Bidirectional switch

Combining two negative power supplies for fault-tolerance? (I use the P-channel MOSFET polarity inversion protection trick all the time on positive supplies.)

Cheers

Phil Hobbs

Or maybe it wasn't the complete French breakfast? No kidding, seen it in the south, and they considered this the whole breakfast: A large cup of cafe au lait, a small croissant, and a Gauloises cigarette.

Seems to be popular in 1-cell apps:

While this requires the use of a charge pump N-channel FETs have better figures of merit.

Le Sat, 20 Jul 2013 15:51:06 -0400, Phil Hobbs a écrit:

Working on a multi-channels SMU (multi as in truck load of) and I"m seeking good candidates for the output switch. The dual package would save a lot of space...

The low tolerated leakage makes this an interesting exercise in itself...

Le Sat, 20 Jul 2013 12:58:08 -0700, Joerg a écrit:

Except for the cigarette I go for this sometimes. Have a wonderful baker

But you can use 2 N channels with the sources connected. ... (reading the IR document...)

Hey, that's it. Look at fig 4. The common drain lends itself to natural monolithic integration and super low RDSon.

Nice but it won't fit my needs...

Some commentary on French and German breakfasts, among other things:

Cheers

Phil Hobbs

I absolutely love American breakfast with bacon, hashbrowns, toast, eggs, and the whole nine yards. Of course we only have that on Saturdays because this is a sure-fire method of sending someone into the cardiology lab.

Doesn't make it any easier to drive though.

That's the beauty of it, plus the FOM advantage.

Not sure what your needs are. But when currents are modest and Rdson is not super important I prefer two P-channels with the sources tied together. Much easier to control.

Greht! I've briefly stayed in Scotland, they have a nice sense of humor.

you can often use an n-channel in the negative instead they are usually available in much smaller packages with low rds on

-Lasse

Depends on whether it's okay to disconnect the ground in fault conditions.

Most of the time my stuff is connected to other expensive gear, so it's not the most considerate idea to let its grounds get dragged to some large negative potential if somebody connects the power supply up wrong. A P-channel just sits there looking at you.

Cheers

Phil Hobbs

it depends what you call ground, if the supply is isolated you aren't really disconnecting ground

-Lasse

Since it only does anything in fault conditions, how can you possibly know?

Cheers

Phil Hobbs

They only included the Drain pinout on the device, not the device schematic. You had me for a moment.

Cheers

Examination question #1: What is wrong with the datasheet drawing?

...Jim Thompson

It's not a 3 terminal device.

Cheers

oops, It's not a 4 terminal device.

Cheers

The body should be common to both devices.

BTDT. ...Jim Thompson

Clearly intended for high current Li-Ion battery packs used in power tools. Go here

and enlarge Series Protection FET Configuration diagram for your application ckt.

Some design philosophy here: "Circuitry in a battery pack, such as a gas gauge, needs to measure the battery-cell stack voltage at all times. This drives the decision to place the Li-ion protector FETs between the ground connection of the battery electronics and the negative pack terminal."

--- I think there's something basically wrong with the schematic on the data sheet in that the substrate diode on the top part of the totem pole is shown the wrong way 'round if the supply voltage is positive.

Turned around, with its cathode blocking the positive supply, the gate voltage will determine the conductance of the channel and the current from the supply into the load.

On the low side of the totem pole, since the load is grounded and its drive is to be bidirectional, its source must go to a voltage more negative than ground.

Consequently, the gates' drives must be separate and the gates not paralleled.

I think...

-- JF