Zener Diode Current Test

[UltimateX]

Hi,

I have a 3V zener diode for which the IRmax and VR are 50uA and 1V respectively.   The zener measures 1.8V reverse biased with a DMM (diode test mode), which appears to make it look like it may be leaky.  Before replacing the zener, it has been suggested to me to test the reverse current of the zener when biased to 1V and then check the zener voltage when running at 10mA to further determine if the zener is actually faulty.  How can I check the reverse current at 1V if the breakdown voltage is 3V?

To test the zener at 10mA, is it a given that this would be in series with, say, a 1k resistor?  If so, I am not quite sure how to make take this measurement without having an ammeter in series.  How should I do this?

Thanks!

[bdunham7]

It is entirely possible that your DMM would only show 1.8V on such a zener, especially if its test current is less than the typical 1mA.  Low voltage zeners have a very soft "knee" and lot of leakage at voltages well below their nominal value. Post the part number of the zener and the model of your DMM.

As for testing, do you have a variable power supply and a DMM with a 100uA range?

[UltimateX]

Thank you for your reply.  The diode is a UZ-3.0B, for which I have attached the datasheet.  My DMM is a Vichy VC99, but I get 1.8V on a White Gold WG020 Precision Series DMM as well and a cheap Machtech DMM also shows 1.8V reverse biased.

I have a variable power supply and, yes, the Vichy has a uA range.

[UltimateX]

I just checked the user manual and my Vichy VC99 has a 600uA range.

[Tomorokoshi]

There are a few details to work through and clarify that hopefully will make it easier to test and evaluate this part.

1. As implied by bdunham7, The DMM diode test mode is meant for the forward direction of a diode, not the Zener (reverse) direction. If used in the reverse direction, it will reach either a current limit or a voltage limit of the test circuit and give some arbitrary result.

2. As the Zener is typically used in the reverse direction, I_Rmax defines "The current that flows from the external circuit into the cathode terminal at a specified reverse voltage (V_R) below the onset of breakdown." What is going on here is that before the Zener breaks down in the reverse direction, there is some leakage current at voltages below the Zener voltage. This could be important for some circuit where this leakage would need to be accounted for. In your case, if it were tested out of circuit, it could be used to indicate if the diode were shorted. From the datasheet this is I_Rmax = 50uA at V_R = 1V.

3. The power dissipation is 500mW, so for the 3.0V part the maximum current would be 166mA unless otherwise specified or limited by thermal performance. Those columns appear to be empty. Therefore, you could test anywhere from perhaps I_Z = 10mA to 100mA and check the reverse voltage. If it works that would be between 2.8V and 3.2V. Perhaps test with the supply set to 6V with a 100 ohm resistor in series with the diode. You should expect around 30mA. This would then be 3V across the resistor, so you can check the current without an ammeter.

[bdunham7]

The diode is a UZ-3.0B, for which I have attached the datasheet.  My DMM is a Vichy VC99...

I have a variable power supply and, yes, the Vichy has a uA range.

The VC99 has a diode test current of less than 500µA and a maximum (by spec) diode voltage of 1.5V, so your observed results are not unexpected.

If you have a 1K resistor, you can set up the PSU, the zener, the 1K resistor and the VC99 in the µA range all in series.  Set the PSU for ~1.1V and observe if the current is less than 50µA.  If so, that's good.

Next, leaving everything connected in series, set the VC99 for a range appropriate for ~10mA--probably a 60mA range if it is a 6000 count meter--and set the PSU for ~13V.  Adjust as needed to get a current of about 10mA and then measure the voltage across the zener with one of your other meters.

Once you've verified that your zener is OK, you can characterize it by putting various test currents through it and recording the voltage across it as measured by a second meter.  You might be a bit surprised by how much current is already flowing before you get near 3.0V.  That's what is meant by a 'soft knee'--the zener action doesn't occur at a clearly defined voltage, but rather fades in as the voltage goes up.

[Kleinstein]

3 V and similar low voltage zener diodes are far from ideal with a rather smooth onset of the current flow. So there is already quite some current flow already from 1-2 V on, not just starting just before 3 V.
A blue/green LED may have a steeper curve.

The diode test is just 1 point of the curve with a not so well defined current. It would need a test with the external supply and series resistors to get more points. Because of the often relative high burden voltage with DMMs the suitable way can be using the DMM to measure the voltage over the diode and the voltage over the series resistor to get the current. Except for really low currents one should get away with 1 meter, reading one voltage at a time. For the low current one may have to correct for the input resistance.  With a few different series resistor values (e.g. 1 K , 10 K, 100 K) one can cover a relatively large current range. The supply voltage can be used for more points in between.

[UltimateX]

Hello all,

Thank you all very much for your replies and advice - greatly appreciated and a useful learning experience.

Here is what I found carrying out the practical tests suggested:

Connecting in series the VC99 DMM, a 1k resistor and the 3V Zener and applying exactly 1.1V to the circuit, I see 10.1uA on the meter.  When I drop the voltage to 1V, I read 5.4uA.

Leaving everything in place, setting the VC99 to 60mA range and applying 13V to the circuit, I see 10.14mA on the meter.  At 12.84V I am able to achieve exactly 10mA.  Measuring the voltage across the Zener, it reads 2.84V.

With 100 Ohm resistor replacing the 1k resistor and an input voltage of 6V, I get 27.6mA on the meter.  I measured 2.74V across the resistor. 

So, given the test results above, I presume it is safe to conclude the UZ-3.0B I have is in fact functioning normally?  Or, do you think I should I carry out further tests with different resistor values and currents to create a graph of the performance curve to be more sure of this?

The advice you see on Youtube by various video publishers is very binary when it comes to testing Zener diodes with a DMM, i.e. OL=working, any voltage observed = faulty.  I think the publishers of such videos really ought to provide at least a caveat for low voltage Zeners; otherwise, it can be very misleading. 

Thanks!

[bdunham7]

So, given the test results above, I presume it is safe to conclude the UZ-3.0B I have is in fact functioning normally?  Or, do you think I should I carry out further tests with different resistor values and currents to create a graph of the performance curve to be more sure of this?

You didn't do a reading across the zener in the higher-current test and that might be useful information.  However, the basic result here is that the zener is a bit on the low side but within the specifications given on the datasheet.  In addition, there is no test temperature specified in the datasheet and your zener will have a negative TC (voltage goes down as T_j goes up).  If you assume that the specs are for T_j = 25C, your zener probably was a little warmer than that with 10mA through it.  So I'd say it is just fine.

If you are having issues with a circuit, it may be useful to plot a curve of the zener voltage (measured by DMM #2) vs current (measured by DMM #1) over a few values--say 1mA, 2mA, 5mA, 10mA, 20mA, 50mA--or something like that.  This may help you understand how it behaves in real life and thus how to use it appropriately in a circuit to acheive a particular goal.  Using a 3.0V zener is particularly tricky.  How is it used in your circuit?

[UltimateX]

Hi bdunham7,

Oh, okay.  I was following Tomorokoshi's instructions and just assumed it would be a matter of 2.84V (the resistor voltage) subtracted from 6V to obtain the Zener voltage. Perhaps I have misunderstood something - apologies.  In any case, I checked it with the 100 Ohm resistor installed in the test circuit and the Zener voltage measurement is 3.21V.  Was it a different test that I should have measured the Zener voltage on?

I will plot a current curve as you suggest and feed back the results - thank you.

The UZ-3.0B Zener is out of a Kenwood TS440S HF transceiver (Service Manual: https://www.oz1bxm.dk/manuals/TS-440/TS-440S_service_manual.pdf, RF Unit, P. 78, D56 at lower right-hand corner) on which I am trying to trace a fault with the 455kHz IF.  The frequency I get at VCO4, connector 3, pin 2 ("3-2") on the IF board is around 405kHz and barely increases or decreases when VR7 is rotated on the "SW "A"" board (about +/- ~2kHz).  The service manual PLL Adjustment instructions (P. 103, step 4, condition 4) instructs you to set the frequency to 456.875kHz, so clearly something is not right.  A connection into J14 on the IF board comes from RF board J11 to which the UZ-3.0B Zener is connected, so I was just trying to verify this diode was working well. 

Thanks for your help with this.

[UltimateX]

Hi,

Here is what I got maintaining 6V with five different resistor values:

R_s        V_R             I_R 
10R     3.76V      220mA (momentarily)
100R   3.21V      27.7mA
1k       2.47V      3.52mA
10k     1.84V      418µA
100k   1.36V      46.2µA

I measured the voltage of the DMM#2's diode test mode and it is 3V.  I then used the VC99 in ammeter mode (uA) in series with the 3V Zener (reverse biased) and the DMM#2 in diode mode and the VC99 reads 357uA while DMM#2 shows 1.83V, so that seems to fit in with the results above.

[UltimateX]

And here is the Zener current vs voltage curve data:

1k resistor R_s in series with 3V Zener and VC99 (uA):

I_R            V_R
1mA        2.07V
2mA        2.28V
5mA        2.58V
10mA      2.83V
20mA      3.09V
50mA      3.40V
100mA    3.60V
150mA    3.66V

[armandine2]

look ok in excel

[UltimateX]

Thank you armandine2.  Yes it looks pretty good and with the soft knee bdunham7 described.