I was reviewing this 2014 reference design by Caldwell,\ (TIDU535), I believe he misuses the "photo-conductive" term here - his circuit at this point is zero bias on the photo-diode which is commonly called photo-voltaic where Photo-conductive is used for reverse biased detectors,
Here is one of many references that make this distinction, If I am missing something, and I have not gotten all the way through this Caldwell document, let me know.
Hi Michael,
photovoltaic is when the photodiode is allowed to work like a solar cell, or by other words, when the anode of photodiode has a positive potential referring to the cathode, including the case that the voltage is zero (zero bias). Photoconductive is when the photodiode is not allowed to work like a solar cell, or by other words, when the anode of photodiode has a negative potential referring to the cathode.
A photodiode is made from p-doped and n-doped semiconductor layers at the contacts. Space charge regions are forming within the pn-junction with a majority of electrons sitting in the p-doped contact layer and a majority of holes in the n-doped contact layer. As result a depletion area is forming with no electrons in the conduction band. And an electrical field "E" is forming by the space charge regions which is oppositely directed to the p-doped and n-doped layers.
When a photon is hitting the depletion area, a free hole/electron pair is created. "Free" means that it is pushed from the valence band into the conduction band. Due to the internal, oppositely directed electrical field "E" the hole/electron pair is sorted and the holes are running to the p-doped anode and the electrons to the n-doped cathode. They both form the so called "photocurrent".
When a load "R" is externally connected from the anode to the cathode of photodiode, a current (= photocurrent) is flowing out of the anode ("technical" current direction!), allowing the anode to develop a positive potential referring to the cathode, as consequence of the voltage drop across the load "R" caused by the photocurrent. ("Technical" current is assuming free positive charge carriers making the current flow.)
This mechanism is called "photovoltaic effect", in opposition to "photoelectric effect", where the electrons are leaving the material being hit by the photons. Concretely, the photovoltaic effect means the creation of free charge carriers, being pushed from the valence band to the conduction band by photon energy. And in the photodiode they are even able to reach the contacts.
The funny thing is, that the photocurrent mechanism is always the same whether the photodiode is operating in photovoltaic mode or photoconductive mode. The "motor" of the photocurrent is always the same internal, oppositely directed electrical field "E" which sorts the holes and electrones and makes them running to the opposite contacts. The only difference is the width of depletion area and the strength of electrical field "E" and the associated effects like junction capacitance, dark current, quantum efficiency, a.s.o.
The current flow is only different, when you connect a high forward voltage from the anode to the cathode. Then the photodiode starts to behave like a simple diode with the forward current dominating the photocurrent. The forward current is flowing internally of photodiode from the anode to the cathode. But if only the photocurrent is flowing, it will always flow internally from the cathode to the anode. The current direction is identical to the current direction in a battery or accumulator then, where a chemical process creates an EMF and forcing an internal current to flow from the minus pole to the plus pole, which outside the battery flows from the plus pole back to the minus pole.
It's interesting to note, that a photodiode being installed in a zero bias TIA, is indeed operating in photovoltaic mode. Not only because zero bias is per definition meaning the photovoltaic mode, but also, because due to the non-ideal, finite open loop gain of a real TIA the photodiode is (at least from theoretical point of view) always operating in forward direction. Of course, the forward voltage will be quasi zero in a good TIA. Also, the input offset voltage of TIA (and its input bias current) should not be forgotten, which usually dominates the required input voltage caused by the finite open loop gain of TIA.
In the following simulation an ideal OPAmp with zero input offset voltage and zero input bias currents but very non-ideally low open loop gain of 80dB is assumed to better demonstrate the effect of finite open loop gain on the polarization of zero biased photodiode in a TIA:
The left side and the right side show the same situation. For a better demonstration the photodiode is connected to make the TIA output a negative voltage. As you can see the photodiode is indeed connected in forward direction 
Yes, I know it's improbable, but maybe this effect of finite open loop gain is the reason why the operation in a zero bias TIA is called photovoltaic mode? At least, it helps me to remember that zero biasing is called "photovoltaic mode".
Kai
Hi Michael,
there's no big diffrence. The photodiode is still connected in forward direction:
Kai
Hi Michael and Kai,
Interesting and good discussion!
Whenever I do photometric measurement, the photo diodes are always reverse biased in order to increase the linearity. For the simulation purposes, the modes of operation in photodiodes and photoconductors are less considered or addressed...
Below is Thorlabs' app note on the topic.
https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=9020
Best,
Raymond
Hi Michael,
such a circuit with a load resistor as shown below might make sense for "event detection" in a physical experminent or for optical communications, but not at all for precise light level measurements, especially not in low light level applications. You will suffer from all the disadvantages a good TIA is avoiding.
Kai
