I would expect a forward biased regular diode to be used. Especially in applications in which you don't expect a change in current direction. Is the curve more predictable in the reverse biased state?
\$\begingroup\$ You mean why don't we use a series of forward biased diodes, instead of reverse biased ones? Because with a single diode you will be stuck with 0.2V, 0.7V and the LED forward values which goes from 2 to 3 volts. \$\endgroup\$– Sredni VashtarJul 20, 2020 at 1:42
The curve is much more useful in the reverse breakdown mode. Especially for avalanche diodes, which every "Zener" diode over above 5V actually is.
simulate this circuit – Schematic created using CircuitLab
Here I have swept the current from 100uA to 100mA on each type of "regulator" diode.
Compare the regulation of each:
Temperature coefficient is also ~4x better, about +6.5mV/°C vs. roughly -25mV/°C
A forward-biased silicon diode (zener or "normal") will have a voltage drop of about 0.7 volts - not too useful for voltage regulation.
Zener diodes can be made with many diffrerent reverse breakdown voltages, so you can use them in suitable circuits to produce many different regulated voltages.
\$\begingroup\$ Forward diodes can produce many voltages as well. For example seven 0.7V diodes will give 4.9V regulation. And yet a 5.1V zener would be better (IIRC, the best you can get). I believe this is what the OP is asking. \$\endgroup\$ Jul 20, 2020 at 1:57
\$\begingroup\$ 0.65 V and 1.35 V diode drops are common in current mirrors and constant-current sources and sinks. Little power is wasted at such low voltages. \$\endgroup\$– user105652Jul 20, 2020 at 6:38
The reason we use zener diodes in voltage regulators is because zener diodes exhibit almost identical behavior to a forward biased diode above the zener breakdown voltage.
It is like having a diode but with a forward voltage of any value you want.
\$\begingroup\$ "Any value you want" implies an adjustable regulator like a TL431, which can have infinite values between upper and lower limits. \$\endgroup\$– user105652Jul 20, 2020 at 6:40
\$\begingroup\$ I didn't mean it literally. \$\endgroup\$ Jul 20, 2020 at 7:41
\$\begingroup\$ I understood what you meant. \$\endgroup\$– Kyle BJul 20, 2020 at 11:04
2\$\begingroup\$ It's pretty far from identical, in fact, as Spehro's answer elaborates on. The zener diode has a much flatter voltage-current curve. \$\endgroup\$– HearthJul 20, 2020 at 16:00