In replying directly to Jon Bertrand, I got an 'undeliverable' message, so I'm taking the liberty of posting this to the PICLIST. Perhaps some others will find it useful. pm --------------------------------------------------------- Subject: Re: Photodiodes Date: Thu, 12 Dec 1996 08:27:59 -0800 From: Paul Mathews Organization: Aengineering Co. To: Jon Bertrand References: 1 Jon Bertrand wrote: > > > Paul, > > Thanks for the response. Great ideas :) > > I'm not familiar with photo diodes - > > From your post I'd guess that placed in circuit reverse bias they will > "leak" more under high light conditions. Does this sound correct? > > Zero light gives < a few microamps, lots of light gives > few > microamps. If so it's exactly what I need. > > (The current gain of the transistor was actually a major problem for > me.) > > Thanks, > > Jon Bertrand > jonb@cirris.com > 2 ways to bias photodiodes: 1. Reverse bias, 'photoconductive', as you describe. Maximizes speed by minimizing capacitance, but increases dark current relative to (2) and may add power supply noise to signal. Highly linear within bias limits. 2. No bias, 'photovoltaic'. Photodiode can generate voltage, hence name. Anode goes positive, so you cannot generate much voltage, since the junction becomes forward biased and shunts away the current! However, is you use a low load impedance and keep the voltage low (say, under 200mV) this mode is also highly linear, and it has no dark current and no added noise from bias supply. Often used with a feedback amplifier: photodiode drives inverting input, which is held at virtual ground, constituting a very low impedance load. In either case, you can model the photodiode as a current source supply a current equal to about 0.5 Amps per incident Watt, provided that biasing is done properly. Lenses can be used to increase the collection area, or, you can obtain bigger photodiodes. Broad daylight has a power density of about 1000W per square meter, so, a photodiode 1mm x 1mm will have about 1uW incident and generate something approaching 500uA in broad daylight. Ordinary office lighting produces power densities about 2 orders of magnitude lower. -- -- Paul Mathews, consulting engineer AEngineering Co. optoeng@whidbey.com non-contact sensing and optoelectronics specialists