@CommanderLake: You just might be in a little over your current level of competence on this one.
Avalanche conduction can occur when a diode is reverse-biased sufficiently. It is distinguished from ordinary reverse-biased zener conduction by being a "positive feedback" mechanism that can quickly lead to self-destruction (through excessive current) unless "special measures" are taken to prevent this from happening. Photo-diodes operated as avalanche diodes are reverse-biased very near the avalanche conduction threshold, which allows an incoming photon to initiate an avalanche event resulting in many more avalanche conduction electrons flowing through the diode junction than the original photo-electron that was absorbed. These devices thus respond to single photon events.
Not ALL photons will be absorbed and produce an avalanche conduction... that depends on the quantum efficiency of the photo-diode... but QEs greater than 80% with some approaching 90% are not uncommon. Hamamatsu (Japan) is a world leader in the manufacture of this type of photon detector.
Here is a link to a very nice discussion of some of their avalanche photo-diode (APD) products.
Back in the late 1960s or early 1970s (I forget which) I was asked to help a newly degreed electrical engineer, who had "come up through the ranks" as an electronics technician, obtaining his BEE degree through part-time study.
He was trying to use an avalanche photo-diode to detect a phenomenon that pretty much everyone who knew anything at all believed didn't exist: a time delay in the onset of Faraday rotation of polarization of light. The Air Force was VERY interested in proving the so-called Allison Effect existed and could be used to identify contamination of lubricants and fuel down to levels of a few parts per billion. They had some in-house money to spend on this project, which was not very sophisticated in its design, but it did depend on a human observer manipulating several lecher-line shorting bars and observing a dimming of the nearly nulled, polarized light, optical field.
The lecher lines were connected across a spark-gap, discharging the lines periodically, and whose brief arc-light illuminated the optical path of the instrument. I never was afforded the opportunity to "try out" the instrument, but all the experiments and adjustments had to made in near-total darkness to allow the human eye to become "dark accommodated," so the dimming of the visible field could be observed as the shorting bars on the lecher lines were adjusted. However, not all observers were equally successful in operating the lecher line controls, nor were their setting of the controls consistent among observers. For this reason, mainly, the Air Force insisted on a demonstration using photo detectors and electronic instrumentation.
IIRC, many types of photo detectors were tried unsuccessfully, beginning with photo-multiplier tubes and finally ending with PIN diodes and avalanche photo diodes. There were lots of problems with stray electromagnetic interference because of the lecher lines (exposed in air) and the high-voltage discharge of the spark-gap illumination source. These lecher-line currents were also used to excite a solenoid around the transparent liquid-sample holder, said current being switched on and off by the spark-gap switch.
By the time I got involved, Herb (the technician who was now an engineer) had managed to destroy his very expensive avalanche photo diode by reverse-biasing it with DC without an adequate means to limit the avalanche conduction current. He hadn't even taken the precaution of operating the device in total darkness while messing around with the bias, much less used a large enough resistance to protect the device. His reasoning was he needed a low impedance because they were looking for wide bandwidth events. Meanwhile, the Air Force was running out of in-house discretionary money to fund this project, so the decision was made by "the powers that be" to write a report with lots of equations and fancy graphs that no one would read and shut the project down. Which they did.
About ten years later I finally got my own BEE degree, but I never again got to work with avalanche photo diodes, although I did get to work with plenty of other photo detectors... and lasers... and other spiffy electro-optical devices. Moral of this story? Probably none, but I would be very careful working with your avalanche photo diode. Without adequate current limiting, it is very easy to damage the device while it is under reverse bias by exposing it to ambient light. Star light is probably okay (except for our Sun), but I would be careful with a full moon in a clear sky. A pin-hole aperture with light reflecting off a spinning wire might be a good way to excite it with a very fast and narrow pulse of light. I would also consider using pulsed-DC reverse bias synchronized to the spinning wire for sensor excitation.