Maker Pro
Maker Pro

Emitter Base Breakdown

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
What happens to the collector current of an npn transistor when the permitted base emitter reverse voltage is exeeded.

Using LTspice for a BC547B does not allow for junction leakage. The base voltage goes to -100V with no limit.
 

Attachments

  • Clipboard01.png
    Clipboard01.png
    35.2 KB · Views: 239

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Interesting question. I thought exceeding that voltage would damage the transistor.

Bob
 

john monks

Mar 9, 2012
685
Joined
Mar 9, 2012
Messages
685
Just for duke37. I just tried the experiment with a 2N2369A. I don't have a BC547B.
I used a 12 volt supply.
I used a -12 volt tail generator 50 ohm output impedance.
Rc is 1k.
Rb is 1k.
I supplied a -12 volt tail pulse to the other end of Rb. The transistor base emitter conducted at about -7.5 volts and the collector remained at 12 volts continuously.
I removed Rb and the collector started getting about a negative 30mv pulse riding on a 12 volt baseline.

So your answer is essentially nothing.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Thank for the replies and the experiment.

Reverse current through the emitter/base junction will not necessarily create damage. I have a Heathkit HW100 transceiver which uses a transistor junction as a Zener diode.

Duke
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
I guess it depends on the amout of current!

Bob
 

john monks

Mar 9, 2012
685
Joined
Mar 9, 2012
Messages
685
duke37, you are quite right. In fact I built a 1000 volt power supply for the purpose doing high voltage experiments with semiconductors running transistors in the avalanche and testing the reverse characteristics of diodes. I discovered that practically no PN junction is damaged by running in the avalanche mode at reasonable power levels. In fact I discovered that the 1N4152 and the 1N4154 make good 90 volt zener diodes.
So in many ways semiconductors are very forgiving.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Here is an interesting document that discuses the potential for failure.
 

john monks

Mar 9, 2012
685
Joined
Mar 9, 2012
Messages
685
Very interesting document. I am looking at the damage characteristic plots and I see that some of there devices were powered over 1 million watts before failure. The transistor plots seem to be close to what I got although I used different transistors.
These plots are very important because in the pulse generator business your running parts well outside of the maximum specification in order to get minimum widths and maximum power levels without failure. Thanks steve.
 

EinarA

Feb 11, 2013
28
Joined
Feb 11, 2013
Messages
28
Breakdown of the BE junction does damage the transistor , the beta is permanently decreased every time this happens. Transistors make excellent zeners: the knee is much sharper but the voltage varies widely from one device to another even within the same type. I have read that you can measure a small negative voltage at the collector, supposedly from IR light from the Zener. Another use is as a white noise source, a2N2222 produces a very strong signal with other types much softer.
 

john monks

Mar 9, 2012
685
Joined
Mar 9, 2012
Messages
685
EinarA, where can I get more information on the damaging affect of a reversed BE junction and the IR affect on the collector voltage? I never thought about the IR.
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
Spice in general doesn't simulate BE breakdown.
You can partly model it by including two diodes (one standard, one zener) into the circuit:
attachment.php


Using this method gives noteably different results if you simulate e.g. an astable multivibrator at different supply voltages.

If you want to study the effect in more detail, see these references:
http://jeastham.blogspot.de/2009/11/modeling-diode-reverse-breakdown.html
http://www.vk2zay.net/article/196
 

Attachments

  • bla.gif
    bla.gif
    4.4 KB · Views: 379
Last edited:

john monks

Mar 9, 2012
685
Joined
Mar 9, 2012
Messages
685
That's interesting but I was mainly interested in the statement that EinarA made. He said that breakdown of the BE junction does damage the transistor, that this reduces beta. I have not seen this and I have not read that anywhere else. If this is true it is worth it for me to research this. I'm in an industry that uses this effect to create random noise.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
That's interesting but I was mainly interested in the statement that EinarA made. He said that breakdown of the BE junction does damage the transistor, that this reduces beta. I have not seen this and I have not read that anywhere else. If this is true it is worth it for me to research this. I'm in an industry that uses this effect to create random noise.

I have heard the same mentioned as well. However I have not been able to find any documented evidence of it.

I suspect that one method for this to occur is due to localized heating of the junction whilst conducting in the reverse direction.

Potentially this could cause migration of doping atoms, or perhaps even damage similar to secondary breakdown.

But all of this is conjecture (by me) and I would not dismiss other theories (or indeed accept these) without some good evidence.

If anyone can find some, I'd be very interested.

edit: This looks like it could be interesting. And this has some background.
 
Last edited:

EinarA

Feb 11, 2013
28
Joined
Feb 11, 2013
Messages
28
John, I'm not sure where I read about the photoelectric effect, it might have been in one of Bob Pease's columns. I don't know why no one is turning up references to beta reduction as this is well known in the professional literature.
 

john monks

Mar 9, 2012
685
Joined
Mar 9, 2012
Messages
685
I intend to run the base emitter junction of a transistor in the zener mode over the weekend and see if the beta degrades. I intend to report back. I suspect it won't.
 

EinarA

Feb 11, 2013
28
Joined
Feb 11, 2013
Messages
28
Why should I bother to post anything if no one believes it. This effect is well reported in real engineering magazines and has been tested by me. To get the data given in the first post I tested about 6 transistors. I didnt test all for beta loss, but two 2N2369s started out at 60 and only measyred 20 when I finished.
 

john monks

Mar 9, 2012
685
Joined
Mar 9, 2012
Messages
685
Sorry, You see I work on transistor circuits that run in the avalanche mode, the zener mode, and very high current pulse mode. It is important for me to know if something is going to fail. I suspect that you drove the 2N2369s with much more current than what I am doing in our company. I drive them with a few hundred microamps. Now I am testing a 2N2369A with about 7 milliamps in the zener mode. I started out with a beta of about 30 and I will measure it again Monday. I thank you for the information.
 

EinarA

Feb 11, 2013
28
Joined
Feb 11, 2013
Messages
28
You can take it out and measure it right now. The damage is done in the first instant. My notebook records show voltage at 1,10, and 100 uA. P S One of those transistors now measures 15. I'm not willing to test any of the 2222s I use for noise sources as I solder B to C so the matel can acts as a shield, but I think this transistor is less affected.
 
Last edited:

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Why should I bother to post anything if no one believes it.

If we didn't believe you, this thread would have been much shorter.

I have speculated on the cause. I am skeptical that it happens at any reverse current, but skeptical means that I'm willing to be shown to be wrong.

This effect is well reported in real engineering magazines

As I said, I've heard of it but I've not seen any references. If you can point us toward any it would be really interesting.

and has been tested by me.

And indeed there are people here trying it for themselves.

To get the data given in the first post I tested about 6 transistors. I didnt test all for beta loss, but two 2N2369s started out at 60 and only measyred 20 when I finished.

Sorry, I can't quite figure what went from 60 to 20? Was it the small negative voltage on the collector?

Going back to your original question, I think you need to test this with real rather than simulated components. The operating region is so far outside where (I suspect) the model is valid that I wouldn't trust it.

An interesting note earlier about IR emitted from the junction. having a photoelectric effect on the other junction. Not something I would have considered.
 

EinarA

Feb 11, 2013
28
Joined
Feb 11, 2013
Messages
28
You don't seem very willing to be " shown to be wrong " if you have rejected everything I have said so far. It should be obvious to anyone that the beta dropped from 60 to 20. Why are you claiming that this is only simulation data when everyone else has said it can't be simulated. I'm sorry I brought this up, I don't care to have my knowledge subjected to the ignorance of others. And no, you don't have to worry about me calling you 'sir'.
 
Top