Solar Eclipse light level data

A few weeks back, I was fortunate enough to see the solar eclipse while in Japan.

Unfortunately, the timing of the eclipse put it early in the morning, right in the middle of my long commute to work. With my trusty eclipse viewing glasses in hand, I was able to catch glimpses of it while dashing between trains in the station. I fortunately got up extra early, so I could spend some time at each stop to take a look.

Sadly, the perfect circle occured while was in the subway, so I missed it. But I DID get several good shots with my pocket camera before and after. I took the shots looking through my viewing glasses. The cool thing was that the dark filter pretty much cut out all the cloud diffusion. Naturally, the sun itself was still diffused, but it was a bright orange circle on a black field.

I may have missed the ring, but my sensor's back home didn't!

The night before the eclipse, I spent most of the evening preparing to catch the morning sun with a TAOS TSL230R Light to Frequency sensor. This IC is mounted in a clear plastic dip package. It outputs a pwm signal proportionate to the amount of light striking the surface of the IC through the plastic case. It has multiple sensitivity ranges as well as scaling factors. The chip utilizes a photodiode array, which feeds a current to frequency converter.

I failed to attach psu decoupling caps on the power lines, which could account for some of the jitter. However, the eclipse light levels are clear, with a solid dip during the transit (around 7:30 AM, the lowest dip in the . The sensor was placed in my window, which faces the morning sun. After 10 am, the light levels here dip considerably due to trees blocking the sun. From here on, it only gets worse, as by noon, the sun is completely shaded by the house.

Throughout the day, cloud passage and sun passing through the treeline just outside the window are easily picked out.

Here is the original googledoc into which I logged the data: https://docs.google.com/spreadsheet/ccc?key=0ArBwm5DI5eb5dGJ1YmlZMWVUaWM4ZjRYQzVreHBpSWc#gid=0

I also recorded temperature and humidity data using a DHT-11 sensor. There are some bad readings (the big one element spikes) on the DHT-11, possibly due to (again) not including decoupling caps, or a bit of error in reading out the DHT11's slightly strange serial protocol.