Odd use of NPN Epitaxial Transistor

[Govmnt_Lacky]

Just saw a layout where an NPN Epitaxial Transistor was used in the following manner:

- Collector AND Emitter tied together and connected to the Anode side of a 1N4148
- The Cathode side of the 1N4148 is tied to the Base and to GND.
-The Emitter/Collector goes to several other connections within the circuit.

Never seen this done before. Any ideas?? This is on the Lovetone Cheese Source PCB.

[Digital Larry]

Well it's kinda like a diode with C and E stuck together (a LOT like a diode, in fact), so it sounds like someone wanted to do the back to back diode thing.  For some reason that transistor presented itself as the component of choice rather than a different diode.  I'd be guessing that they wanted the voltages to be different on + and - clipping to influence the relative even and odd harmonic content of the overdrive.

[R.G.]

Yep.

One interesting (if only trivially related!) item with transistors is that tying the base to the collector makes the collector/base to emitter diode follow the "diode law" equations better than most simple diode junctions do. And you can use two resistors and a transistor to make an "amplified diode" which has a larger diode "forward voltage" than the normal diode, if somewhat polluted by the conductance of the parallel resistors that do this.

[bool]

looks like a "mojo" clipper to me (bjt-diode + a "real" diode)

...don't have a clue about this particular pedal btw...

[merlinb]

IIRC connecting the collector and emitter together makes it behave as a zener diode, doesn't it? In this case that function is not used, since the 1N418 is in parallel with it, but maybe it was used that way in an early prototype?

[PRR]

> IIRC connecting the collector and emitter together makes it behave as a zener diode, doesn't it?

The E-B junction is a fine Zener. In some devices there is a residual tempco which can be reduced by throwing the C-B junction in series forward-biased.

I _do_ wish Lacky had used a crayon. Text descriptions are mind-puzzles.

IF I have worked the puzzle correctly (and he posed it correctly), then I get this:



While C-B has a ~~~40V breakdown, and E-B is usually 7V breakdown, with this 1N4148 in the direction I *think* it is, both are clamped to 0.6V.

So it is two back-to-back diodes, one marked "diode", one marked "transistor", and is ~~0.6V either way.

Commercial diodes are not perfect junctions. There will be some difference in curve between diode and transistor. I have doubt that this could be audible.

Another factor is that from time to time transistors (especially "junk" parts) come on the market cheaper than good diodes. In this case, the transistor could be stupidly low gain or even half-blown (one junction open) and still clip signal. Then the question is: why not two junk transistors instead of a several-penny 1N4148? Some things we will never know.

[Govmnt_Lacky]

@Paul

Your drawing (although not crayon... still artful!) looks correct to what I see in the circuit. Thanks a lot for the explanation!