IRF7319 dual Mosfet Latching switch

[dumpystig]

I have breadboarded the circuit below but can't get it to work as intended. As soon as power is connected the load (an LED just for test purposes) illuminates and stays lit. Operation of the momentary switch doesn't have any affect. I have transposed the Mosfet pin numbers from the datasheet to the diagram in red, but think I may not have done it correctly.
Can anyone help me out with this?

[dumpystig]

*bump*

[HackedFridgeMagnet]

Hi
Your circuit wont work, I think it needs another transistor to give it some sort of memory.

Have you watched this Blog? Dave has a good circuit.

http://www.eevblog.com/2012/03/30/eevblog-262-worlds-simplest-soft-latching-power-switch-circuit/


Probe around your circuit and check the voltages you can then identify when and what is holding the P channel fet on.

Also there may be a FET only version in the comments somewhere.

[Harvs]

Actually it does work.

Yes you got the pin numbers right.

The only thing that was apparent in when I simulated it, was without any output load to ground you could end up charging the gate of the N-channel FET from leakage of the P-channel FET.  This makes the switch in-effective, as it will switch back on within a few hundred ms of pressing the switch.  Maybe this is what's happening...

LED's have forward voltage probably above the threshold voltage for those FETs.  So try just putting a resistor directly from the output to ground, something between 1k to 10k (and leave your LED in place.)

[dumpystig]

Nice, adding a 1K resistor across the load has got the original circuit almost functional; but still when initial power is applied the circuit is on instead of off, although subsequent switch presses work as intended.

What about changing the cap value? Incidentally, the cap I'm using is a ceramic that I bought specifically for this build and I don't have a stock of any other values - would an electrolytic (polarised) suffice, providing I have the correct value?

[dumpystig]

Just an update on this circuit which is now operating as required. The load needs to be of low impedance so I used a 12V coil to switch the actual circuit I wanted to operate.

But it was still in the ON state when initial power was applied and I needed it to be OFF.
A guy on another forum suggested a cap across the GS of N MOSFET  should limit the gate voltage from turning on, and that was the solution.

If you want the circuit to be ON when powered up omit the cap across 1 +2 in the circuit below.

I've attached an updated diagram, maybe this be useful to someone else in the future.

[morato]

Hi, my name is Paulo and I'm from Sao Paulo, Brazil.

I used this circuit with some modifications to "resolve" cheap caliper consumption. I cut all original traces of power push button and + battery. The circuit consumption is ~6uA, caliper ~17uA (on or off) in original state. After modification 24uA ON and 0.013uA OFF (Keithley 196A).

Attached is a LTSpice file and pictures.

OBS: R1 = R2 = 470k, baterry voltage 2.27~3.20V

[perieanuo]

Waw that's perfectionism I just put 2 wires with some Aa battery


Envoyé de mon iPad en utilisant Tapatalk

[kondzio]

Hi
Any one try to exchange physical switch with mosfet and obtain the kill function? Or there is any onther way to obtain kill function?