Is it bad if base voltage exceeds collector voltage?

I was experimenting with this circuit for inverting RS232 signals:

Transistor: BC546. Diode: 1N4148.

It worked fine when feeding in ±5V, but testing with ±10V it stopped working and the transistor was destroyed.

I suspect either:

  • It is bad if the base voltage exceeds the collector voltage
  • I exceeded the emitter-base voltage (the maximum being 6V)

What do you think?

If I am right, I thought of perhaps:

  • A 5V zener diode in parallel with the other diode to clamp the input voltage to 5V
  • Another normal diode between Base and Collector (or Base and +5V) to clamp the input voltage to the collector voltage

What would you recommend?

[quote author=Nick Gammon link=topic=190459.msg1408080#msg1408080 date=1380320510] it stopped working and the transistor was destroyed. [/quote]

Collector - base short?

I’ve used similar circuits for many years… with +10V on the input the base current is 1mA and the collector current would be about 480 - 490 uA ( - Vcesat) I = 5/10000 + .5 mA Or 1 mA X 100 {Typical median Beta value) = 100 mA Ic… means that the transistor
1mA of base current is not… I wouldn’t think excessive and the data sheet provides little information counter to my experience.
At 20 deg C the base voltage would be about .66 V @ 1mA (the BE junction is mainly a diode to ground for this measurement)
My experience is with the 2N3904 transistor… But the’547 is close enough… or I’m really losing it.
Did you try another… transistor just for laugh’s ?

Doc

BC547.pdf (146 KB)

Not cranked up to that voltage I didn't. I could, I suppose, I've got quite a few of them. :)

The datasheet says max 6v to base.

I've never looked into this but, does that mean 6v max regardless of current eg.. 10v 0.00001ma enough to kill it?

If so, Nick simply buy a 5v1 zener diode and clamp the output to base (and skip the transistor :p)

I still need to invert it.

I think I'll sacrifice another transistor to see if my theory is correct.

Oops yes sorry ... to invert it yah!

Stick a zener to cap it's base voltage, you could probably use a couple of LEDs to limit the max voltage if you don't have a zener eg (2 leds 10v-2v-2v = 6v to base)

[quote author=Runaway Pancake link=topic=190459.msg1408109#msg1408109 date=1380323634] [quote author=Nick Gammon link=topic=190459.msg1408080#msg1408080 date=1380320510] it stopped working and the transistor was destroyed. [/quote]

Collector - base short? [/quote]

Nick, I was asking how the transistor failed. Test that with an ohmmeter or, better, the DMM diode-test function. + to B and - to C should be a diode equiv. (or Low ?, but not no-?) - to B and + to C should be "OL" + to B and - to C should be a diode equiv. (or Low ?, but not no-?) - to B and + to E should be "OL"

BE junctions have a reverse breakdown voltage that is relatively low, and may be destroyed with somewhat mild currents, from what I recall. What that current is, I cannot find any information. They are sometimes used this way as noise sources, so they will withstand modest currents in breakdown. I would think with 10k on RS232, that would not be too much current at less than 1mA at -10V, but it is all I can think of.

BTW, to be specific, it is not “6V to base”, it is VEBO. Note it says the “E” first, that indicates + on the Emitter and hence reverse breakdown. The “O” (Open) indicates the Collector is disconnected.

I don’t think it is VCBO breakdown is the issue, as that is only 15V and the BC547 is rated at 50V VCBO.

If you look at that circuit it is virtually impossible to damage the transistor with any configuration with the stated voltages the base 'lead' is a max of 1 mA current available, the collector 500 uA and those are at the transistor itself to ground from 5V Vcc and + 10V input. If in fact the transistor is/was damaged then the diagram doesn't match the components...

Doc

Derp, with the diode there is no reverse breakdown.

It is possible this is actually breaking into oscillation. I don't know how you have this connected, but a few loops of wire on a protoboard with the capacitive coupling between connections, I can certainly see making an oscillator. If you have any ferrite beads, try putting some on the transistor leads. If you only have one, put it on the emitter lead. It must go directly on the transistor lead, not on a wire coming off of the protoboard somewhere else.

+ to B and - to C should be a diode equiv. (or Low ?, but not no-?) - OL
- to B and + to C should be “OL” - OL
+ to B and - to E should be a diode equiv. (or Low ?, but not no-?) - OL
- to B and + to E should be “OL” -OL

My component analyzer thinks it is a diode with C=cathode and E=anode

I tried the circuit again (new transistor) with ±20V (so max +10V) and it hasn’t failed. Also the base voltage was only 800 mV.

Docedison:
If you look at that circuit it is virtually impossible to damage the transistor with any configuration with the stated voltages …

Yes, well I was trying something else at the time it actually failed. I had a diode in series, give me a minute and I’ll knock up the circuit.

So it sounds like the transistor shorted B to E. It still could have been oscillation, as a different transistor may work fine in the same circuit.

Of course the base to ground voltage is only 800mV, the breakdown voltage is the -reverse- breakdown from E to B. The forward bias voltage will still be that of a typical silicon junction.

Yes, well I was trying something else at the time it actually failed. I had a diode in series, give me a minute and I’ll knock up the circuit.

Hrmph.

Put a radio next to it. If it is bursting into oscillation, it will probably be rather noisy.

Can we get a picture of this circuit?

Yes.
VB, with the Emitter at Gnd, = VBE (a “diode drop”.)

Now I'm not absolutely certain if that was it. Either that, or the diode around the other way, or the resistor (R2) to ground rather than in series.

Oh, my.

I meant, 20V peak-to-peak.

[quote author=Runaway Pancake link=topic=190459.msg1408268#msg1408268 date=1380336590] Oh, my. [/quote]

OK, what did I do?