I've been trying to understand the proper usage of Zener diodes, but I've not been able to get the behaviour I expect based on my understanding.

For example, in the test circuit below I would expect the voltage across the Zener to be approximately 3.3 V (the datasheet indicates the 1N4728A has a Zener voltage of 3.3 V at a test current of 76 mA), yet I measure it as 4.05 V. I've tried multiple replacement diodes to try to rule out a faulty part, but I get a similar result for each, so clearly there's something faulty in my understanding. Can someone explain the behaviour I'm seeing?


simulate this circuit – Schematic created using CircuitLab

For reference, here are photos of the diode in question, and the kit it came in:

enter image description here enter image description here

  • \$\begingroup\$ Try to reduce current, datasheet specifications shows voltage tolerance and test current, at which you should get expected voltage with tolerance band. Zener diodes have differential resistance (slope of characteristic). \$\endgroup\$ Commented Jan 1, 2023 at 8:43
  • 2
    \$\begingroup\$ How are you determining the actual current? \$\endgroup\$ Commented Jan 1, 2023 at 11:47
  • 1
    \$\begingroup\$ I believe the 1N4728A is supposed to have a voltage tolerance of 5% at the test current if I've understood correctly (so anywhere between 3.135v & 3.465v would be acceptable). I'm calculating current based on the voltage across the resistor divided by its resistance, which in the above example will indeed give 43mA. If the voltage across the diode were 3.3v as expected though, the current would be 77mA which is almost exactly the test current quoted by the datasheet. I've tried using a 10 Ohm resistor to increase the current closer to 76mA, this gave 4.17v across the diode and 83mA current. \$\endgroup\$
    – Mark Lucas
    Commented Jan 1, 2023 at 13:11
  • 2
    \$\begingroup\$ The actual voltage across the zener diode should be the rated voltage +/- the rated tolerance at the Iz test current. The dynamic resistance only comes into play when the current is different from test current. Either your resistor is something like 2.2Ω (in which case the diode would be getting very hot indeed), your meter is wonky (dying battery?- measuring the 5V on the same range would eliminate this possibility for practical purposes), or the diodes are not the part number you think. \$\endgroup\$ Commented Jan 1, 2023 at 13:48
  • 1
    \$\begingroup\$ Do you trust Bojack of China? Is he a dumpster diver who gets and sells manufacturer's rejects? Buy Name Brand parts instead. \$\endgroup\$
    – Audioguru
    Commented Jan 2, 2023 at 16:02

1 Answer 1


Behavior found with simulator microcap v12.
If you measure something as 4 V, then it is not a 1N4728A.

NB: Note that if this diode has been "used" and "mistreated" for some reason, some characteristics have been "changed".

Simulation seems ok, current too.

enter image description here

  • \$\begingroup\$ I get the same result in an LT Spice and CircuitLab simulation too. So, I'm not sure if the components I have are faulty or the simulation model is flawed \$\endgroup\$
    – Mark Lucas
    Commented Jan 1, 2023 at 16:38
  • \$\begingroup\$ Marking error probably. \$\endgroup\$
    – Antonio51
    Commented Jan 1, 2023 at 20:09
  • \$\begingroup\$ All Name Brand zener diodes I have used are rated at 5mA and had accurate voltages. Low voltage zener diodes have poor voltage regulation then of course the voltage is much higher when the current is massive. \$\endgroup\$
    – Audioguru
    Commented Jan 13, 2023 at 16:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.