Low voltage zeners are poor....zener voltage has too wide tolerance

[treez]

Hi,
Are there any 3v3 zener diodes that pull less current at voltages lower than 3.3v than A BZX84-C3V3?
BZX84-C3V3 datasheet
https://www.diodes.com/assets/Datasheets/ds18001.pdf

It seems all low voltage zeners are equally poor?

We have a divider from a high voltage rail (up to 150v) feeding into an opamp which has a vdd of 3v3.
[RHI=136000, RLO=3300]

 

[KlausST]

Hi,

there are a lot of discussions about using zener for protection.

As said in the other threads: I´m no friend of it.
* The current drawn at lower voltage decreases measuremtn precision a lot.
* with some "higher" error current the voltage goaes alot above 3V3.
* when power is switched OFF they dont protect the OPAMP input at all.

Thus I usually use two diodes, one to GND and one to VCC. An optional series resistor to the OPAMP input additionally limits the current.

Klaus

 

[betwixt]

Assuming you are using the diode for protection, I concur with Klaus, in fact if you can live with their slightly higher leakage current, use Schottky diodes to GND and VCC. Their lower Vf helps to keep voltages closer to the supply rails than a normal diode.

Brian.

 

[asdf44]

Yeah diodes are a good clamp solution.

If you really need zener functionality a TL431 or variant does the job accurately down to about 1.2V.

 

[dick_freebird]

Another option might be to multi-tap the divider, and
use a higher-current shunt reference (some app notes
may be found on boosting) at a "5V" tap point just
above the 3.3V-max tap; there should be no way for
the 3.3V not to follow the "5V" clamp limit.

 

[asdf44]

You’re sometimes ok in such a case simply because your large divider inherently limits the current (I’m guessing something like 1ma) and if the manufacturer is nice they’ll have an input current spec on the input internal clamp diodes.

Also I’ll often use 2 resistors in parallel on the low side to mitigate the impact of one of them going open.

 

[dick_freebird]

Often the input current spec is low, like <10mA and is set
to prevent pin induced electrical latchup. Manufacturers
never will tell you whether latchup will happen at 11mA
or 100. I would not recommend using pin protection clamps
this way (unless you have data to say that you have a
whole lot of margin, at high temp).

 

[KlausST]

Hi,

The darsheet will tell the current limit.

The protection should not be used as standard operation. But it's O.K. to use them as protection.
I don't think a voltage divider for 150V takes more than 10mA, because this means 1.5W of dissipated power.

Klaus