When taking measurements of the total light power output of an LED using an integrating sphere the results vary each time by up to 50%. I might expect a variance of around 5% due to the device itself, but such a large deviation is obviously impractical.
Aside from the position of the device and detector not being constant, can anyone suggest a reason for this? Any help and suggestions are welcome.
Hello,
I think the main aspect is the position and orientation of detector and LED. Furthermore, the geometry is actually the only problem if the integrating sphere is a good (R=98-99%) sphere. Are you measuring in a 90º configuration (i.e. angle between LED aperture and detector is 90º, and both of them radially aligned. Therefore the sphere only has two orifices)? If so, light from the led cannot reach directly the detector (a white obstacle should be placed in between if from the detector's position you can directly 'see' the LED) and geometrical consideration might be avoided.
OK thanks for that Juan, I thought I might be missing something else obvious. I'll post back to let you know how I get on.
Hi,
Perhaps you selected the wrong diameter of the sphere. And to assess the sensitivity of the detector dynamic range.
Dear Matthew, I suggest to use a thermoelectric cooler as well as a temperature sensor in back of your LEDs to make sure that you are doing your experiments in a same temperature (of course if you are not using them now!)
- Badges
- Science method