After measuring the forward voltage drop over a diode, I should also have measured the reverse leakage current, i.e. how much current the diode lets through when it’s supposed to be blocking. I never did until now, because I couldn’t detect any current in a quick check I did a while back. Time to build a better setup – here’s what I used:
The voltmeter’s own 10 MΩ or so internal resistance will skew the readings by 10%, but that’s no big deal.
It turns out that the reverse leakage current is pretty small when applying 5V:
- 1N4004 – a high power diode: 1.3 mV = 1.3 nA
- 1N4148 – a low power diode: 3.4 mV = 3.4 nA
- BAT34 – a Schottky diode: 50 mV = 50 nA
That’s nanoamps, i.e. milli-milli-milli-amps. The Schottky diode does indeed leak a tad more than the others. Here are the specs of that BAT34 diode – note that the reverse current could even be used as temperature sensor!
FWIW, I found a minuscule “RB751S” SMD Schottky diode, about 1 mm long, which does a bit better at 7.0 nA:
It was quite a challenge to get some wires soldered onto it. I used the core of 30 AWG Kynar “wirewrap” wire:
Anyway – the BAT34 is good enough: 50 nA leakage is acceptable while dealing with circuits which consume µA’s.