BAV99 low-voltage IV curve

[matthuszagh]

I have an application for which I wanted to know the IV curve of the BAV99 at low voltages. I measured a single diode from the pair in the forward direction on my HP 4145A for 200 values between 0.001 and 0.2 V. I think these results may also be useful to others, so I'm posting the plot. The blue curve is just the calculated static resistance, i.e., V/I.

[TimFox]

Another curve that might be more useful for some applications is the "dynamic resistance" dV/dI.

[David Hess]

I would like to see a small signal transistor like the 2N3904 emitter-base and collector-base junction also for comparison.

[mawyatt]

Nice!!

Maybe extend the voltage range to 0.5 or 0.6V forward biased, and also the reverse biased junction capacitance would be useful to "see" if the OP has the time.

Best,

[TimFox]

For some applications, it is useful to see the dynamic resistance as the voltage goes through the origin (from slightly positive to slightly negative).
Near zero bias, the junction capacitance is maximum, and the dynamic resistance is finite.

[Kleinstein]

The fast diodes are usually rather high in leakage and also tend to have a relatively low differential resistance. The small signal transitor (unless a gold doped one) tend to be much lower leakage and higher resistance.

I think the resistor should not go down again for the very low votlages. The expected form is more like increasing resistance down to a bit in the reverse direction and than more like constant for a while.
So the drop in the curve could be offset or maybe extra leakage due to light.

[matthuszagh]

Here are some more measurements to show more of the range. I've kept the ranges separate, because it makes it easier to read off values from the plot. Note that not all scales are the same. In the reverse plot, I swept in the reverse direction from 0.1 to -0.1 V.

[matthuszagh]

more plots

[matthuszagh]

all raw data (except reverse sweep).

[matthuszagh]

I'll keep this setup for a day or two, in case people want me to try anything else. I'll try to get around to measuring a 2N3904 for comparison.

I also have a Keithley 617 electrometer, in case people want to see a comparison with that. I use that with a much shorter triaxial cable, but the same HP 16058A test fixture.

[dmendesf]

 Could you please plot IV graphics for an IR led? They seem very steep...

[matthuszagh]

Could you please plot IV graphics for an IR led? They seem very steep...

Are you saying the BAV99 IV curve appears very steep, or an IR LED? I don't think I have an IR LED on hand, but if you send me one I can measure it and share the results.

[David Hess]

Here are the results of various diode curves that Robert Pease measured.

The slope of the curve is "conductance" with high conductance diodes having a steeper slope.  Common transistor base-emitter junctions have an ideal 60 millivolts per decade of current but most silicon diodes are inferior because they are optimized for other things, although even the low leakage diode he tested was not as good as a transistor base-emitter junction.  Schottky diodes come in different types with different values of conductance; HP had different categories for low, medium, and high conductance.

He also mentions that red LEDs can have very low forward currents, like 1 picoamp at 0.6 volts forward voltage which is useful in some circuits.