The Zener diode is often used to create a reference voltage. In tutorials and even college texts, there are mentions of creating a Zener diode based regulator. The idea is that the Zener maintains a known voltage drop. The problem is that current matters. This post looks a quick Zener diode overview and shows what happened when I tried to power a microcontroller using a “Zener diode regulator.”
Zener diodes overview
Just a brief overview if you are not familiar with Zener diodes. Like typical diodes, Zeners have low forward voltages. Typically you voltages are around 0.7. However, different material sets can offer different forward voltages.
Also like generic diodes, there is a reverse breakdown voltage. If you look at a hefty diode like the 1n4001, you find breakdown voltages starting at 50 volts.
Zener diodes are unique because their reverse breakdown voltages are relatively low. For example, I have some that are 3.3, 5.0, 9.1, and 12 volts. (Interesting numbers, aren’t they?)
The curve is showing that above the forward voltage and “below” the reverse voltage, the diode conducts. I put below in quotes because it suggests a negative potential. This comment does not mean you need a negative voltage supply, just that the diode is reverse biased. Also known as turned around.
Zener diode regulator
As mentioned, the idea behind a Zener regulator is that the diode drops a stable voltage when reverse biased. Moreover, with values like the 3.3 and 5.0 I mentioned before, it starts to sound like a good option, doesn’t it?
Let’s use the BZX79C3V3 as an example Zener diode. Notice in the characteristics table, that the reverse voltage is 3.3 volts at 5.0 milliamps.
The idea is that you pick a resistor value, perhaps even a precision value, to create enough current to reverse bias the Zener diode at 5.0 milliamps.
However, there’s a problem with this basic circuit. The current that flows through the load must also flow through the resistor. Based on Ohm’s law, the resistor’s voltage drop changes based on its current flow.
Powering ESP8266 with Zener Regulator
Using the above circuit, I tried to power an ESP8266 with a 5.0 volt supply. Before building this circuit, I measured the ESP8266’s current draw to be 60 milliamps when powered by a 3.3 volt supply.
Using a 3.3 volt Zener, the series resistor drops 1.7 volts. With 60 milliamps at the load and 5 milliamps for the Zener, Ohm’s Law tells us we need a 28Ω resistor. The closest value I have is 22 ohms.
When I hooked the circuit up, nothing happened on the ESP8266. The VOUT node measured around 0.9 volts. What’s worse is that no matter what I made the source voltage, the VOUT node stayed at 0.9 volts.
On a hunch, I dropped resistor value by about 10 ohms.
When I measured with the multimeter, I saw only 1.8 volts at the divider. However, the ESP8266 was operating. After resetting the ESP8266, I saw 2.5 volts. And depending on how much weight was on my left or right foot, any value in between.
So what the heck is happening here? Well, first, thank you for continuing to read before jumping to the comments to say the following statement. You cannot treat a microcontroller, especially a system-on-chip (SOC), as a constant load.
When I press and hold the RESET button, the Vout node jumps to a nice clean 3.4 volts. At that point, most of the active circuits in the chip are turned off.
Since the ESP8266 was a high impedance load, nearly all of the current in this circuit is flowing throw the series resistor and the ESP8266. The amount of current was staggering, almost 200 mA. Well staggering when you might have only expected about sixty.
More Zener Regulator Problems
This entire exercise was to show why the Zener diode is a lousy regulator. The voltage drop depends too much on the current flowing through the junction. Which means the “regulator circuit,” depends on a constant load. Any active device is going to cause the VOUT node to be unstable.
So what good is this Zener regulator circuit? Well, it is not a regulator. Instead, it is a reference.
For example, you could use a similar circuit on Arduino’s AREF. Let’s say you are using an analog sensor that only outputs 3 volts maximum. Using a Zener reference could give the A/D more resolution.
You could use the Zener circuit as a reference into an op-amp. This circuit is not too different from how linear regulators work.
The lesson here is that if you want to use a Zener diode regulator circuit, you need to reconsider your design. In some very rare or corner cases, it will work fine.
If you have used a Zener as a regulator, and not as a reference, leave a comment below. I would love to hear how you used it.