatever you build either wont be right or you wont be happy wih it but you have to start somewhere. If you do go with the pulley and cable idea then I would be interested to see how it works out as I think it has potential.

:beer:

HiltonSteve
11-06-2009, 01:06 AM
Knock the bottom out of a cardboard box for an idea of how to visualise/prevent/minimise this phenomenon.

I like your theory on this one, it was what I was thinking but maybe I have had a couple of pints too many!:beer:

fasteddy
11-06-2009, 11:02 AM
Hi Hiltonsteve..

Thanks for the advice... To answer some of your questions... I want it as accurate as I can get it using the cheapest, nastiest materials I can get my hands on.. from what I have seen on the net, if you get the design right then a mortgage is not necessary to make a reasonable CNC.. This is why I am picking everyone's grey matter on design and sourcing of electrics..

i intend, once this machine is built, to make the next one on it.. Will be using a lot more ally on the next one so if I can get the prototype rigid and sturdy enough, this should not be a problem... Might even just upgrade the prototype using the prototype..! Its all a bit of an experiment, will be using a lot of the suck it and see methodology..

As for leadscrew, have found Marchant Dice (ebay) do a 16mm x 4mm pitch trapezoidal one that seems cheap enough, although I am not sure about getting the ends machined, might look at knocking up a couple of weldable reducers.. Also got to locate cheap nuts for this thread...

I agree that expanding the gantry sides and extending the bearing rails will eliminate a lot of yaw, but I am trying to utilize as much of the work area as possible, also with a 48" y-axis, the potential for yaw while machining at the extreme ends of the y-axis is increased..

Really try to utilize design as much as possible to avoid having to win the lottery for this one.. More expensive materials, ballscrews, motors etc will overcome most of these problems, but cost is a big issue at the moment.. If the first one comes out ok, I will no doubt put more funds into the next one.. But for now, cheeeeeap!!!!

Am using angle ally and skate bearings for the runners...

fasteddy
11-06-2009, 11:12 AM
Thanks Kip..

Will modify my design to put the additions tracks (under the bed, see my posted cad pic) closer to the gantry edges... I was originally thinking to spread the load a bit, but your comments make more sense...

John S
11-06-2009, 11:12 AM
As for leadscrew, have found Marchant Dice (ebay) do a 16mm x 4mm pitch trapezoidal one that seems cheap enough, although I am not sure about getting the ends machined, might look at knocking up a couple of weldable reducers.. Also got to locate cheap nuts for this thread...

.

Doesn't he do the nuts ?

fasteddy
11-06-2009, 11:17 AM
At a price John, not a nice one either.... Would be cheaper to buy a tap a go visit my mate and his lathe....!

HiltonSteve
11-06-2009, 03:06 PM
At a price John, not a nice one either.... Would be cheaper to buy a tap a go visit my mate and his lathe....!

If you speak to Michael nice I am sure he would knock some nuts up for you. I think he has got a 10 x 4 tap not the 16 x 4 that you are looking at but why not drop down to 10 x 4 trapezoidal screws to save a bit of money, especially if they are temporary and you are going to upgrade the screws to ballscrews later, will be speaking to him in a bit so I will give him a nudge for you.

Tom
11-06-2009, 05:36 PM
Hello Fasteddy,

It sounds like your build's very similar to mine.

I don't want to badmouth Marchant Dice needlessly, and Kevin may still come through with the goods, but i'm struggling to find out what the state of my (overdue) order is.

Anyway, best of luck with your build. I'll update on the leadscrews when I hear more...

Ross77
11-06-2009, 11:01 PM
I agree that you should definatly get on and build it, probably learn more in the process. good luck.

May have retract my statement about cable stretch being low...
Had a look around today and seems that 1m of 3mm dia. stainless cable will stretch 0.5mm under a 200N load (20Kg ,not much if cuting Al and a heavy gantry)....:geek:

May be start a new tread If anyone is intested in following this topic up? Im not used to designing things that move! traditional to keep structures stationary. There must be a way to get cheap and accurate control of CNC machines........

Ross77
11-06-2009, 11:40 PM
Somebody is going to have some fun calculating it all up! :toot: not me though :lol:


I'm a great believer in horses for courses....Milling machines for milling routers for routing......Now engravings a different thing.....It never ends does it :(


I hear you, I wouldnt ever use anything but ballscrews. But I like a challenge and am happy to do the math.



Would the SS cable stretch the same every time? What if it was already tensioned (stretched) ? What if the antistretchyness of the cable was more than the chiploadedness?


Yes the stretch is a product of the load (hookes law not mine) The bit that i dont understand is that if the cable is pre-tensioned and then a load is apllied, is this always an additional load (and therefore more elongation) as the cable is in an elastic state or is only a factor if it is greater than the preload stress?

if its the later then problem solved, if not can you use plastic design on moving parts?.........

Ross77
11-06-2009, 11:48 PM
Somebody is going to have some fun calculating it all up! :toot: not me though :lol:


I'm a great believer in horses for courses....Milling machines for milling routers for routing......Now engravings a different thing.....It never ends does it :(


I hear you, I wouldnt ever use anything but ballscrews. But I like a challenge and am happy to do the math.



Would the SS cable stretch the same every time? What if it was already tensioned (stretched) ? What if the antistretchyness of the cable was more than the chiploadedness?


Yes the stretch is a product of the load (hookes law not mine) The bit that i dont understand is that if the cable is pre-tensioned and then a load is apllied, is this always an additional load (and therefore more elongation) as the cable is in an elastic state or is only a factor if it is greater than the preload stress?

Ross77
12-06-2009, 12:19 AM
Easy...
got another week of college work yet and my brain is still in geek mode



Oi! stop thinking and make something...good, bad, indifferent just make it :D


Collets and ER32 chuck arrived today, so getting there....

fasteddy
12-06-2009, 02:30 AM
Thanks guys, buckets of info to digest...

Cables... Would also be concerned about thermal loading, stretchy in the summer, shrinkage in winter (if you can tell the difference nowadays, where is our global warming gone?).. Would have to insulate my workshop to the nth degree just to minimize this problem...

Nuts.. Damn right they are expensive Hiltonsteve.. Could you hang back on your offer for a few days please? Better get the leadscrew sorted and Tom's point on sourcing it could be a bit of a hurdle.. If I can get to my friends lathe I should be able to screw cut some, 16mm would be easier as a tiny bore would mean a tiny tool, also will reduce screw-whip..

Gonna get all the mdf and angle this weekend and get right into it.. Leadscrews and motors can wait till next week... Could still do with a bit of advice on motors chaps...

Will hopefully fill this post with progress piccies as I do each stage... Loads of thank you's for all your advice guys (are there no women on this forum?) and please keep it all coming, two heads are better than one etc (thats what I keep telling my girlfriend)...

fasteddy
13-06-2009, 10:32 AM
Oh bum....! Have been thoroughly corrupted now having seen this set of videos http://buildyourcnc.com/blacktoe.aspx
Reckon I can do this build as it is very straightforward, also like the vee bearing rails.... Although I am somewhat shocked at the prices of the vee bearings, anyone tried bolting two bearings together and using the chamfers on the od between them as a vee?
Also, as mine will have a 48" y-axis, I am back to the idea of twin motors and leadscrews.. Sigh... Need to get on with it now...
Gonna start building next week..

fasteddy
13-06-2009, 03:06 PM
Yeah, mdf... But doesn't it run sweetly? The designer is an architect and seems to know what he is doing, have also chatted to some of his customers who are utter converts to this design... And of course, it answers my original question about twin leadscrews on the x-axis, the answer being 2 motors and a 4 axis controller..
Got to be worth a bosh for the cost of some mdf and ally angle...

John S
13-06-2009, 08:40 PM
Here you are then Hemsworthlad, job oppertunity for you, no excuse now.

Kip, loads of John S's on the zone :whistle:

.

fasteddy
13-06-2009, 10:33 PM
I am going to try your first 'suggestion'... Bandsaw, couple of sheets of mdf and some bits and bobs.... Dont have metal cutting equipment, am a cabinet maker...
Will let you know how I get on..

John S
13-06-2009, 10:40 PM
Never gave any excuses or need any bud. . . . . . . I Wouldnt rip another blokes idea off just like wouldnt waste good money on that!! . . . . . . . . Mdf is cheap as chips and a few bits of bearings and metal shouldnt cost that much money.
I know the blokes put a bit of work into design but Cmon £728.00 for a 2 sheets of Mdf and £50.00 quids wuff of bits n bobs. . . . . . . Thats a rippoff
and it,s not like it,s a unique design or owt is it. . . . . . Like Kip said its. . Old as the Noah.:nope:

Most routers are, I have seen loads and although some have a unique part the total design is very similar.

Drive systems is where most differ, screw, ballscrew, rack, timing belt etc.

.

John S
13-06-2009, 11:05 PM
I have more toys than Santa, or so Gert says. :naughty:
Problem is remembering where they all are.

True story,
Guy put an advert on the Homeworkshop.org site.

"For sale Rung Fu major milling machine, I bought this machine off John Stevenson about 3 years ago and never used it but took the DRO off to use on another machine, here's John description of it when I bought it ...............etc"

I wondered what happened to that, I thought I'd misplaced it... :rolleyes:

.