Phototransistor vs photodiode ?

[TechSpec]

HI all,

i'm currently working with IR sensor diodes, and trying to find a proper device for my project, however i haven't found a good unit which fills my needs. Now that i have extended my search to phototransistors, i wonder are they compatible with photodiodes?

Im using Atmel's IR-amplifier IC and the datasheet only suggest of using photodiodes, as the IC have only 2 input pins for diode. But can i also use phototransistors?, they have 3 pins, so how should i connect them to device which only have two input pins for diode?

Im thinking that, if the IR light represents as a base control "current" then i should use collector-emitter to be as a anode-kathode in diode, am i right?

TechSpec

 

[berkeman]

What IR source are you wanting to detect, and at what distance? What is the bandwidth of the modulation of the IR source?

PIN photodiodes are best for fast IR signals. You reverse bias them at several volts, and use an opamp circuit to convert the photocurrent to an output voltage. Phototransitors are slower to respond -- the photocurrent acts as base current into an NPN transistor. So you connect a phototransistor up with a grounded base and an collector resistor to the + supply rail, and use the collector voltage as the output. The advantage of the phototransistor is that you get the gain of the transistor, so you get more output response for the same input IR level, as compared to the PIN photodiode.

Does that help?

 

[TechSpec]

Actually, i skipped OP-circuits because some manufacturers offer ready made, highly developed amplifiers in one chip, like for example Atmel, those have built-in interference filter circuits and such as AGC(auto gain control) and ATC (auto strong signal adaptation) to be a complete amp unit, its tuff task to produce such a thing with discretes...

http://www.atmel.com/dyn/resources/prod_documents/doc4905.pdf

I'm thinking that since they have connected the diode internally to their amplifier, connecting a phototransistor would not work at all because of a different biasing etc.

My IR source is a 850nm 50mW LED and the detector is Pi detector at 850nm.
Emitter will emit 0.5ms bursts of 38kHz square at 50% ratio.
System uses optics so the distance will be >30m, and because the photodetector does not have daylight blocker, i will use IR-bandpass filter behind the optics to do the job.

TechSpec

 

[berkeman]

Yeah, from a quick look at the datasheet, it looks like it is made for a PIN photodiode. The biasing for a phototransistor would be different. Sounds like a fun project!