****Calculating the thermistor temperature*****
From the voltage divider and AtoD converter:
Rthermistor = R0*(1024/D1-1), R0 = 10(k)

From fitting a function to the thermistor data sheet temp-resistance conversion table:
Temp = -18*ln((Rthermistor-0.7912)/29.7746)


*****Calculating the solar panel output*****

Vpv = 5V*D0/1024

Ppv = (5V*D0/1024)^2/Rshunt, Rshunt = 27 ohms

*****Photometers*****
Photometer 1 is on the side and photometer 2 is on the top (with the solar panel)

*****Magnetometer*****
The registers are set to 110 which means the gain for X and Y should be 320 units/Gauss and the gain for Z should be 285 unit/Gauss. Using these gains, the values agree fairly well with what is expected (see http://www.ngdc.noaa.gov/geomagmodels/IGRFWMM.jsp).

*****Accelerometer*****
Taking the average of the first values of AccX, AccY, and AccZ, squaring them, and summing them, should result in 9.8^2 because the payload was initially at rest on the ground. Thus, dividing (mean(AccX(1:150,:))^2+mean(AccY(1:150,:))^2+mean(AccZ(1:150,:))^2)/9.8^2 should give the square of the scaling factor the accelerometer values should be divided by  (c^2). Computing this gives c=103. Calculating mean(AccX(1:150,:))/103 now for each axis will show that the x-axis is actually oriented pointing from down to up (z-axis). However,  there are small values for the Y and Z accelerometers even though they should be zero because the board is not perfectly flat in the payload. Thus to calculate the correct X, Y, and Z accelerometer values plug the raw values into:
AccZ = AccX/103+0.029
AccX = AccY/103+0.604
AccY = AccZ/103-0.936

*****Making the DNT and HAM time axes agree*****
DNTtime = floor(time/100)*60+mod(time,100)-1395
HAMtime = 3600*(hh-hh(1))+60*(mm-mm(1))+(ss-ss(1))-346

*****Explain PV readings*****
The correlations between the solar panel current/voltage/power and the top photometer were 0.6515, 0.6253, and 0.6515 respectively. This makes sense since I and V are proportional since R is a constant. The top photometer had the greatest correlation with the solar panel, and even showed dip and rise behavior. It does not explain what caused this behavior though, except that it had to do with sunlight (not temperature, because the other photometer looks flat in that region).

Currently, my calculate rotation around the Z axis and from the Z axis, have lower correlations with the solar panel than the raw Mag axes alone. However, I think I might be able to combine the acc and mag data to calculate the angle between the solar panel's normal and a straight line to the sun. I expect this will have a higher correlation than any of the mags alone since it will take into account not only how much the panel is turned from straight up, but also in which direction it is turned.

The temperature shows a slight positive correlation with the solar panel voltage, which suggests the panel works more efficiently at higher temperatures (in contrast to what I expected). On the data sheet: Temp. Coeff. Vmp = -6.2 mV/°C which suggests that the voltage at maximum power decreases as temperature increases. However, I am not sure the solar panel was running at max power (the resistor was sized to be close the the MPP, but it is possible the conditions, light and temp, shifted it in one direction or another). Supposedly, the open circuit voltage is what is most effected by temperature. The 27 ohms resistor I used is closer to a short circuit.

Sun at 10:10am at the Mt Washington Airport.
Az 155.71 deg
El 65.58 deg
Source: http://www.esrl.noaa.gov/gmd/grad/solcalc/

The sun was pretty close to straight along Earth's magnetic field, so whichever axis of the magnetometer was aligned with the solar panel would be strongly related to the solar flux.