Overcurrent trips in the EVALL6472H-DISC due to defects in the stepper motor

[scope]

Posted on January 27, 2015 at 21:42

I'm testing the EVAL6472H-DISC boards. In the first tests, one worked fine with an overcurrent setting of 1.5A on a motor rated at 1.4A. There was and is no mechanical load on the motor, its just lying on a work-bench.

Then for no obvious reason, overcurrent faults began to appear. The motor would go one step and stop with an overcurrent fault indicated on both the board and GUI. That the motor only went one step with the overcurrent trip set a little more than maximum phase current indicated a locked rotor condition. The faults were traced to both a defective motor AND GUI settings problem. I had disassembled one of the motors to inspect the internals, but did not completely tighten the screws that hold the end caps on the motor and this does VERY bad things to the motor. In the motors Im using, the end caps have a machined shoulder that interferes with the rotor pole pieces to center the bearings. If the end caps are not tight, the bearings and rotor will tilt and cause mechanical binding which raises winding current- the motor appears to have a heavier load attached to the shaft which results in excess winding current or a locked rotor. What was odd is that after tightening the screws, I turned the rotor a few steps, thinking this would reseat the bearings. It did not. After tightening the screws, the motor did not run properly, it made a growling noise at 200 steps/second. After a few stops and starts, the growling sound was eliminated and was replaced with a higher pitched squeal. That probably means that the bearings reseated themselves in the end caps. There must have been some minor bearing damage or misalignment that remained after the screws were tightened. Also, beware that stepper motors, in general, use the wrong type of bearings. They use radial ball bearings and that is a bad choice for stepper motors which exhibit the highest torque at near zero speed along with sudden, frequent starts and stops. The balls in radial ball bearing units tend to skid when the races start to turn and that can cause internal damage and binding which may appear as motor overcurrent. Its like, for an instant, that the bearing is seized and rotor is locked. Ball bearings also have minimum mechanical load ratings and a stepper motor lying on a table has essentialy no load on the bearings. With no radial load, the balls are likely to skid. See SKFs Engineering data for more information. Stepper motors should have needle or plain bearings, not ball bearings especially if used at slow step speeds. Be very careful to not strike the motor shaft or allow it to strike anything while being installed or shipped. This could cause bearing damage. Overcurrent trip continued to ocurr until I raised the overcurrent level in the GUI, now the motors seem to work as before. (GUI-Tools- Device Configuration-Others tab- Overcurrent threshold was set to 1.875A instead of 1.5A). The overcurrent threshold was raised until the overcurrent trips stopped. Just because a motor is rated at 1.5A does not necessarily mean the winding current will be that value, especially if there is mechanical binding or damage inside the motor. Once the motor bearings are seated and the motor is working normally, the current might be reduced. There may also be a break-in period in a motor after which phase current is reduced. Also, avoid running this board at less than 5 steps, it ran but overheated grossly. I tried one step per second at default settings, it ran a few seconds then overheated the driver. Lowering the Hold current may solve this problem. At such slow speeds, the driver is sourcing maximum current as the motor windings are no longer charging- the winding current becomes DC.