Wackiest Trouble?

I'm sure that most of us have experienced trouble similar to Timothy's post on a servo drive.



Just curious, what are some of the most unexplainable you have seen?



After replacing a DC motor, the machine ran without a hitch for two weeks exactly, no trouble.



Then two weeks almost to the minute, the drive went into runaway for no good reason. Two seconds later $80K of parts were toast. Six good technicians called for help 24 hours later. A control guy from the west coast showed up. Two days later we were all scratching our heads, no progress.



Somebody says out of the blue, "Hey, lets reverse the field windings and see what happens." It ran, but with the labels clearly indicating that the field windings were reversed. The drive ran for ten years and never had a similar problem again.



I was watching it when it went into runaway. No one was near the controls. It went into runaway in mid cycle.



This was TTL logic, and Hg wetted relays. Parts were almost non-existent even then. (Before the days of runaway safety controls)

I was called to look at a drilling rig where the mud pumps and rotary table would get unstable and settle down. The engines would also vary 1/2 to 2 hertz at the same time. Several others had looked at the issue with no luck. When I got there things were running smoothly and they were not allowed to drill unit it got fixed. They were circulating the mud to keep the hole open. It did not matter what engines were running or what drive the motors were assigned to or if they were running 1 pump or 2 pumps. Most drill rigs have the ability to assign the motors to 2 different drives. The system is 600 VAC 3 phase and 750 VDC at 1800 amps DC drives.

I looked for shields that were grounded, noise on the engine signals and found none and for things that usually caused instability problems. The motor man mentioned he noticed when one of the compressors grounded out and tripped it got really bad then. I decided to look at motor control. The motor J box was opened and the insulation was replaced and the compressor was tested good. I looked at the ground fault circuit and pressed the test button as B phase was brighter. The motor control system is transformered to 480 vac. All 3 lights got the same bright. It was then I realized there were leakage grounds on A and C phase in the 480 volt system. They were traced to an air conditioner on opposite ends of the control house. Both AC units were set to max output and all pumps and engines began to get unstable. When 1 unit was turned off everything smoothed out. Needless to say both AC units were shut down. There were 4 total and it was early May so they could get by until the units could get replaced.

I suspect that having A phase on one end of the metal house and C phase on the other end of the house with about 15K ohms to ground each caused a bit of ground currents thru every shield there. Oh Yeah, the compressor was a hard ground on B phase.

My wackiest problem:

I once worked on a feeder system that had sections of vibratory track. A small conveyor with a 2HP motor was used to transfer the parts out of the system. Here's the problem I found:

After the system ran for a while, the drive would fault out (It was a small A-B drive. Can't remember the model or fault code.)

Turns out that the motor leads were taped up in a junction box located just under the vibrating track section. The electrician did a lousy job of taping the leads and an exposed section of motor leads would ground out ever so slightly (but not continuously) while the track vibrated and fault out the drive.

The wiring problem was not clearly visible upon inspection. I really had to look at the tape job closely. This problem took a full day to find.

I got called to a customer's plant where their machine had been running for about 4 years. The machine had stopped running and it appeared that the main 24VDC power supply had failed. When I measured voltage at the power supply itself, though, I found that it was working properly with the 24VDC output wire soldered firmly to it. After tracing the wire back to the terminal block, the 24VDC was gone again.

Assuming I had a broken wire, I eventually cut the wire to within 6 inches of the power supply and stripped off all the remaining insulation leaving a bundle of wire strands sticking out from the power supply. Still, I was able to measure voltage at the terminal but not on the wire itself. Somehow the wire or solder was "open" even though all the wire strands were securely soldered to the power supply terminal and nothing else. Apparently they had a cold-solder joint with a 4 year time delay.

After I de-soldered the original wire and soldered on a new one, the problem was gone.

Steve

Randy,

If you are still curious about your wacky SPST problem, here's a possible explanation...

When you "took it out of the circuit", you may have been reading it when it was oriented differently. Gravity may have temporarily "made" the contact with a loose or intermittent switch "wiper" touching the contact point, but only when you held it a certain way.

Remember, a momentary switch needs some sort of spring mechanism to return to its de-energized position. This spring mechanism is subject to fatigue over time and is the most likely part to fail.

Jimmy, I agree, but I should have included it was a toggle SPST. Sorry for the confusion, if it were momentary the orientation may well have been a problem.

Hmmm.. When I started here, we had one of the earlier Compumoter???? (I think) servodrives running a Japanese brand motor, indexing part of the machine. Every so often, for no reason anyone could find, they would kick on(at full speed). Since the part of the machine it was running was round, it was dubbed "doing the helicoptor".

This was before I knew anything about servodrives or programming, so I wasn't involved in the troubleshooting at all. (Ug.. Me change bearing good)

They never did figure out exactly why it would do that sometimes... The fix??? (I guess you could call it that) was to change them out for Kollmorgen Drives and motors.
Knowing what I do now, I'd say it had something to do with the fact that we don't have any earth grounds in the plant other than the ones on the main transformer coming in. (I've been trying to get more installed for a couple of years now... No luck)

One more with vibratory feeders. The first shipment of machines we recieved from Austria all had vibratory feeders. All tuned to 50Hz. That sure was fun.