The effect of downtime ranges from a minor inconvenience to a total loss of productivity. Some industry average costs are as high as $80,000 per hour, burning four hours per event and nine downtime events per year for a loss of nearly $3 million per plant per year. Another survey reported annual losses of $350,000 to $11 million per organization, with an average annual loss of $5 million.
Eliminating downtime would be ideal, but it’s not practical. Reducing downtime duration and severity is one of the goals of technical training. While doing motion control training of maintenance electricians, engineers and technicians, we’ve observed a number of common misconceptions and confusion about motion-control applications. Sometimes these misunderstandings can prolong your troubleshooting efforts.
Continuity counts
One such misunderstanding involves three-phase motor connections. In a conventional three-phase motor, reversing any two of the motor leads reverses the motor rotation. Interestingly, in motion applications, misconnecting the U, V and W connections won’t reverse the rotation. Instead, it causes the motor to run slow, hot and perhaps erratically.
What was overlooked in this case is that motion drives use feedback to monitor motion. This feedback compensates by electronically reversing the order in which the drive switches phases to keep the motor turning in the same direction. Unfortunately, the drive can’t do this well and the motor won’t run efficiently.
Another overlooked fact is that modern variable-frequency drives can operate with one missing phase. In fact, they can function with only one phase if de-rated. Older drives had no missing phase diagnostic, and would manifest a missing phase as overcurrent faults, undervoltage faults and overtemperature faults. These tend to point to a drive problem, not a supply problem.
We ran into this problem on a robotic application in which movement downward was fine, but upward movement immediately caused these faults to trip the drive. Replacing the power supply modules didn’t help. After some thinking, we realized that this robot was used for training and was subject to uncorrected “bugs.”
Sure enough, carefully tracing the input power circuits revealed a disconnected phase. Of course, in real life, you don’t have “bugged” equipment. But phases can get disconnected for various reasons, including tripped breakers or broken connections. Missing phases can be hard to spot in simple three-phase motor circuits because the symptoms of running hot or not developing proper horsepower aren’t immediately attributed to a missing phase. Late-model drives probably include a missing phase diagnostic, which makes this condition even easier to overlook in older drives because people start assuming that every drive has this diagnostic. Fortunately, detecting a missing phase is easy – simply check the phase-to-phase voltage. The values should be the same. Follow appropriate safety precautions for the voltages you’re measuring.
Getting tuned
An area of limited information – and misinformation – centers on tuning a drive. Modern motion drives shouldn’t need tuning very often and, with autotune features, this should be a simpler procedure than in the past. Most autotune procedures involve selecting the variables to be adjusted and clicking on a software button to initiate the process. You then tweak the values if you feel it’s necessary.
Manufacturers indicate that retuning shouldn’t be necessary when replacing a motor with an identical motor having the same part number. Others have doubts about whether motors really are that identical.
Another argument for retuning involves a weak motor or increased bearing resistance as the system ages. Obviously, the correct solution is to replace the motor or bearings. However, at $80,000 per hour, waiting for a replacement motor could be quite costly. Somewhat like PLCs, which can be reprogrammed to compensate for mechanical problems and thus allow the postponement of the proper repair until a more convenient time, motion drives can, to some degree, be tweaked to compensate for mechanical problems.
Tweaking the PLC to compensate for mechanical degradation requires “un-tweaking” it after the proper repairs are completed. Similarly, electronic adjustments to motion systems require retuning after the proper repairs have been made.
Should you tune with – or without – the load? With some thought, you realize that you’re trying to tune the total movement, which includes the entire moving mechanism, not just the motor. But, with a little more thought, you realize that tuning procedures are going to move the motor and the connected mechanism in a somewhat unknown and vigorous manner. Maybe tuning with the load connected isn’t a good idea.
Manufacturers suggest using fairly rapid movements for successful autotuning. But high-inertia loads, such as flywheels, fans and pump impellers, can’t change speed and position suddenly without risking shaft damage and bent fan blades. So, don’t impose sudden changes on these devices.
One approach is to perform the initial tuning with the load disconnected to verify motor hookup is correct and to capture initial tuning values. After confirming that motor movement is in the expected direction under the initial tuning, connect the load, enter appropriate speed and acceleration limits, and tune with the load attached.
Because system hardware varies so widely, we can’t give any specific values here. Have someone knowledgeable about your system and application derive a procedure and document it for others to follow. This procedure should indicate when to tune, the qualifications of the person performing the tuning, the specific tuning sequence to follow, and the maximum speed and acceleration values to use when the load is connected.
Safe recovery
Another area of possible confusion is how to recover from a system failure after it reaches a hard stop. This is problematic because your system should never have reached a hard stop – a soft stop should have caught the problem. This is considered a serious fault because the drive shut down and disabled itself. You probably don’t want to disable hard stops to allow any movement farther in the wrong direction. But, you can’t clear the fault before you get it off the hard stop. This means you’ll have to move it back into the normal operation range manually.
Although not recommended, you might be able to move a small system manually without disconnecting the load. However, on larger systems, you’ll probably need to disconnect the load to reposition it. This needs to be done with respect for stored (potential) energy and the resulting movement likely to occur when the load releases. Give careful thought to how the load will be moved manually and what sources of energy need to be locked out or controlled.
Energy can be stored in the form of unsupported weight, air pressure, hydraulic pressure or water pressure. With so many possible configurations, it’s impossible to give specific guidance, but developing a well-thought-out procedure before the emergency strikes can prevent a lot of problems.
Such a procedure includes, but isn’t limited to, lockout/tagout procedures, support of the load, and a method to move the load back within its zone. Will you use human power and shove it back into position or will you need a mechanical assist? If mechanical, what form? Can you do it with a come-along or do you need something more powerful?
The recovery procedure also should include how the load is to be aligned and positioned while it’s reconnected, and steps on homing and a system checkout before returning the equipment to production. You also might want to specify the qualifications of the person performing the recovery process. It might require specific skills, or perhaps the procedures and training are sufficient to allow any maintenance technician to do the work.
These issues aren’t difficult, and with proper training, the issues of motor connection, missing phase and when to tune are easy to identify. Good training, procedures and communication greatly reduce the time needed to recover.
Leno Pederson is chief instructor/controls engineer and Howard Loveless is an instructor/controls engineer at Intellect Controls Group Inc., Louisville, Ky. Contact them at (888) 468-3591.