# Why does this Zener diode circuit not produce the Zener voltage?

Why does the measured voltage at the op-amp input equal 2.02V? simulate this circuit – Schematic created using CircuitLab

• Probably too little current flowing through the zener for it to reach the exponential part of it's curve. Reduce R1. – DKNguyen Feb 22 at 2:29
• Not sure if this schematic is the same you are using to simulate but V1 is floating. – Wesley Lee Feb 22 at 2:36
• 1. The negative terminal of the input voltage source is floating. 2. There is no negative feedback. – The Photon Feb 22 at 2:38
• always observe Vcm input range and ensure Vin+=Vin- with negative feedback – Tony Stewart EE75 Feb 22 at 2:53
• Assuming it really is open loop as you show, the two inputs will be far apart from each other and it is very likely the (+) input is sinking significant current. – jonk Feb 22 at 3:18

## 3 Answers

Assuming you actually grounded the negative side of the +5V supply, you provided the op-amp with adequate supply voltages, and closed the loop around it to make it a voltage follower, it would still likely give you considerably less than 2.4V with only about 2mA flowing, for most US part number zener diodes.

The "knee" on low voltage (less than about 5V) zener diodes is notoriously "soft" so the zener voltage is rather dependent on current. Further, most of the older US-numbered zener diodes such as the 1N4370A are specified at rather high Izt, probably 20mA in this case. 20mA is quite a bit more than 2mA, so you can expect at 2mA the voltage will be much lower than 2.4V.

Some of the newer (relatively speaking) Japanese part numbers are specified at lower currents such as 5mA but the physics is essentially the same so they'll still have soft knees. For a solid 2.5V reference, I suggest using a TL431 or similar, which is an IC that simulates a 2.495V nominal zener diode, and only requires 1mA to be in regulation.

Here is the typical behavior of Motorola 500mW zener diodes from an old datasheet: Ideally there would be a series of vertical lines, but the low voltage types have a very soft knee (and the higher voltage ones don't like a lot of current- but they'd be burning up at those currents anyway).

• An important take away on Zener is that old ones needed more current (20mA) to achieve 100 Ohms @ 2.4V and the newer ones can do that with 5mA. Meanwhile 3V white LEDs in 5mm are 15 Ohms at 20mA as the epitaxial quality has improved the Rs on both over the years. So it is important to remember the Vzt if 5% tolerance is nominal only at the test current for Vzt. In both LEDs and Zener 5% tolerance is easily achieved at 5 mA while 20 mA has a much wider tolerance due to Rs and not the threshold voltage which ON semi calls or called at one time Vzk for knee voltage @ 250uA – Tony Stewart EE75 Feb 22 at 9:06
• Thanks for making all the right assumptions. I will try your suggestion. – Koot33 Feb 22 at 16:39

This works better simulate this circuit – Schematic created using CircuitLab

This OpAmp configuration produces good results. simulate this circuit – Schematic created using CircuitLab