With the August temperature in the low hundreds and corresponding high humidity, a lightning event was almost a sure thing.
Wireless communications have become a routine part of our lives. We expect to place and receive calls wherever and whenever we want, in any kind of weather. Last August, however, one cellular service provider (CSP) experienced a mysterious problem at several of its newest cell sites in western Texas. The CSP hoped to attract a significant number of new customers with its new call capacity. Instead, the company found that the digital switching equipment would disconnect and then reconnect (drop out) at random sites.
The CSP requested the utility company's assistance in determining the problem. Utility personnel monitored the AC supply side of the AC-to-DC conversion equipment with several power-disturbance analyzers. Test results indicated that the power company delivered high-quality AC power with no interruptions, sags, or swells — in short, a pretty picture.
Next, the CSP contacted the manufacturer of its AC-to-DC power conversion package. With every cell site dropout, there had been severe thunder and lightning in the general vicinity. The power conversion manufacturer suspected that a high-speed, high-voltage disturbance on the AC supply side caused a DC fault, which interrupted the DC supply voltage to the cell sites.
To test this theory, the manufacturer consulted our application engineers, who recommended using a disturbance analyzer that could monitor high-speed transients on both the AC current supply side and DC current load side of the power conversion equipment. With their gear in tow, the engineers arrived at the manufacturer's design center for a few tests before heading to the field.
At the design center, simple tests were conducted to answer the following question: Can you measure a 6000V pulse with a rise time of 1 microsecond? The engineers assembled the monitoring equipment and connected the high-speed disturbance analyzer and a laptop PC to a 115VAC line monitor. A pulse generator injected a 6000V pulse into a 115V line to simulate a lightning strike in the AC system. When a team member pressed the start button on the pulse generator, a sharp, loud crack sounded. The laptop's display screen went blank, but the flashing annunciator on the front of the disturbance analyzer indicated it was online and acquiring data. The analyzer was capable of measuring spikes up to 6000V, but what about the laptop? With crossed fingers, the team rebooted the computer. To the members' delight, the screen display on the laptop returned. With the computer operating properly (see Fig. 1, on page 48), the team downloaded the test data from the still-monitoring disturbance analyzer. The next stop: a remote cell site in western Texas.
In the Field
The local utility's power quality engineers joined the team at the chosen site. With the August temperature in the low hundreds and corresponding high humidity, a lightning event was almost a sure thing.
The team connected the high-speed disturbance analyzer to the supply and load sides of the conversion equipment. The utility's engineers connected their most reliable power analyzer to the AC supply side at the same time, so both groups could compare data.
That evening, the skies grew dark and thunder crashed overhead. The team wondered if impulses caused the cell site to go “off the air.” To find out, they dialed up the disturbance analyzer's modem to check for any impulse hits at the site. There were several, each with a corresponding interruption in the DC load side.
The next day, the team compared the AC supply data of the different disturbance analyzers. The utility's analyzer, which was unable to measure DC, did not record the previous evening's disturbances because its fastest sampling speed was 1 million samples per second. The high-speed analyzer sampled at 10 million samples per second, detecting the lightning disturbances that caused the mysterious dropouts.
Once the team knew that lightning was responsible, they took the appropriate preventive measures. After that, the CSP had no further problems.
What this case history demonstrates is the need for appropriate monitoring equipment in certain high-speed disturbance situations and, equally important, the successful cooperation between all involved parties.