# Why are there two diodes in series in this test circuit?

Look at Figure 2 in the following datasheet: http://www.onsemi.com/pub/Collateral/MMBT3904LT1-D.PDF

Why are there two diodes and not just one?

Thanks.

• I think that 2 diodes gives a higher threshold than 1, eg: 2 100V diodes can barrier just under 200V. I've got nothing to back that up, though. Commented Jun 22, 2017 at 14:06

The circuit includes two diodes to set the negative bias at the base during the turn-off phase of the measurement.

A single diode would only give about 0.7v reverse, two give twice that.

If they didn't have any diodes at all the voltage would exceed the reverse breakdown of the base-emitter junction that is in the region of 7v with silicon planar transistors.

I agree that it is somewhat unusual way to do the test rather than defining the voltage from the pulse generator.

• so it's a 1.4V regulator of sorts? Commented Jun 22, 2017 at 15:05
• @tuskiomi, yes that's right, you could make a shunt regulator this way and it's sometimes done. Regulators are for continuous power supply though. This is a simple clipping function. Commented Jun 22, 2017 at 18:09
• @TonyM isn't this wildy unstable with temperature? Commented Jun 22, 2017 at 18:18
• @tuskiomi, no it varies with temperature and slightly, not wildly. It's always stable. Let's not start a semantics debate in lots of comments, though. You can look up the temperature variance in the datasheet. Commented Jun 22, 2017 at 18:30
• @TonyM i'm just a bit rusty, is all. not looking to debate. Commented Jun 22, 2017 at 18:32

I would say the schematic is incorrect. They should be pointing "down". If they were pointing down, they would give a combined voltage drop (taking the average) of 0.63 *2 = 1.26, putting the voltage in the middle of the active region of the transistor.

The diodes in such configuration are used for temperature compensation when such circuit gets to work in hot/cold environments so that the target output remains temperature invariant.