Gantry & Coupled Axes
A portal carried on two rails needs both driven, and it needs them to stay together. One axis leads and the other follows its command, ratio applied; the control watches the two for drift and stops both if they part company.
The same coupling drives a geared axis, where the ratio is not one.
The settings
Section titled “The settings”On the coupling, under Synchronous Coupling:
| Field | What to enter | Unit |
|---|---|---|
Coupling ID, Name | which coupling this is, and what to call it | none |
Leading axis | the axis that is commanded | none |
Synchronized axis | the axis that follows it | none |
Coupling factor | the ratio, as a numerator and denominator: 1 / 1 for a gantry pair | none |
Sync error limit | how far apart the two may drift before the pair is stopped | mm or ° |
While the coupling is engaged the following axis ignores its own motion commands. It is driven from the leading axis and nothing else.
Square is captured, not configured
Section titled “Square is captured, not configured”There is no squareness number to type. The pair takes its baseline, the offset between the two rails, at one specific moment: when both axes are referenced and both are standing still. From then on the follower is held at that relationship.
Each rail homes to its own switch, independently, and the coupling stays out of the way while they do it. Those two switches are what defines square, and the baseline records the beam as they left it.
This is why a portal is squared by adjusting the home switches, not by jogging a rail. Free one rail, move it by hand or by jog to “square” the beam, engage the coupling, and the control records that as the baseline. A skewed beam is taught as correct, with nothing on screen to say so.
What stops the pair
Section titled “What stops the pair”| Condition | What happens |
|---|---|
The two rails drift past Sync error limit | both axes are stopped |
| The drift approaches the limit and stays there | a warning, without stopping |
| Either axis faults on its own | the other is stopped too |
The drift that is watched is the measured one: what the encoders say, not what the control commanded. A commanded follower tracks perfectly by definition and would never trip; a rack, a loose coupling or a rail catching is a real difference between two feedbacks. A single noisy sample does not trip it either; the difference has to persist.
The cross-fault case is the one people are surprised by, and it is deliberate: a healthy rail that keeps driving while the other has faulted is exactly how a portal gets racked.
After a trip
Section titled “After a trip”Clear the fault on both axes. The pair then re-takes its baseline at wherever the two are now standing, so re-engaging does not jerk the beam. The baseline follows the machine. If the stop left the beam skewed, that skew becomes the new relationship, and the way back to square is to reference both axes again.
Moving one rail on its own
Section titled “Moving one rail on its own”Disengaging releases the follower so it can be jogged alone, for maintenance or to free a beam that has jammed. Re-engaging takes a fresh baseline.
What this does not do
Section titled “What this does not do”- It does not square the machine. It holds whatever relationship the two rails had when it latched.
- It does not replace the mechanical setup. A pair that needs a large correction to look straight is a machine to be adjusted, not a limit to be widened.
- It is a licensed option.
Where to go next
Section titled “Where to go next”- Axes & Channels: the homing settings that decide where each rail references to.
- Backlash Compensation: the per-axis corrections, set after the pair holds together.