A strange thing occurred this week, several days of cloudless skies. This afforded me an opportunity to measure the variation and trend in PV output over a day without the interference of cloud. The fact that this happened in the middle of February is surprising but less onerous than having to get up at the crack of dawn in mid Summer! I was not surprised at the basic results – I always assumed that the output over a day would be non-linear and I also expected my peak output to occur in early afternoon as my panels are facing SW. In the event, I was pleasantly surprised at the perfect curve I got and the peak output came about 2 hours after true noon. The falloff in late afternoon was more sudden than the slow build-up over the morning.