Why is Y axis "squaring" necessary only with twin motors?

Hi All

In another post regarding a new build, one of the contributors commented:

Quote:

---

If you are using two screws and two motors you will have to use two home switches to make sure the gantry is put square every time you home.

You can of course use two screws with one motor and a belt driving both screws then one home switch will be fine.

---

This might be another example of "the blindingly obvious", but to me it's not......yet!


This was a moving Y gantry machine.


If the machine was well designed for maximum rigidity, and accurately built with good quality components (as best this can be achieved in "home" circumstances), why does it go "out of square" such that "squaring" is necessary?

I can only think that it might be caused by one motor missing a step.......but then why doesn't the gantry resume "squareness" naturally when the system is powered down?


But this suggests that in a single motor, twin belt drive moving gantry machine, the gantry always remains square?..........does it?


Then I wonder what is regarded as "square" or unacceptably "out of square"?

5um/m, 50um/m?


Can someone explain and put this into perspective for me please.


Many thanks

Martin

If the machine is rigid enough that it can't go out of square you could drive just one side, if it's not rigid enough to drive just one side then it can go out of square and zeroing on one side may not set the machine square.
The exact figures will depend on the individual build and requirements but why wouldn't you want to make a machine be the best it can be? ;-)

- Nick

-Nick

I WOULD want a machine "the best that it can be". My 40+ years of Engineering has instilled in me the importance, no, necessity of (appropriate) machine precision, accuracy and rigidity.

I am currently trying to get my head around ALL of the elements of what many you guys have been doing for that long that it is probably second nature.

So, expanding on what you have said:

If a machine is not rigid enough such that it needs twin homing squaring, do the belts in the single motor approach keep an adequate squareness?


Can anybody put um numbers on this?


- Martin

Look up belt manufacturers specifications on length change versus temperature and tension and combine that with your pulley diameters and screw pitches to get a number.
Using a tensioning system which applies equally to two both upper and lower belt runs will almost eliminate errors from temperature and tension, good grade ground ball screws with pre-tensioned ball nuts will help too.

- Nick

Thanks Nick

..............yes I'll do just that.

You can get belts with various reinforcements, fabric, more exotic fibres and steel wires which will all perform differently with changes in temperature and tension.

I have a machine with an approx 1000mm long gantry, driven at each end. Couple of ways in which it can get out of square:

One motor stalls for some reason, or sometimes when I hit the e-stop button. If the two motors do not stop at the same time, or come to a halt at slightly different rates, you have lost sync.

When you first power on the stepper drivers, the motors may or may not be at a full-step point. If not, it's likely that the driver will initialize it to the nearest full step, although I believe that some drivers are supposed to remember the last state when powered down. It's possible that if this happens often enough, the gantry may drift slowly out of square, a tiny amount each power-down/power-up cycle.

Homing both ends to a known position gets round these problems; on my machine it will hold these settings for an entire session unless one of the above happens.

So Neale

Quote:

---

One motor stalls for some reason, or sometimes when I hit the e-stop button. If the two motors do not stop at the same time, or come to a halt at slightly different rates, you have lost sync.

When you first power on the stepper drivers, the motors may or may not be at a full-step point. If not, it's likely that the driver will initialize it to the nearest full step, although I believe that some drivers are supposed to remember the last state when powered down. It's possible that if this happens often enough, the gantry may drift slowly out of square, a tiny amount each power-down/power-up cycle.

---

whether it is motors jumping sync/steps, and/or the movement of a less than stiff gantry/endplate/bearing setup, the TWIN homing switches correct the gantry squareness.

Makes sense.

And presumably, within a single motor/twin belt system, the "stiffness" within the drive belt arrangement tends to hold the gantry square.


Great got it thanks.

-Martin

Quote:

---

Originally Posted by MartinS

So Neale



whether it is motors jumping sync/steps, and/or the movement of a less than stiff gantry/endplate/bearing setup, the TWIN homing switches correct the gantry squareness.

Makes sense.

And presumably, within a single motor/twin belt system, the "stiffness" within the drive belt arrangement tends to hold the gantry square.


Great got it thanks.

-Martin

---

Yes that is correct, but with using one motor you have to use a bigger one like a nema34 which usually requires a bigger volatage to drive it. The single motor setup does not go out of square, and using 2 motors hence the duel switches you also.need stall detection etc in case one drive packs up etc

Quote:

---

Yes that is correct, but with using one motor you have to use a bigger one like a nema34 which usually requires a bigger volatage to drive it. The single motor setup does not go out of square, and using 2 motors hence the duel switches you also.need stall detection etc in case one drive packs up etc

---

Noted Clive, thanks

Yes, stall detection is important with a two-motor setup. Modern digital stepper drivers will do this for you.

As Clive says, a typical single-motor setup will need a Nema 34, where my machine (gantry with spindle, etc, is probably around 40-50kg) is driven by two 3Nm Nema 23 motors with 2005 ballscrews. Not an optimum configuration, but it works ok up to about 5000mm/min.