Help regarding fast response photodetection

[Shikhar]

Hi there...
i'm working on my college project and wanted to design a high speed photosensor. The intensity of light source is quite high (actually a bomb explosion), so noise signals are not a problem. The photosensor is required to trigger other sensing devices.
So which type of photosensor would suit me best- a photodiode or phototransistor?? and why...

i would really b obliged if you could provide me with the required circuit(implementation point of view)...

Thanks
Shikhar

 

[berkeman]

Hi there...
i'm working on my college project and wanted to design a high speed photosensor. The intensity of light source is quite high (actually a bomb explosion), so noise signals are not a problem. The photosensor is required to trigger other sensing devices.
So which type of photosensor would suit me best- a photodiode or phototransistor?? and why...

i would really b obliged if you could provide me with the required circuit(implementation point of view)...

Thanks
Shikhar

Use a photodiode with a reverse bias of 5-10 volts or so. Use it in the traditional current-to-voltage converter configuration, with a fast opamp. You should be able to get fairly fast response on the order of a few us, I think.

Quiz question -- why should you reverse bias the photodiode instead of running it with zero reversse bias? It's an important concept related to the speed...

An even faster option would be to use a fiberoptic receiver module, which has the photodiode and converter circuit built together. You can get them with logic output capability. HP made some that I've used in the past.

 

[Bob S]

I have used photomultipliers (RCA 2020) and vacuum photodiodes (RCA 935) to measure sub-nanosecond transit times, using an analog time to voltage converter..

Bob S

 

[berkeman]

I have used photomultipliers (RCA 2020) and vacuum photodiodes (RCA 935) to measure sub-nanosecond transit times, using an analog time to voltage converter..

Bob S

Show-off!

 

[Shikhar]

Quiz question -- why should you reverse bias the photodiode instead of running it with zero reversse bias? It's an important concept related to the speed...

thanks Berkeman!

hmm..i think...it's because photodiodes work on the principle of generation of electron-hole pairs when a light of appropriate wavlength strikes the junction. Now, more the reverse bias, faster would be the speed with which there carriers reach the measuring circuit...
am i right?

please tell me one more thing..can i use a comparator for the voltage output from the photodiode circuit so that i get a TTL output (this is what i actually want as the light sensing is just to trigger other instruments)

Shikhar

 

[Mike_In_Plano]

The PIN photodiodes will get you there fastest. Most of the carriers form in the I region and as long as you apply a reverse bias to the device, they will drift out rapidly (i.e. 5ns) driven by the e-field.

You will need some gain with the PIN diode. Most times, it is used in a trans-conductance amplifier configuration. However, if you have plenty of signal, and really want a really fast response, it may prove easier to output it into a 50-100 ohm load and follow it with a video amplifier.

Photo transistors are slow dogs which often rely upon slow carrier recombination to improve their gain. It also takes time for an unbiased photo transistor to reach operating point. If your looking for just the flash, you can bias the photo transistor and it should turn on fairly quick (not near as fast as a PIN). It just won't turn off very fast.

- Mike

 

[berkeman]

Quiz question -- why should you reverse bias the photodiode instead of running it with zero reversse bias? It's an important concept related to the speed...

thanks Berkeman!

hmm..i think...it's because photodiodes work on the principle of generation of electron-hole pairs when a light of appropriate wavlength strikes the junction. Now, more the reverse bias, faster would be the speed with which there carriers reach the measuring circuit...
am i right?

please tell me one more thing..can i use a comparator for the voltage output from the photodiode circuit so that i get a TTL output (this is what i actually want as the light sensing is just to trigger other instruments)

Shikhar

Yes, you can follow the opamp with a comparator to give you logic levels. Be sure to use a fast comparator, and add some hysteresis to the comparator feedback.

On the quiz question, no, the main effect is capacitance. How does the capacitance of reverse-biased PN junction vary with reverse bias voltage? Why is that important in the photodiode detector case?

 

[snarfherder]

On the quiz question, no, the main effect is capacitance. How does the capacitance of reverse-biased PN junction vary with reverse bias voltage? Why is that important in the photodiode detector case?

the interface of a reverse biased pn junction is depleted of majority carriers (depletion region = comprised of minority carriers, which respond extremely fast compared to majority carriers, meaning low capacitance). Light impinging on the pn junction excites majority carriers into the conduction band, and they are swept across the junction by the imposed electric field.

a forward biased pn junction has a build up of majority carriers, resulting in an extremely thin depletion region. Carriers excited into the conduction band move across the junction slowly because of this build up.

 

[Shikhar]

On the quiz question, no, the main effect is capacitance. How does the capacitance of reverse-biased PN junction vary with reverse bias voltage? Why is that important in the photodiode detector case?

ohkk...i get the clue!
on reverse bias, width of depletion region increases, so capacitance decreases as C is inversely proportional to 'd'.
NOw as time constant = RC, so time constant decreases ie. faster operation!