Atmospheric CO2 – Photodiodes & Signal Range

A receiving photodiode operates in the same manner as any reverse-biased diode (pn-junction) except for being optimized for such operation. The diode is used in the “OFF” state and only very small leakage current flows. This leakage current is called dark current in a photodiode; it is the current which flows absent any optical stimulus and may vary with applied bias (if any). The specific details of semiconductor operation are deferred.

Dark current defines the optical detection floor; adding photons only increases this current. Dark current may be reduced by cooling, but only as a factor of square root – a large decrease in temperature is necessary to make a significant difference .

It is usually more important the temperature be stable.

All diodes – semiconductor junctions – have an optical response; additional noise can be introduced in non-photodiode applications by optical leakage through pin leads of certain electronic components such as operational amplifiers or ADCs.

The photodiode is optimized to respond to specific optical wavelengths. In the majority of the response range, the relationship between optical power and current is linear – often on the order of 1A/W. Certain multiplier structures can increase this to perhaps 20 – 200 A/W at a cost of increased noise and complex circuitry. Dark currents may be on the order of 50pA – 200 nA. This variation in structure does not affect the following discussion.

The diode also has a maximum limit; too much optical power can damage or destroy the device – this limit may be in the mA – mA range. A diode may have an effective output range of perhaps 50 pA to 5 mA which corresponds to 8 decades of current. However, the linear response region is less than this but 5-6 decades of usable signal range is not unusual.

The above figure illustrates the detector compliance range. The optical signal needs to remain somewhat above the dark current and somewhat below the saturation current to minimize distortion.

Of greater concern is the range of common-mode signal as shown below:

Each differential-mode signal is identical; perhaps the ratio of absorption to scattering elements. However, the effects of common-mode variation move the difference-signal up and down on the responsivity range.

Common-mode level 1 is too high; the difference signal is distorted at the high end. Measures at levels 2 and 3 would result in identical values, while common-mode level 4 is too low; the signal is not sufficient to properly activate the photodetector.

Measures would be obtained in all 4 examples, but there would be no way to differentiate between them and the inclusion of the distorted measures would add unquantifiable uncertainty to the result.

It is the task of defining the expected optical signal range which led to this discussion. Design of the detector system needs to be based on the expected range of intensity levels.

That’s all for now.

Next 10: Avalanche Photodiodes