Z axis runaway

[camaro_dan67]

Ok this is my other problem I posted at the end of my other thread. Machine is a Kiwa Excel center-4 from 1984. Control is a Fanuc 3m model c. I was having trouble with a watchdog alarm that started all this. Tried a replacement motherboard which ended up throwing a z axis feedback cable is disconnected error. I then took off the z axis servo motor cleaned it with QD electronic cleaner and reassembled. After the cleaning and reassembIing the motor the z wanted to slam into the table. I then checked the signals from the encoder with a scope. Found the encoder had 2 separate square waves when rotated by hand also it had one single pulse per rev on another wire. I checked the same signals at the board. They were the same. I determined by this the signal wasn't getting into the board hence the error it had on the screen. I then bought another motherboard and that's where I'm at now. I hooked up the new board and entered all the parameters and got the control to come up with no errors. When I release the estop and hit the reset button the z axis tries to rapid into the table. When it does this the 2x4 I have holding the head up stops it and the drive throws an ocv error. If the encoder is putting out signals when rotated that means it is working right? What would cause the z axis to take off in the negative direction as soon as the servos are powered up? I also tried another velocity control unit for the z axis with no change. What do I check next? Is my encoder bad even though it puts out signals when rotated? Dan.

 

[cngbrick]

I'm not entirely familiar with the Fanuc servos but I assume this is a brushed DC motor that you took apart? Is it possible that during reassembly the stator magnets were assembled at 90 degrees from their original orientation? I know from experience this is possible on Baldor DC servos and results in the behavior you describe.

RT

 

[camaro_dan67]

Yes it is a brushed motor. It has 4 brushes. I did scribe a line at each end before taking it apart. I suppose I could have put one of the ends back on differently then taken apart. Is there a way to know that the end caps are in the correct orientation? The magnets are also fixed to the center sleeve and I did not remove them from the sleeve.

 

[atex57]

The first thing I think of is reversed leads to the motor.

Ed.

 

[gustafson]

And since it was showing an encoder error, incorrect encoder hookup.

 

[cngbrick]

The first thing I think of is reversed leads to the motor.

Ed.

This would be a quick way to check if the motor was assembled with the magnets/brushes at 90 degrees -- just reverse the motor power leads.

RT

 

[camaro_dan67]

How would I reverse the leads? The motor connection cable and the encoder cable are both pre made cables with crimped end connectors.

 

[atex57]

Check if there are leads from the brushes to the connector or leads at the drive.

Ed.

 

[michiganbuck]

Z ax is run way..guess I gotta use a saw.

 

[cngbrick]

How would I reverse the leads? The motor connection cable and the encoder cable are both pre made cables with crimped end connectors.

One option is to compare direction of rotation between two motors. Apply something like 12 VDC to the motor and compare direction of rotation between the suspect motor and a known good motor (other axis?). If it's the same then it points to the encoder or cable (assuming you have already ruled out the drive).

RT

 

[gustafson]

(assuming you have already ruled out the drive).

RT

blown shorted output transistors would do this also

 

[michiganbuck]

You might check polarity to see that the safety stops or a motor are not somehow reversed.

 

[TheBigLebowski]

If the leads are swapped on an encoder it will cause a runaway. The PID loop will see some following error in one direction then command the motor to move in the opposite direction to correct. If the motor moves in the same direction as the error (due to swapped leads) the error grows, the PID loop takes the same action again to correct, cascading into a runaway until something stops it.

Swapped leads on a dc motor would probably have the same effect, as the motor/encoder disagree on which direction to go.

 

[camaro_dan67]

The thing that confuses me with all of this is all I did was dissamble the motor to clean oil and brush dust out of it and put it back together. With the encoder all I did was take the cover off the top to check for signals on the circuit board then check for the same signals at the motherboard and put everything back together then it went crazy.

 

[Vancbiker]

blown shorted output transistors would do this also

Ummm. A Fanuc DC drive doesn’t use transistors on the output. SCRs are the output devices. I suppose it’s possible, but it would require two units to be shorted simultaneously. Pretty unlikely IME.

 

[TheBigLebowski]

can you swap motors (plug the x motor into the z and vice versa) for a quick test to determine if the problem is related to the z axis motor assembly vs the ebay motherboard? If swapping the motherboard didn't actually fix your original problem you might be going down the wrong rabbit hole.

If the X axis runs away when plugged into the Z slot on the board - you know the board is bad.
If the Z axis runs away when plugged into the X slot on the board - you know the Z axis motor/encoder is wired incorrectly.

 

[camaro_dan67]

I'm not sure about changing motors or drives between xy and z the z is different in the motor and the drive.

 

[cngbrick]

I'm not sure about changing motors or drives between xy and z the z is different in the motor and the drive.

Start by checking rotation direction with a 12 Volt battery. If it's different than the other motors then take it apart again and see if the brush assembly or stator magnets can be rotated 90 degrees. Make sure you don't take the rotor out of the stator for too long (or at all if you can avoid it) or you risk weakening the magnets.

RT

 

[camaro_dan67]

Ok Sounds like I need to remove the motor again to do some testing. I would also like to test the encoder again before removing the motor. Its a Fanuc model A860-0300-T001 2000p. Can someone give me a testing procedure and values I should be seeing for this encoder.

 

[cngbrick]

Do you have a 17 pin or 19 pin connector for the encoder?

RT