As far as seeing what happens when the lights go out, all of my tests were run, with a fully charged battery of course, by shutting off power at the circuit box to the outlet into which the RS 1300 is plugged. The first test took place with a load of 241Watts as indicated by the unit itself and the APC PowerChute software. They also reported a run time of 18 minutes on battery. Note that APC’s runtime chart would indicate closer to a half-hour. The alarm sounded as expected and the software informed me from the taskbar that I was now on battery power. Power usage remained constant while I surfed the web and loaded some games. The alarm continued sounding every minute or so in a very annoying fashion.
Of course the power usage reporting from the software is anything but highly accurate and real time: its job is to let you know if you have too much stuff plugged in and to ball-park the remaining run time. The historical reporting of power events is off target as well; the first pic shows the results of my first test and a few previous events, claiming a total time on battery of 5 minutes and 15 seconds. The test ran 21 minutes 43 seconds, so numbers reported by the unit and software should definitely be taken with a grain of salt. The first shut down request (second picture) came after 15:42, when there were actually six minutes of time left. The last warning came at 19:31.
Test two took place with a reported consumption of 273 Watts (the increase due to CPU benchmarking) and runtime of 16 minutes. Note that APC’s website indicates the RS 1300 as having a 23 minute runtime at a 300 Watt load. Shutting off incoming power caused the alarm to sound and on screen notification to appear. This time I turned off the alarm. Playing music, running a scanner, and talking on a VoIP phone got the reported power consumption up to 288 watts. This fed my suspicions about the software’s accuracy as the scanner alone draws a good deal of power. The system ran for 18:07, with warnings after 13:07 and 17:00.
The third test involved turning off the printer and speakers, which resulted in a reported draw of 241 Watts and runtime of 20 minutes. This time the alarm, mercifully, did not go off. I would have liked a setting to have the alarm go off at the beginning of the power event then be quiet, but as we see in the third picture there is no such choice. I will be running silent from now on, and just use any other electrical device in the room to monitor changes in the state of incoming power instead of the incessant alarm. The overall run time was 22:27, with warnings at 17:43 and 20:47.