WTH!? Why BJT act like a photo diode/transistor?

[ali6x944]

Hi everyone,
I was playing around with a BC107 NPN transistor and I accidentally fried it, so I went and removed the metal cap to reveal the transistor's silicon die just for fun and I notesed that the resistance of the collector-emitter junction droped as soon as light shined on the die , similarly as soon as a shadow is casted on the die the resistance increased clearly
So I ask what is actually happening here and how and is it possible to use other types of transistor s in the same manner as a photoresistor/ photodiode/ phototransistor?

[Audioguru]

A photodiode is a diode in a see through package. Try a 1N4148 diode or an LED.
A phototransistor is a .... You tried it and it works.

[ali6x944]

But how dose it work and why?
It is absolutely confusing bucause from what I know BJT's are current operated devices thus thay need a current flowing into or out of the base to the emitter, so whyyyyyyy
It is doing so!?

[Zero999]

It works the same as any other phototransistor. The base-emitter junction generates a current when exposed to light, which is internally amplified by the BJT and can be used to control a larger current flowing from the emitter to collector.

If you fried the transistor then I'm surprised it worked. I suppose you didn't completely destroy it, only damage it.

[retrolefty]

But how dose it work and why?
It is absolutely confusing bucause from what I know BJT's are current operated devices thus thay need a current flowing into or out of the base to the emitter, so whyyyyyyy
It is doing so!?

 Photons striking the junctions, most silicon semiconductors exhibit this.

 http://ecee.colorado.edu/~bart/book/book/chapter4/ch4_6.htm

[ali6x944]

Thanx everyone that was really helpful
Thanx

[newbrain]

Reported as true story at my job (I was not there at the time, but shortly after, relata refero):
Rome 1990 World Football Championship.
A brand new telecom equipment installed at the stadium, tested, and perfectly working.
Then it started.
Every day, at approximately the same time, system failures with system restart (the expected down time should be less than 2.5 minutes/year).

The best technicians could not understand, until was discovered that tours with various countries' representatives were organized to show off the new equipment, and hence it also had a board on a board extender (to have it out of the rack, usually for debugging/test).

Some of them used a camera, with flash!
The flash was enough to have some diodes (in glass package) reverse conduct and bring down the system.

[ali6x944]

Reported as true story at my job (I was not there at the time, but shortly after, relata refero):
Rome 1990 World Football Championship.
A brand new telecom equipment installed at the stadium, tested, and perfectly working.
Then it started.
Every day, at approximately the same time, system failures with system restart (the expected down time should be less than 2.5 minutes/year).

The best technicians could not understand, until was discovered that tours with various countries' representatives were organized to show off the new equipment, and hence it also had a board on a board extender (to have it out of the rack, usually for debugging/test).

Some of them used a camera, with flash!
The flash was enough to have some diodes (in glass package) reverse conduct and bring down the system.

Man! Really !? That must took an eternity to pin down ...
How did they figured this out?

[uncle_bob]

Hi

Back in 1975, we are "big stuff" making the guts of these cool new digital watch gizmos. Our main marketing guy (my ex-boss) is driving down the road in Phoenix AZ. Arm out of the car window. Watch on the wrist hanging in the breeze. Looks at watch ... watch is dead. Swaps watch with spare, goes on to meeting.

Back at the factory, we get to figure out what went wrong with the magic watch. Turns out, the LCD display is somewhat transparent. With enough sunlight, you illuminate the silicon die on the ceramic behind the display. Enough photons into the poor thing and it latches up. No more clear die coat on that IC from then on ....

Bob

[ali6x944]

Dose this happen with fets and cmos devices?

[uncle_bob]

Dose this happen with fets and cmos devices?

Hi

Yes it most certainly does. The IC that latched up on us was a CMOS part.

Bob

[rs20]

Let's not forget the photo-shy raspberry pi!

https://youtu.be/SrDfRCi1UV0

[T3sl4co1l]

Light forward-biases the transistor, in nearly exactly the same way that forward-biasing the B-E junction does.

Indeed, a photoresistor is a piece of consistently doped semiconductor (no junctions).  Which describes a UJT (unijunction transistor), except that its resistance is controlled by forward-biasing a diode junction attached to it (the "emitter" spews free charge carriers into it, decreasing its resistance).  Suppose we added a diode junction, externally, that produced the same effect?



Tim

[TimFox]

This experience is relevant to a problem often seen with op-amp circuits.  It is common to protect the differential input with two silicon diodes back-to-back across the input terminals.  Most leaded signal diodes are in glass packages, and the photocurrent can affect the offset of the amplifier.  Better to use black plastic packaged devices, or to encase the glass diodes in black heat shrink tubing.  Glass zener diodes are especially photosensitive, since they have large junction areas.

[albert22]

Many years ago this was a "poor´s man phototransistor"...

The OCP71 was a Photo Transistor being a clear enveloped version of OC71 germanium transistor. It could be used as a photo diode.
 The original OCP71 germanium photo-transistor (filled with clear gunge) cost an obscene amount of hard earned cash - twelve shillings and six pence whereas the OC71 typically cost four shillings and six pence (still a lot when pocket money was earned rather than just given). Since the only discernible difference was the thin black paint of the outer envelope and the first OC71 transistors had clear gunge, so the black paint was often scraped off to achieve a lower cost equivalent photo-transistor thus saving eight shillings

http://www.museumoftechnology.org.uk/expand.php?key=1178

[T3sl4co1l]

FWIW, I measured a common 1N4148 as developing a tiny voltage (< 0.2V? I forget) and leaking very little (~10s nA?), when exposed to a rather intense white LED flashlight (~1W optical output) at point blank range (LED lens touching glass diode body in the middle).

The difference is small (only about 2-3 times the normal dark/ambient leakage), but measurable.

Other types or brands may vary.  Flash photography is probably even more intense, even at a distance, so could conceivably cause problems in sensitive circuits (10s uA?).

But probably not well-designed circuits; since, after all, why are you using a ~nA leakage diode around ~uA circuits, regardless of flash sensitivity?  You should use a transistor C-B junction instead (which is also hermetically sealed, or encased in black plastic, as the case may be).

Tim

[TimFox]

I use the base-collector diode of low-noise bipolar transistors (emitter shorted to base), such as 2N2484 (metal) or 2N4250 (plastic).  Gate-channel diodes of low-current JFETs are good for even lower leakage, but have low maximum current ratings.

[ali6x944]

What about we all do a series og experiments on semiconductor devices recording the light frequency and intensity and the device's conductivity , est...
Everyone make measurements on one device and by the end we could get a huge database of devices and their values, how dose that sounds?

[retrolefty]

What about we all do a series og experiments on semiconductor devices recording the light frequency and intensity and the device's conductivity , est...
Everyone make measurements on one device and by the end we could get a huge database of devices and their values, how dose that sounds?

Boring, sorry.   

[ali6x944]

Ok ....
Thanks for all the help thou