Accurate Strong Gantry

Hello Everybody,

My target is to built a gantry for home use since I do office job and miss the miller work. (I have three sons so hope they follow the way)
The travel is 1000 x 700 x 250 mm with 1000 N cutting force in Plastic, AL, with 0,05 mm total accuracy.
After studying the relevant threads some contradiction still need to be solved.
I.e. my friend who build some CNC machines(not like this) says you only need one ball screw in the middle.
There are many ways but the proved, reported solution is my solution.
I hope with your help and skill I can reach my goal so please correct me where I am wrong.
Base is not finished in the attached figures.
My queries:

1. 1500 mm long rail position-side or top ?
2. Is there any good carrige pattern to eliminate the double ball screw?
3. At belt synchronizing version1 or version2 is the better? -my logic says the equal belt length in any drive direction give more accuracy . http://www.mycncuk.com/attachment.ph...id=13921&stc=1http://www.mycncuk.com/attachment.ph...id=13922&stc=1http://www.mycncuk.com/attachment.ph...id=13923&stc=1

Thank for your cooperation
BR
István

Quote:

---

Originally Posted by vargai

1. 1500 mm long rail position-side or top ?
2. Is there any good carrige pattern to eliminate the double ball screw?
3. At belt synchronizing version1 or version2 is the better? -my logic says the equal belt length in any drive direction give more accuracy .

---

1. As far as I'm aware the bearings and rail do not care which way you have them but you need to think about how you will mount them so that they are level with each other across the frame, and that they are the same distance apart along the frame.
2. On a 700 wide machine I think you will need 2 ball screws.
3. As you say, the equal belt length method looks best.

Thanks form the prompt answer,
[QUOTE
3. As you say, the equal belt length method looks best.[/QUOTE]

When I am saying equal length I am thinking of the extra synchronizing belt that has the same upper and lower length .
In the lower figure the drive is symmetrical so I am confused which of you mean the best

Quote:

---

Originally Posted by EddyCurrent

1. As far as I'm aware the bearings and rail do not care which way you have them but you need to think about how you will mount them so that they are level with each other across the frame, and that they are the same distance apart along the frame.

---

I am planning to machine the base in one operation with bearing in mind the cost so at the moment side mount seems stronger and safer for the rail

Quote:

---

Originally Posted by vargai

Thanks form the prompt answer,

3. As you say, the equal belt length method looks best.
When I am saying equal length I am thinking of the extra synchronizing belt that has the same upper and lower length .
In the lower figure the drive is symmetrical so I am confused which of you mean the best

---

Version 2 where the motor is central.

The price of the belt and pulleys will be same as the price for one driver more. So if you invest in one more motor you would have it properly made. Not that it would not work like that with one motor.

Thanks your comment, because this is one of the most critical issue here. You are aware of JazzCNC's intention namely two motor can cause big problem at synchronizing failure .
Just example for calculation with overestimated values:On a 1 m long belt 1 mm elongation on a perimeter of a 360 mm pulley gives 1 degree delay- derived to the 20/10 ball screw means 10 mm/360=0,027 mm deviation-looks not bad
But this is just theory again-reported data overrides the best theory.
I am planning to use 200-300 W servo here and master slave drive is quite expensive.
I saw you used two motors on your just made machine-I think steppers that gives torque protection but with potential step loss during operation
You know the target is 1000 N cutting force.
I need to see about this issue in the threads and your opinion is appreciated in the belt versus dual motor question.
Br
István

Hi, calculations are one thing, real life another. I have belt driven machine on all axis. So i believe i know something about belts. The only thing positive in this design is that the belt is perpendicular to the ball screw, means its not driving directly the machine but transferring the movement.

What size type belt and pulleys?

How heavy is the gantry with all motors, rails and so? Not sure if you could drive it with one 300w servo at all.

I know Dean /JazzCNC/ is driving his machine like this, but you have to make sure about the proper belt size, corresponding motor and make sure that the accumulated inertia of the pulleys and ball screws will not limit the machine speed.

And looking at 0.05mm means tthat you must strive at 0.01mm. That means cheap chinese pulleys will NOT work. Have that in mind. And if you need to lower the inertia you must buy expensive aluminum pulleys or custom made that small size. I have 20 crap pulleys laying around cause i thought they were a deal from aliexpress.

PS. It would be better if the spindle center is inside the bearings that move the gantry looked from side. Means you should offset a bit more the gantry to the back. You have to imagine that when all is mounted and weights what it weights the center of the weight must be in the center between the 2 bearing that move the gantry.

understood what you mean, just wanted to reach the perimeter of the machine table with the tool to have access for edge drilling that is advantage at high work piece -but this is not main issue.
When I think of a standard VMC the real heavy consoled spindle house running up and down on a vertical rails with about 400 mm or more offset and the bearings stands it- it is their job.
I am using 25 mm bearings that is IMO strong enough for normal operation, however I will consider your suggestion-every detail can be important.
Perhaps it will help to support the rail side form above against the constant torque and lowering the console lever according to your suggestion

Quote:

---

Originally Posted by vargai

When I think of a standard VMC the real heavy consoled spindle house running up and down on a vertical rails with about 400 mm or more offset and the bearings stands it- it is their job.

---

Yeah, but with sliding counter weight at the back to balance the weight of the head so at the end the bearings have less push when static than a gantry cnc

belt is common and useful item (coupling) between motors and ball screw reducing the dynamics and revolution with effect of the inertia -I saw it in the Hurco VMC as well that a had worked on earlier.
Short belt is accurate:long synchronizing belt -I do not know . The only thing is sure - producer gives 0,1-0,2 mm accuracy on a belt driven linear unit on 1 m long.(so the elongation is the accuracy)

You are right double ball screw means double inertia and its effect to the required power fourfold.
That is why I am here to overcome this type of doubt , so thank you to share your experience

According to the charts HTD 5M is well in the charts for what you will use it for. I would say 25mm wide and steel corded. I wouldn't go with other braids cause Kevlar and so stretches. Pulleys should be 20 - 22 teeth. Make the ball screws placement and design so that they are only long for the exact necessary travel. Dont overextend them cause its easier to mount or for other reason. Pulleys should have fitted bronze bushing and the bolts/shafts should be proper ones. Shoulder bolts if you use bolts and steel shafts, don't leave them to rotate without bearing in between.

PS, some additional info:
http://www.mycncuk.com/attachment.ph...id=13925&stc=1
http://www.mycncuk.com/attachment.ph...id=13926&stc=1


http://www.mycncuk.com/attachment.ph...id=13927&stc=1

I am a fan of taper locks or similar shaft lock- simple fast and not expensive - is it good without parallel key?

Then you will need dial indicator and mess around just to center it properly on the shaft. Cause they are not self centering to the needed tolerance. Plus are heavy. Best would be tight fit and you or somebody localy to machine you the pulleys from aluminum.

thanks, already I can go further with design
BR
István

Quote:

---

Originally Posted by silyavski

So if you invest in one more motor you would have it properly made. Not that it would not work like that with one motor.

---

It is properly made and it actually works better than twin motors being more accurate and reliable if done correctly.
My machine use this setup and I'm just going to upgrade and replace worn components and seen has this means new custom length belts and new gantry setup etc then it would be perfect opertunity to change to twin motor setup if it was better.? . . I can tell you now the thought never entered my head and I wouldn't entertain swapping to twin motors unless I could use servo's or Hybrid servo system with encoder feedback.

This doesn't mean Twin motor setup is bad or worse because it isn't if setup and run well within the motors speed curve. What it does offer or mean is piece of mind that you never loose sync or have racking issues. No chance of damage to screws if one motor stalls at high feeds etc. Easy to setup and square gantry and never changes after belts have settled down.
Less electronics to deal with etc and while more mechanical items they are very low maintenance and easily available.

There are very few negatives but with some Big positives and if I had any negs it would be belts look messy with more initial work involved and need covering but other than that nothing really and when settled down after a few weeks running they are fit and forget other than yearly maintenance check.

What i meant was, lets say it in other words- that if you don't have a machine shop at home or access to one, it seems more easy to do it with 2 motors, cause no money will be saved buying all that stuff from internet, postage, custom bores and so, plus later the very careful alignment. It could pay off if 1 servo instead of 2, money wise, but like you say Dean, servo is the good way to go with 2 motors. So at the end is just a personal preference. I simply don't like that long belts. In reality my small machine works in a lot of dust and never had any problem with the belts.

Quote:

---

Originally Posted by silyavski

So at the end is just a personal preference. I simply don't like that long belts. In reality my small machine works in a lot of dust and never had any problem with the belts.

---

Yes Boyan it really does boil down to preference and components used not really so much money. My machine doesn't use belts based on money it's soley down to accurecy and repeatabilty/reliabilty within a budget based on using a stepper system.
Any machine using closed loop or some form of error catching setup Ie Servo's/encoders then I wouldn't use single motor and long belts etc.

Regards belt accuracy etc then to any belt/pulley doubters out there then just see this made with a 1300mm wide router with medium weight gantry.!

https://www.youtube.com/watch?v=wjDvimSdH8M

Silyavski, Dean,

Thanks for the comments.
This is the case when everybody has right.The proof of the pudding is in the eating. Dean likes his pudding and you like yours.
I will do my one and will taste it.
I am going to install two screws anyway and link them with belt-it seems more simple now than two servos-later it can be upgraded

Hi,
Incredibly detailed piece, very nice done.
The patterns are so sharp even in a mm .
What sort of material is it?

Quote:

---

Originally Posted by silyavski

What i meant was, lets say it in other words- that if you don't have a machine shop at home or access to one, it seems more easy to do it with 2 motors, cause no money will be saved buying all that stuff from internet, postage, custom bores and so, plus later the very careful alignment. It could pay off if 1 servo instead of 2, money wise, but like you say Dean, servo is the good way to go with 2 motors. So at the end is just a personal preference. I simply don't like that long belts. In reality my small machine works in a lot of dust and never had any problem with the belts.

---

I red about dual drive on other sites and heard of several complaints- homing, running faiulre, etc. I also do not say it is wrong rather say it is an elegant solution but like anything it can get out of order.
If it is happens I can not find on heaven and earth in Hungary expert to repair or without being robbed.
So that is way I get closer to the things from the simple side and it will be a future development.

Quote:

---

Originally Posted by vargai

Hi,
Incredibly detailed piece, very nice done.
The patterns are so sharp even in a mm .
What sort of material is it?

---

Aluminium 6082T6 cut with a simple 90degree spot drill nothing fancy.

Regards the belts and cutting conditions let me just say my machine 99% of the time it's been built has cut nothing but aluminium and the belts are completely uncovered being exposed to millions of flying sharp and very hot chips so are working in the worst enviroment and I've only ever had to replace two belts. This wasn't because of wear it was because I left a pulley loose which fell off after adjustments for a bracket for sensing belt breakage. (It was a very good real world test of my new belt E-stop.!:ambivalence:)
The other I trapped and damaged moving the machine due to not being covered. (So do as I say, not do as I DO, and cover your belts from the start other wise it never get done.!!)

at first sight thought painted wooden, but it cannot be done so sharp

Quote:

---

Originally Posted by vargai

at first sight thought painted wooden, but it cannot be done so sharp

---

Yes probably because of poor video done with phone camera.! . . .Looks much better in real life and very shiny.

I have a similar set up just 90 mm wider. I would not do it again, setting it up without backlash is a pain.

Next one wil have a dual stepper, a leadshine easy servo, which is a stepper with encoder.

In this design I would box up the 4 beams so they become one, that is much stiffer.

Hi Sven,

Yes , the four box should be redesigned somehow, originally I wanted to tie the boxes with 3, 4 plates to work together after placing the
ball screw
What does the similar exactly-how did you make the synchronising?
Nothing is written in stone-just thought to skip the step with stepper and make higher grade machine but do not know too much about stepper.
I hear of lead shine and certainly stepper technology becomes better so if there is an reliable option instead the more expensive servo why not.
My friend who is member of a CNC DIY group made a 5axis machine with steppers for woodwork upon order.
They mates from the group says: everything is ok but why did not you use servos.
Probably stepper is undervalued in public awareness.

I have the second setup in your diagram, with the motor in the middle. It is impossible to get enough tension in the belt without putting big loads in the spindle bearings.
I added a guide to push the middle section of the belt above the motor to reduce floppiness. still not good.

So you are a doubter of using belt based on your experience.
Others have well working machine with this set up.
I am just an outsider at the moment, the best thing is to ask Dean to show a short video of a working machine .
Would you be so kind to do it, Dean?

Quote:

---

Originally Posted by Sven

I have the second set up in your diagram, with the motor in the middle. It is impossible to get enough tension in the belt without putting big loads in the spindle bearings.
I added a guide to push the middle section of the belt above the motor to reduce floppiness. still not good.

---

Let me know about the data bearing type, size, belt type, size and experienced phenomenon
angular bearing on drive side it can bear more load than single row
20/10 ball screw with Dia 15 machined end -bearing can hold let's say max 1 Tons here

Quote:

---

Originally Posted by vargai

Let me know about the data bearing type, size, belt type, size and experienced phenomenon
angular bearing on drive side it can bear more load than single row
20/10 ball screw with Dia 15 machined end -bearing can hold let's say max 1 Tons here

---

Hey,
no tons here!! You have to read well the belt documentation! Find the belt tension formula on Gates web, calculate the tension for your exact setup and then using guitar tuner / or app on your phone/ tension the belt to the proper tension . This is a must to do it properly.

So the force you apply is not tons. Lets say its the force i can apply with one hand and using as a lever say 15mm screw driver against the motor and the frame. With say 20-25cm screw driver i exceed that force and over tighten the belt easily. Of course i have done it at the beginning with tuner, but now just know how a properly tensioned belt feels and sounds.

That means you have to design it so that you could lever somehow the motor with a wrench , screw driver or make some tightening mechanism. That does not mean introducing unnecessary idlers and tensioners.

Quote:

---

Originally Posted by Sven

I have the second setup in your diagram, with the motor in the middle. It is impossible to get enough tension in the belt without putting big loads in the spindle bearings.
I added a guide to push the middle section of the belt above the motor to reduce floppiness. still not good.

---

Now if you have read what i posted before and belt documentation may be you don't have the proper belt cord or you have not designed it to be tensioned properly. I tension the belts with one hand only and a screwdriver or a wrench, using them as a lever against the frame. So if you don't have where to lever them, drill a hole and fix something to be used as a support for the lever.

So are you idlers bigger than 20 tooth pulley as per belts specification? Cause if they are smaller, misaligned or whatever not properly followed, may be thats the problem.

Yeah , i advised against long belts exactly cause people don't bother to read carefully 300 pages belts manual and maybe they miss something or do some basic mistake. Yeah, Deans machine works, but he has knowledge and experience so his machine is an example that there is no problem with the design, not that if say you, me or sb else could do it with the same success.

I will give you an example - Honestly how many people know point 6 below. About the idlers size and placement. Most people i have seen on the forums think that they could place idlers where they want and 99% of the idlers are made by 3x 608 bearings stuck together, which in diameter a much smaller than 20 tooth pulley/which is ok for 10mm wide belts but not 25mm as is this case/


Also look at number 4. The machine sides should be stiff enough.




What i mean is RTFM, do everything properly and all will be smiles and sunshine.


http://www.mycncuk.com/attachment.ph...id=13940&stc=1

Quote:

---

Originally Posted by silyavski

Hey,
no tons here!! You have to read well the belt documentation!

---

"bearing can hold let's say max 1 Tons here mean only it has a capability to hold it but I have never said there would be so big load.
I just want to understand Sven's problem.
At the moment I am watching and listening to collect the proofed working solutions in order to make a good design.
It is useful information that you shared, I wish I were there in building
I read and will read any documentation what is necessary.
Thanks you more and more people knows about point 6.

Quote:

---

Originally Posted by Sven

Next one wil have a dual stepper, a leadshine easy servo, which is a stepper with encoder.

---

Hi Sven,
I read a short 20 page long description of leadshine easy servo and did not find mentions of dual purpose.
Can you point to it ie. link or so?
Thanks
István

Quote:

---

Originally Posted by vargai

Hi Sven,
I read a short 20 page long description of leadshine easy servo and did not find mentions of dual purpose.
Can you point to it ie. link or so?
Thanks
István

---

Is this the one you are thinking about as shown by Dean :- https://www.youtube.com/watch?v=ZMZdCcLQc4M

Thanks Clive,
I have finished watching it with sound off, complete it later.
As far as I know two standard steppers can simultaneously operate two ball screws
Hope listening to the narrator will make clear everything with easy servos

I was looking into the Hybrid servo from leadshine. There were some things that did not convince me:
-the price of drives and motors and all below combined cause at that price i jumped at top 400W second hand AC servos with drives.
-No AC, so still need a transformer, so rectifier circuit, ~45eur,transformer 80-90euro, time, shipping etc.
-basically the characteristics which are no better than normal stepper
-low encoder count

They could have gained me with absolute encoder for example.


I am not saying they are bad, just that according to me, and in my particular case, the servos seemed better option or stay with steppers. Still time will show, when the machine is ready.

Quote:

---

Originally Posted by silyavski

Still time will show, when the machine is ready.

---

Thanks, yes I am not planning machine to motor, in fact I am planning the suit motor to the machine , but first I have to reorder the frame because
learned my expectation is to idealistic:angel: with this gantry desing generally.So I get back to the ground:ambivalence:

Hi,
I would like to know where is limit in weight with the moving gantry that can be drive on reasonable way.
My redesigned one is still 100 kg made of steel with a 18 kg motor just for cheking.
I don't think I can go below 80-90 kg.
I don't care if it requires 20-50 W more just pls advice if this is over sized for a general strong gantry
Thanks

The general opinion is that if >70kg you can not move it adequately with 2x3nm motors so you need bigger motors, which now puts you into position to find bigger but at the same time fast enough or go with servo, which to me seemed the better option.

Now i am looking back at your gantry design. What size are the profiles, why 4, why are not connected each other? Whats your idea there? At that size machine
2x 100x100x3 or 100x100x4 welded together with separation for ball screw or not /if back/ will be ideal. 100x100x3 weights ~9kg/m and the 4mm weights 12kg/m, So you co8uld still move that gantry by 2x3Nm nema 23

Quote:

---

Originally Posted by silyavski

The general opinion is that if >70kg you can not move it adequately with 2x3nm motors so you need bigger motors, which now puts you into position to find bigger but at the same time fast enough or go with servo, which to me seemed the better option.

Now I am looking back at your gantry design. What size are the profiles, why 4, why are not connected each other? Whats your idea there? At that size machine
2x 100x100x3 or 100x100x4 welded together with separation for ball screw or not /if back/ will be ideal. 100x100x3 weights ~9kg/m and the 4mm weights 12kg/m, So you co8uld still move that gantry by 2x3Nm nema 23

---

The modification is being processed hopefully more optimal now but from steel I have no to big margin with Z+Y plate gantry legs and all stuff (some item was hided on my dwg ) so it will be 100 kg anyway
When I saw my friend machine with the moving bed I said I would not want to move to and fro 120-150kg mass with material so it would be moving gantry.
Entering to this 700x1000 mm machining zone with rigid steel design I have to revise myself.
I am seriously thinking of the moving bed option now it is just 5-700 mm longer space and can forget many trouble i.e the dual motor, heavier spindle
My goal is a bit better an more accurate machine than a router so called semi industrial so within a 4000 -4500 GBP budget I would like to reach the best performance.