The main reason balancing efforts sometimes fail is because the problem is simply not due to unbalance. Balancing fixes only one problem: unbalance. Vibration from misalignment, eccentricity and other problems may be reduced in just one direction, but not by balancing.
Simply stated -- To succeed at balancing, fix problems due to misalignment, eccentricity, bad bearings, resonance, and looseness first.
• Do everything with precision. • Check initial reading for repeatability. • Wait for readings to stabilize before recording. • Take notes.
Measure the vibration amplitude with precision. If amplitude readings fluctuate there may be difficulties. Average the amplitude if you can.
Measure phase with precision. Try to get phase readings to within 2°. If phase is not steady there may be problems. Average the phase readings if you can.
Weigh trial and balance weights precisely. Use a precision scale to measure weight accurately. Digital scales are best.
Position weights exactly at the angle required. Use a fixed phase reference card or phase pointer. Position all weights within 2° if you can.
Note that errors reduce your effectiveness in arriving at a solution.
• An angle error of 7.5° limits balance improvements by 15%. • An angle error of 15° limits balance improvements by 25%. • An angle error of 30° limits balance improvements by 50%. • An angle error of 60° renders balance improvements to 0.
Check that balancing is progressing as it should, and use common sense in adding weights. Make certain that it is closing in on a solution and that you are not "chasing your tail".
Further considerations in the field....
• Ensure a good solid mount of the accelerometer. • Ensure a good solid mount of the laser. Make certain that it is not moving in any fashion. • Make certain not to store data until phase is not varying any more than 3 degrees • Set the Fmax value to a reasonable number that is about 6-10 times the running speed. • Set the number of averages to at least 2.