can a encoder be mounted on a ball screw?

[replicapro]

I have it on the servo, and my z axis is simply not ballance correctly. I saw someone mentioned putting the encoder on the ball screw before and was wonder how to do this and if it is better than on the servo. Im off by -.003 over just 20 cycles moving from z+2 to z-2 and it is causeing the cuts to go deeper on each down stroke.

[One of Many]

A few other things to look into.

Something could be slipping or that encoder is missing pulses for some reason. Either one of those problem will create an accumulative gain. It should not matter where the encoder is mounted as long as it is coupled without slip. Connecting it to the screw will show actual motion, but both positions should be repeatable. If there were an encoder scaler parameter, the pulse per unit could be off, but still repeatable.

DC

[replicapro]

im pretty sure the problem is my counter ballance for my head.
The more weight I add, the less it messes up. After a period of time though, the weight may need to be changed as the encoder and servo wear out. So I think the best thing to do is put it on the ball screw. The amount of weight the servo is throwing up and down is well over 200 lbs, so I fingure the inertia of the head and counter weight is causeing the problem. What I really need is zero gravity. LOL

[NC Cams]

When properly counterweighted, the axis you're trying to move can weigh almost NOTHING to the drive. That's the cool part.

However, controling the inertia which will be higher is another issue altogether. Try programming in a ramped slow down as opposed to an instantaneous start/stop. Can make a big difference with some drives.

Inertial and mass are the bane to our existance sometimes. Yet without them, flywheels would be useless.

[One of Many]

I cannot offer anything as written in stone. Long distance troubleshooting is more like speculation based on symptoms.

It is has been my experience that if the encoder is tracking actual motion, the weight should not be throwing that position count off. If this is a stepper driven Z, and the encoder is supposed to correct for lost pulses, it should be easy to trace. Possibly somewhat self correcting IF the software has true closed loop to restore true position at the end of a move. The only thing I might add here is that if during the accel and/or decel slips steps at the start or end of a move, the system thinks it did a correction, but never completed the physical move. Then the chance for a positional reset is gone and so induces your accumulation error? In that case, your "anti-gravitational device" would seem appropriate.

If this is servo driven, then coupling between the screw and the motor, motor and encoder or loss of encoder pulses would be the first things I would investigate. Servo systems are true closed loop, so if the weight were an issue it would drop out the servo amp on overload, not lose positon.

DC

[Karl_T]

Originally Posted by replicapro I have it on the servo, and my z axis is simply not ballance correctly. I saw someone mentioned putting the encoder on the ball screw before and was wonder how to do this and if it is better than on the servo. Im off by -.003 over just 20 cycles moving from z+2 to z-2 and it is causeing the cuts to go deeper on each down stroke.

FWIW, my hardinge CHNC lathe came from the factory with resolvers mounted on the ball screws. I put encoders in the same spot. Works great. I don't know what will happen if I ever break a belt.

Karl