Maximum current for analog switch

the gate of a FET to turn it on with 5V and turn it off shunting to 0V

or some tens of ns.

ide the 4053 IC to yield the same Rdson, so it should have the same robustn ess and same peak currents etc?

There are plenty of products made to handle surges like this http://www.ti. com/product/ts5a3159-q1 ,

ESD "diodes" aren't used on later versions. ...Jim Thompson

What do they use instead?

Cheers

Phil Hobbs

They're called "snap diodes", but they're not diodes... that's all I can reveal. ...Jim Thompson

Okay, but how do they behave, if not like diodes?

Cheers

Phil Hobbs

You could add a resistor or bead which somewhat defeats the purpose because it slows switching.

This one has better switch current ratings and lower Rdson values but you can't use it at 5V:

I kept some really old databooks because they contain data on parts that are still available and used on hi-rel gear. Much more detail than found in "modern" datasheets.

if it is anything like the 5V tolerant input on various MCU/FPGAs looks like it is a ~zener

-Lasse

That might cause weirdies. But you can put resistors into either/both of the inputs to a SPDT switch whose common drives a fet gate.

Or inductors even.

Like an SCR. ...Jim Thompson

Okay, so something like this then:

Good to know.

Cheers

Phil Hobbs

Digging a bit deeper, it seems that they're actually really low voltage avalanche transistors, which now rejoice in the name "snapback".

Funny how these things come round again and again. :)

Cheers

Phil Hobbs

This device I think is pretty much purely a one/few of those:

They show a BJT-boosted zener in the equivalent diagram, but that would run out of hFE (or need too much junction capacitance), unlike an avalanche or thyristor structure.

I haven't tested one to see if it latches, though.

Tim

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

It's not ESD diodes, it's a parasitic BJT structure. "The voltage drop across the bidirectional switch", i.e., between terminals Y (positive) and Z (negative), when both are within nominal common mode range (not touching the ESD diodes as such).

At least, that's what I gather from it. The diagram doesn't show an obvious B-E path (but then, why would they bother?). But there are plenty of NPN (and PNP) structures around just waiting for some charge injection to happen.

Tim

"Snap diode" was a traditional name for step-recovery diodes.

Sounds like a classic ESD SCR latchup blunder. There ought to be a law.

Lots of analog switches will sort of turn on in strange ways if the signal i/o swings past the rails. Most HC405x parts do. That's just turnon of the series fet switches, not an ESD diode problem, but it's almost as annoying.

I remember that from 1982ish. I think "snapback" is the cool kids' lingo for these LV avalanche transistor things.

Of course since Jim's lips are sealed, we can tease him *ad lib*. ;)

Cheers

Phil Hobbs

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

I'm looking for work... see my website.

Does a transistor avalanche different from a zener... I'm thinking a B-C connection?

George H.