Zener diode reverse leakage current is a subject of contradiction in its datasheet

[opampsmoker]

Actually i notice that the UDZV8.2B (8V2 Zener) is said in its datasheet page 2 to have a Reverse Leakage current of 500nA at VR=5V and 25degC. However, its datasheet on page 4 states that it has a Reverse Leakage current of <10 picoAmps at VR=5V and temp = 25degC.

Do you know which one is correct?

UDZV8.2 zener
https://fscdn.rohm.com/en/products/databook/datasheet/discrete/diode/zener/udzvte-178.2b-e.pdf

Also,
Please can you state what is the overall leakage current through these attached back to back zeners?
One is a 5V1 zener, and the other an 8V2 zener.


UDZV8.2 (8.2V Zener)
https://fscdn.rohm.com/en/products/databook/datasheet/discrete/diode/zener/udzvte-178.2b-e.pdf

UDZV5.1 (5.1V Zener)
https://datasheetspdf.com/pdf-file/1261604/ROHM/UDZV5.1B/1

(BTW Page 4 and page 2 of their datasheets contradict each other over leakage current)

 

[OBW0549]

Do you know which one is correct?

They're both correct. The spec on page 2 gives you the maximum reverse current, while the graph on page 4 gives you the typical value.

 

[opampsmoker]

Also, do you think the “modified back to back zeners” circuit, (also attached) with a MMBD4148 diode in series with each Zener, will reduce the amount of leakage current whilst having the same “back to back” Zener effect?

MMBD4148 datasheet

https://www.diodes.com/assets/Datasheets/ds12003.pdf

 

[opampsmoker]

They're both correct. The spec on page 2 gives you the maximum reverse current, while the graph on page 4 gives you the typical value.

Thanks, normally i would agree, but we're talking several orders of magnitude difference here.

 

[OBW0549]

Thanks, normally i would agree, but we're talking several orders of magnitude difference here.

Several orders of magnitude difference between typical and maximum values is exactly what you'd expect when it comes to reverse leakage current.

One more time, speaking from almost a half-century of engineering experience: THERE IS NO CONTRADICTION HERE.

 

[opampsmoker]

OK Thanks, much appreciated.
Do you think that the attached way of modifying the back-to-back zener circuit results in lower leakage current?

 

[Diver300]

Thanks, normally i would agree, but we're talking several orders of magnitude difference here.

You will find this with many secondary characteristics of components, especially leakage or standby currents.

With a zener, most users don't care about the forward leakage, so there is no commercial advantage in specifying a low leakage current, and the disadvantage of specifying a low value is that testing is more difficult and there could be more rejects. Also a change in process might make it worse which for nearly all customers is unimportant, but if the leakage is worse, the specification and part number would have to change, so it would be a huge amount of work.

 

[Diver300]

OK Thanks, much appreciated.
Do you think that the attached way of modifying the back-to-back zener circuit results in lower leakage current?

No, they don't. All the circuits end up with the supply voltage across the zener, so the leakage current of the zenner will be taken. The more complicated circuits have other paths which may leak, so they can only add to the current taken.

As you have noticed, the typical leakage is tiny.

Moisture could increase the leakage a lot.

 

[opampsmoker]

Moisture could increase the leakage a lot.

Thanks, i tend to agree, and up to about 2uA or so for a 5mm square damp bit on the PCB?

Also, i hear what you say about leakage current in the UDZVTE-178.2B zener, but its quoted on page 2 as having 500nA of leakage at 25degc...then on page 4 the graph shows even at 125degc, it only has leakage current of less than 1nA......which sounds unbelievable, because leakage current typically increases dramatically with temperature in most components. Surely?
Sub 1nA of leakage current in any semiconductor at 125degc surely sounds unbelievable?

 

[Diver300]

Sub 1nA of leakage current in any semiconductor at 125degc surely sounds unbelievable?

No. The leakage goes up a lot at 125 °C, but that is from a very low base.

Semiconductor leakage can be extremely low. If you take a CMOS gate, an input pin will just sit at a voltage when left with no connection.

An EEPROM has conductive areas that are charged up during programming and just sit there. You can now get a 1 TB micro-SD card. I struggle to imagine how small each charged volume is for each of the eight trillion bits in an object the size of a fingernail. Each of those charged volumes will keep their charge for years, so the leakage current must be vanishingly small.

 

[opampsmoker]

woops sorry, here is the corrected cct of post #6

 

[opampsmoker]

An EEPROM has conductive areas that are charged up during programming and just sit there. You can now get a 1 TB micro-SD card. I struggle to imagine how small each charged volume is for each of the eight trillion bits in an object the size of a fingernail. Each of those charged volumes will keep their charge for years, so the leakage current must be vanishingly small.

Thanks, i agree, but those memory elements are specially doped so that they have extremely low leakage current, whereas zeners arent made with low leakage in mind.

Regarding leaky components in series, doesnt the resultant leakage go to the level of the lowest individual leakage current?

 

[crutschow]

Do you think that the attached way of modifying the back-to-back zener circuit results in lower leakage current?

I see no advantage to adding the diodes.
In either case the leakage below the breakdown voltage of the reversed-biased zener will dominate the leakage and I see no significant difference in the total leakage either way.
What is the application that you are so concerned about leakage?

 

[opampsmoker]

sorry im having a bad day..here is the correction for my post #11

 

[Diver300]

sorry im having a bad day..here is the correction for my post #11

Re-arranging things with the zeners can only make things worse. The zeners have leakage, so any circuit that is using them to detect voltage will suffer from the leakage of at least one zener. Other complications just add more current paths, which can only increase leakage.

Leakage reduction circuits work by turning off devices that can consume current. You can turn off the zeners, but they won't be active in the circuit any more.