Zener diode problem

Thread Starter

surfdabbler

Joined Oct 7, 2013
7
I am having issues with Zener diodes. First time, I assumed I must have a dud zener, but I have bought two more and they are doing the same thing. I'm trying to get a reference voltage for a voltage comparitor circuit, but the zener is not giving the correct reference voltage. I've take the zener out of circuit to figure out what's going on...

My test circuit is pretty simple - V+ (11.5V), connected through resistor, then 3.3v Zener (reversed, with black line toward V+), then to gnd. At first, I was using a high resistor of 1Mohm, to get the reference voltage without drawing much current, but the zener voltage was very low, like 0.9V. I wondered whether the zener needed more current to achieve the required voltage, so I tried decreasing the resistor, and tried two different battery supplies, and got a big range of results...

R zener v zener v
(7.7V batt) (12.2V batt)
1.2M 0.93v 1.01v
400K 0.93v 1.01v
10k 1.7v 1.8v
150 3.07v
50 3.5v

So, I had to drop the resistor right down to get close to the right voltage, and at 50ohms, the zener voltage exceeded the spec voltage of 3.3v!

Out of curiosity, I reversed the zener to be fwd biased, and repeated some of the tests, and the voltage varied between 5.5 to 6.5 through the same range, so fwd biasing seems to give a better result, as reasonably expected, but the reverse biased zener should be giving a more consistent 3.3V, right?

The batteries were freshly charged, and I measured the battery voltages in circuit, so that should not be an issue. I have tested diodes from two different stores, and they are two different wattages (400mW, and 1W), and they are giving consistent results.

What is going on?
 

crutschow

Joined Mar 14, 2008
28,189
I assume you haven't looked at the data sheet.
Look there anf note the current used to specify the voltage rating of the zener and the dynamic resistance of the zener.
Low voltage zeners have a high dynamic resistance and thus poor voltage regulation as the voltage varies significantly with current (called a soft "knee").

Don't understand your results when forward biasing the zener.
For that, its voltage should be comparable to a standard diode or about 0.7V.

If you want a voltage reference that will operate at a low current (1ma) with a sharp knee, then use an IC reference, such as the 2.5V, TL431 (it can also be adjusted from 2.5V up with two resistors).
For really low current you could use a 2.5V, LT1634 which can work well down to 10μA.
 
Last edited:

Thread Starter

surfdabbler

Joined Oct 7, 2013
7
Thanks for the update. Looking at my results (apologies for the formatting - it didn't come through, and I'm trying to edit that at the moment), I need to be looking at about 100ohms to get the correct voltage. at 7.7-3.3 = 4.4V, 100 ohms gives me 44mA, so should I assume my diode needs 44mA to operate correctly?
 

Thread Starter

surfdabbler

Joined Oct 7, 2013
7
Hmm, I give up on trying to format my data!

Thanks for the info. I had assumed that zeners were a lot better at their job, but I understand the 'soft knee' now, and I will adjust my circuit to aim for 20mA current through the zener. At this level it looks like the zener might be closer to around 3V, but at least the data makes sense now! Thanks for your help.
 

crutschow

Joined Mar 14, 2008
28,189
Higher voltage zeners, such as 5v-10V units have much sharper knees, with the 6.8V unit having the lowest dynamic resistance (sharpest knee).
 

Thread Starter

surfdabbler

Joined Oct 7, 2013
7
I adjusted the resistor to provide the correct current to the zener, and the whole circuit is now working perfectly! Thanks for your help!
 

hp1729

Joined Nov 23, 2015
2,304
I am having issues with Zener diodes. First time, I assumed I must have a dud zener, but I have bought two more and they are doing the same thing. I'm trying to get a reference voltage for a voltage comparitor circuit, but the zener is not giving the correct reference voltage. I've take the zener out of circuit to figure out what's going on...

My test circuit is pretty simple - V+ (11.5V), connected through resistor, then 3.3v Zener (reversed, with black line toward V+), then to gnd. At first, I was using a high resistor of 1Mohm, to get the reference voltage without drawing much current, but the zener voltage was very low, like 0.9V. I wondered whether the zener needed more current to achieve the required voltage, so I tried decreasing the resistor, and tried two different battery supplies, and got a big range of results...

R zener v zener v
(7.7V batt) (12.2V batt)
1.2M 0.93v 1.01v
400K 0.93v 1.01v
10k 1.7v 1.8v
150 3.07v
50 3.5v

So, I had to drop the resistor right down to get close to the right voltage, and at 50ohms, the zener voltage exceeded the spec voltage of 3.3v!

Out of curiosity, I reversed the zener to be fwd biased, and repeated some of the tests, and the voltage varied between 5.5 to 6.5 through the same range, so fwd biasing seems to give a better result, as reasonably expected, but the reverse biased zener should be giving a more consistent 3.3V, right?

The batteries were freshly charged, and I measured the battery voltages in circuit, so that should not be an issue. I have tested diodes from two different stores, and they are two different wattages (400mW, and 1W), and they are giving consistent results.

What is going on?
Congrats on doing a successful exercise on Zener diodes. Yes the voltage across a Zener varies with the current through it. They are rated at a certain voltage at a given current. You can tweak the current to get a precise voltage. Using a constant current source instead of just a resistor gets you a pretty stable voltage.
 

Bordodynov

Joined May 20, 2015
2,938
I took a resistor value of 1 meg as the author. Maybe his battery-powered and the currents of the order of one milliampere and longer is no suitable. There are available comparators and operational amplifiers with micro-ampere currents.
REF1112.png
 

Thread Starter

surfdabbler

Joined Oct 7, 2013
7
Thanks for all the updates. If I had to do the circuit over, I would certainly look into these improved voltage references, but I can live with the higher current for this particular application.
 
Top