Gantry dual servo motor with absolute encoders

[nicubila]

I am facing an issue in regards to a router with 3 axys - X, Y, Z, while Y gantry has a further slave - A.
All motors have absolute encoders.
The system is Delta - the controller (NC300), the drives (Asda A2) and the motors.

My issue is that even though the gantry is quite rigid (steel, about 100kg), when the system is stopped, the gantry can be moved by hand, and the Y and A get out of sync.
If I move by hand the X axis, the absolute encoder will know the position even after restart, as it is normal. But even if this is the case for Y sometime, in some cases the Y motor turns more than the A motor when moving the gantry by hand, and I will get an error when starting the system.
In my view the A (slave) axis should reposition itself in sync with the Y axis automatically of the system detects this.

I opted for absolute encoders to have this benefit of no needing to go through a homing procedure.

One option I consider it to have on the Y and A axis motors with brakes, so when stopped the gantry cannot be moved. Of course that will not come cheap, but not having to do the homing is a great benefit.

I would appreciate inputs from anyone having a similar setup, and how it handles the situation of accidental moving of the gantry and a desyncronization of the master-slave motors (and encoders).
Using home switches defeats the purpose of absolute encoders.
Also, important to know that X and Y axis must be squared, and this procedure is very time consuming

Marius

Similar Threads:

• Dual motor gantry homing?
• Dual motor gantry and Mach3 features that don't...
• cheap absolute encoders? Oxymoron? Decent Incremental encoders?
• Absolute encoders?
• Absolute encoders?

[wizard]

How is it that absolute encoders eliminate the need to home an axis? I guess the answer depends upon the type of encoder. There are various approaches to absolute encoders that may or may not retain “absolute” position value. One type simply offers up a value representing rotational position. Often this is 360 degrees with some odd amount of resolution. Beyond shaft rotation it is the controllers responsibility to maintain actual position. Beyond that there are absolute encoders that can track position to XX revolutions and report those values via a serial connections.

For linear encoders there also a variety of ways to do absolute or give you absolute like results. The trouble here is telling us that you are using absolute encoders really doesn’t tell us much of anything.

[nicubila]

How is it that absolute encoders eliminate the need to home an axis? I guess the answer depends upon the type of encoder. There are various approaches to absolute encoders that may or may not retain “absolute” position value. One type simply offers up a value representing rotational position. Often this is 360 degrees with some odd amount of resolution. Beyond shaft rotation it is the controllers responsibility to maintain actual position. Beyond that there are absolute encoders that can track position to XX revolutions and report those values via a serial connections.

For linear encoders there also a variety of ways to do absolute or give you absolute like results. The trouble here is telling us that you are using absolute encoders really doesn’t tell us much of anything.

Hi Wizard, and thank you for the reply.
Indeed, I was quite vague about the description. There are indeed many types of absolute encoders.
Even though the encoders on my motors are said Not to be multiturn, they record values even when I turn off the machine and I move one axis much more than one turn. For example if X axis is 0.000 mechanical, and turn off the system, then move by hand the X anis to the other end (1600mm), when I turn on again the system (controller), the new position is 1600 in mechanical positioning.
I guess there are about 300 turn of the motor (1:10 reduction). So that is fine.

On the Y axis, with A slave, if I move the gantry by hand, I there is no perfect rigidity in the gantry and one of the motors (Y or A) moves slightly more. When turning on again the system, I get the error of non alignment of gantry.

My question is how others deal with the issue. As mentioned, my guess is that there has to be a way that the motors get again aligned, so both A and Y get to same count.

Marius

[ger21]

If you are going to use the encoders like you want, then you need to make sure the machine is perfectly square when the motors are at their "home" position. Or your control needs to be able to deal with this issue.

We have a $150,000 router with absolute encoders, but we till need to home it every time we turn it on. Once homed, we can hit E Stop and push it around, and it will always know where it is. But once we shut it off, it needs to be homed again.