Zener diode below-threshold reverse current

Clear as mud ;-) Post a schematic of what you mean... 25V -> 4V ?? ...Jim Thompson

I don't understand your numbers.

Higher voltage zeners, like 8 volts and up, have very low leakage just below the rated voltage. A 1w 12 volt zener might hit 1 uA at 11.5 volts.

Low voltage zeners are awful.

Test some! Or consider using a mosfet.

Disable the regulator when the input voltage drops below some threshold? Regulator with a UVLO? A regulator that drops out instead of saturating? BTW, why a linear regulator? With any current, you're dissipating a lot of power (in both the zener and regulator).

What voltage is the zener? George H.

There are a number of regulators which have a quiescent current in the

Sorry. When the circuit is nominally on I want to make 12V from the 25V rail with a linear regulator. When the circuit is nominally off the 25V rail is at 4V, and I would like to effectively shut off the 12V line entirely.

A 6.8V zener would make for around 18V at the input to the regulator, so the "on" part works fine. I'm just wondering if, with 4V on one side and a regulator input on the other, if the zener will flow some predictably low current.

There seems to be ones that advertise 100nA at a 5V drop -- that would be enough for me.

That number is hugely smaller than the number of 3-terminal regulators out there, though.

I don't mean to bitch or insult, but I could do s*it like that when I was 13 years old. I also knew how to make the rails rise and fall in time so as not to damage anything.

Sounds pretty low for a 6.8V zener. At what temperature? Note that the zener will have to dissipate some amount of power.

I don't know about that. Tons of LDOs have enables.

Enable, AND >25V rating, is probably on the rarer side though.

I know! Drop the Vin with a zener so you can use a modest rated (20V?) LDO!

;-)

As for zeners, yeah, a >5V zener will do what you want.

On a related subject: there's a good reason why you don't see TVS diodes for very low voltages (under 5V). Zener devices cannot be made with a usefully steep curve. They "leak" like a sieve, and under surge conditions, the peak clamping voltage isn't much less than a regular 6.2V zener diode's (which is mainly what an SMAJ5.0A is, give or take the qualifications for surge duty instead of voltage regulation).

You can only get a steep, low voltage knee, using active circuitry (think TL431). But you can't handle huge pulses with small, cheap and fast parts: the circuitry incurs delay, and you end up using a power device in its linear range -- there's no current-multiplying breakdown effect available. The traditional TL431 + thyristor crowbar is the closest, but is only good for surges (slow enough to deal with) on self-clearing circuits (since the thyristor stays latched on otherwise).

So consider that, too. Your 4V requirement and 13V overhead means this is easily solved with standard zeners. :-)

Tim

I think that is pretty irrelevant. Digikey lists over 58 thousand linear voltage regulators. How many do you think will meet the requirements? I don't have any idea how it matters that there are many that don't fit the requirements. The design only needs one.

BTW, by definition there are NO 3 terminal regulators with an enable.

that the leakage should be in the tens of nanoamps at room temperature, even with a forward voltage slightly below the knee. We have not measured a 6.8V zener, but think that 100nA leakage near its knee is a reasonable expectation.

Well, if a 6.8V zener does not meet your fancy, try a reverse-biased E-B transistor junction; somewhere around 8V with tens of nA leakage..

Technically, the input IS the "enable"; 0V in for 0V out; min operating V in for rated V out (="enabled"). Just be picky....

Hi Tim... It's Friday. This is data from a little 6.8 1n5235 (I assume the part number is right.. it's written by hand on the package.)

V bias Current (volts)

George H.

The datasheet is your friend:

For example, on page 3 a 6.8V zener is listed at a max reverse current of 2uA when at 4V. As John pointed out higher voltage zeners are better but you might not have much choice because of the drop-out voltage in your regulator.

Of course, this also depends on what size zener you need. A big old fat one will have proportionately more leakage current. However, it's hard to find fat zeners that are stocked. So if you are dealing with lots of current I'd consider a transistor circuit instead. If this is for longer term production I would not rely on the continued availability of a larger size zener diode because that can turn into a boomerang.

There are giant TVS zeners around. We use them with polyfuses to protect DC power inputs, from wall warts. I should measure leakage on some of them.

I wouldn't call it "picky"...

How so, IMHO, they're close to optimum in many ways.