[blim]

I am building av CNC router size 1000x2000x200. On the longest axis I use two 20 mm ball screws having 10 mm pitch. Gantry weight is about 25 kg. My first thought was to run it with one motor located between the screws and driving them with a single timing belt to avoid having two motors on the same axis. I do not like the idea of having two motors too much since I could risk the gantry being teared to death if one of them stops, and I find no way to make sure the axis is squared after i stop the machine. So I have these options. Which would you go for?

1. Single motor, on a timing belt.
2. Dual motor with synchronization timing belt to prevent screws to get out of sync
3. Dual motor with individual end switches to keep the axis squared.

I tried to set up option 1 with a XL timing belt and pulleys. but I managed to tear the timing belt which I ran on 15T pulleys. Which timing belt / pulley size should be used for this kind of applications, XL, L, HTD, GT - there is a jungle out there. Also are any of you aware of a controller that lets you set individual endstops on a dual motor axis?

[ger21]

Most everyone uses option 3.
Motors slaved in the software, and separate switches to make sure it's square.

[blim]

Thanks, that leads to the next question. I can't find an affordable controller that has supports separate limit switches for the dual motor axis (gantry squaring). Do you have any suggestions?

[CitizenOfDreams]

Thanks, that leads to the next question. I can't find an affordable controller that has supports separate limit switches for the dual motor axis (gantry squaring). Do you have any suggestions?

The controller just needs to have enough inputs for the homing/limit switches, everything else is the responsibility of the control software. The minimal configuration requires 2 inputs: one for X, Y, Z and one for Y2 (the slaved axis).

[JerryBurks]

I use version number 2 with the timing belt. Works fine, as long as the belt pulleys screws are tightened with both motors "on" to make sure they are stepping in sync. Belt does not need to be super tight since it will not influence the accuracy.
I never use the homing procedure since the gantry is always guaranteed perpendicular. The limit switches are only on one side and strictly for just that, i.e. limiting the motion.

[blim]

The controller just needs to have enough inputs for the homing/limit switches, everything else is the responsibility of the control software. The minimal configuration requires 2 inputs: one for X, Y, Z and one for Y2 (the slaved axis).

That means I wire the limit switches for all axis in series, and home one axis at a time. And the last one for the extra stepper on the dual axis?

[ger21]

You could put X and Y on one pin, and Z and A on the other pin, then home Y and Z together, then the slaved X and A.

[CitizenOfDreams]

You could put X and Y on one pin, and Z and A on the other pin, then home Y and Z together, then the slaved X and A.

It is probably safer to always home Z first before moving any other axis. Just in case there's a tool in the spindle and some obstacle on the work table.

[CNC_DE]

Hello everybody,

I am working on a project right now, where i need to control 4 axis with the cnc programming.

the axis are built in this way: one seperate vertical axis, which moves up and down. the second axis is a horizontal one, which moves left and right. and the two last ones have one horizontal axis with a vertical one.

All those axis need to be synchronized.

So i thought , i write for every axis a cnc program: CNC_Axis_1 , CNC_Axis_2 and CNC_Axis_3_4. so i think i will have at the end 3 cnc programs that need to be synchronized.

How can i do this synchronisation between the 3 cnc programs. Is there any tricks or options or globale variables , which can be used, to execute the 3 programs in a define time or define position.

Please Help )

[fretman_2]

I drive both my Y motors with two different drivers, but the drivers come off of the same port on the controller. I'm limited with my setup, Arduino Uno, so that's the only way I could do it for now. Both motors stay in sync as long as they start out at the same place. I have a hard stop on both sides and I run the gantry up close to that, then move both sides against the stop by hand, then I energize the drivers. Seems to me you'd risk the machine tearing itself apart if you lost a motor no matter how it was driven.

[raychar1234]

Agree, version 2.

I used it in my stepper driving motors router machine a long time. However, I am going to build another one with easy servo motors. It will have two gantry(s). So
totally needing 2+2+1=5 motor (servo). It costs me too much. I am going to move to version 1. Price for one Nema.34 servo is much less than two Nema23 servo motors!
But, in that case, the problem as mentioned happens. The flexing of the long timing belt will contribute backlash (inaccuracy). Can it be solved? Will tightening it eliminates this problem?

In X-axis, one 18mm width 5M timing belt runs through two lead screw pulleys and motor pulley, with motor one at center.
In Z-axis, two 9mm width same type timing belt to connect individually two lead screw pulleys, stacked motor pulley at center.

Thanks,

[wizard]

Old thread!

Agree, version 2.

I used it in my stepper driving motors router machine a long time. However, I am going to build another one with easy servo motors. It will have two gantry(s). So
totally needing 2+2+1=5 motor (servo). It costs me too much. I am going to move to version 1. Price for one Nema.34 servo is much less than two Nema23 servo motors!

Technique #1 is doable but like all thing there are qualifications.

First off many timing belt implementations are undersized in my opinion. In order for the belt not be a factor in the dual driven lead screws it must be tensioned tight, probably tighter than most would care to attempt. This means the bearings(all of them) must be able to handle that tensioning load. Ideally this means that you relieve the steeper motor of the radial load associated with the tight belt.

Second you will be limited by the distance achievable between lead screws. This is due a significant amount of belt being unsupported and subject to oscillation (the guitar string effect). This can be moderated some what via careful tensioning, and providing support for the belt to ride against. Ultimately though there is a real limit where is makes far more sense to look into dual motor control.

Third one leanscrew pulley will need a phase adjustment capability. This is to allow for proper loading of the leadscrews. It may also help with squaring if the gantry doesn't have the stiffness to maintain square on its own.

But, in that case, the problem as mentioned happens. The flexing of the long timing belt will contribute backlash (inaccuracy). Can it be solved? Will tightening it eliminates this problem?

Size is everything here so keep that in mind. However unless the belts and pulleys are worn or otherwise mismatched you will not get significant backlash out of a belt drive. I state this on the basis of what backlash is which is clearances in machine components. This is important because the issues you do get from belts can be dealt with in other ways when you stop thinking in terms of backlash. As mentioned above the tension in the belt is critical as this preloads the belt but it also is a factor in the belts singing. So one very good suggestion is to have a sound method available to tension the belt. Ideally this happens on the length opposite the drive motor. The idea here is to keep the unsupported lengths of belt short and of equal lengths. The equal lengths means tensioning should have the same impact on all belt segments.

In X-axis, one 18mm width 5M timing belt runs through two lead screw pulleys and motor pulley, with motor one at center.
In Z-axis, two 9mm width same type timing belt to connect individually two lead screw pulleys, stacked motor pulley at center.

Thanks,

You make no mention of size so this is completely off the cuff but I woiuldn't use HTD belts smaller that 1/2" wide which is around 13mm for any axis where you attempt this. Wider belts may be required. Note I must repeat this is only viable if the distance between leadscrews is keep modest.

I'm not sure how you arrived at your costs but the reality is a type 3 solution is generally cheaper which is why it is rather popular. It also saves on precision machining because to do a type one you need the components all mounted on one plane with every axis orthogonal to that plane. Otherwise belt tracking and wear become a real issue when the various shafts are not 100% parallel. In other words it takes a bit more effort and I dare say precision, to do type 1 well. So I'd look closely at how you come to the conclusion that type one would by default be cheaper, that is likely only so if you have the equipment to make a proper mechanical implementation.

[raychar1234]

Firstly, thanks for your detail advice.

By the way, if using the version 3, it seems that Mach3 only supports one gantry slave limit switch and not the two gantry(s), doesn't it?
When having very tight belt tension, one bearings on the pulley may wear easily. Will consider adding another one on its opposite side.
But without excessive protrusion of shaft length, it is another (headache) problem..Attached is the picture of the cad model, it is
supposed mainly for aluminium machining and occasionally for light steel machining. With all new linear actuators purchased, I am now
in the stage of settling the motors arrangements. Weight for the steel table and Z(Y?) bridge are around 30kgs each.
The respectively leadscrews separations for X and Z are 500 and 690mm.

Attachment 396428

[skrubol]

Are you using linear actuators rather than separate screws and linear bearings? If not, why have 2 ball screws on the table? One in the middle should work.
Interesting concept with the vertically moving gantry. Not something you see very often.

[ger21]

By the way, if using the version 3, it seems that Mach3 only supports one gantry slave limit switch and not the two gantry(s), doesn't it?

No, Mach3 supports separate switches for a slaved gantry.

[raychar1234]

Are you using linear actuators rather than separate screws and linear bearings? If not, why have 2 ball screws on the table? One in the middle should work.
Interesting concept with the vertically moving gantry. Not something you see very often.

Using linear actuators are easily for me to build. One in middle is doable but two are more strong especially when machine cutting head moves to leftmost or rightmost positions during cutting.
Vertical gantry will make a longer Z travel which still keeps strength when it is a lower position in during low level cutting.

[raychar1234]

No, Mach3 supports separate switches for a slaved gantry.

Understand. But what will happen if one of switch fails, one of the leg will continue move while another one is stationary?
If the gantry structure(let's say a thin extrusions) is not strong enough while the power of drive motor is high. Will
it deforms it to a permanent damage?

Regards

[CitizenOfDreams]

But what will happen if one of switch fails, one of the leg will continue move while another one is stationary?
If the gantry structure(let's say a thin extrusions) is not strong enough while the power of drive motor is high. Will
it deforms it to a permanent damage?

A properly designed machine should not be able to self-destruct like that. The motor should stall or fault without inflicting any permanent damage.

[raychar1234]

For the homing purpose of two leadscrews in gantry machine. Which type of switch is suitable in aspect of precision: mechanical level arm or Hall sensor type as shown in the attachment.
The costs for them are around 4 and 5USD per piece respectively.

Attachment 396850