“A twin cell tower arrangement was installed as a replacement on the roof of an old 20-story building. Vibration was noticed by office personal on the 20th floor about a year after the installation. An initial survey indicated that the vibration was very high in both cells at a frequency of 2,625 CPM. The #2 cell also had a high level of vibration at fan speed —675 RPM. The cause of the 2,625 CPM frequency was not clear; the #2 cell needed to be balanced. Initially it was thought that balancing the fan would be simple, but after repeated attempts to take phase data with an optical photocell tack, it became clear that the stainless steel sides on the tower were reflecting back and being picked up by the photocell. At this time an unusual solution to the problem was tried. A short length of PVC pipe was cut and attached to the photocell eye. The PVC pipe was moved close to the rotating shaft, the shaft was painted, and the problem of the reading at three and four times the frequency of operating speed was resolved (Figure 1). The residual balance on the #2 cell was reduced at fan operating frequency from 0.82 IPS to 0.06 IPS filtered to 1X.”