Soft Latching Power Switch Too Fast

I have made this soft latching switch to control the power to my project but it is way to fast. I have to flick the switch as fast as I possible can to make it work. I have tried changing some of the resistors and the cap which does seem to make some change but not enough.

Ultimately I don't understand how the circuit works. I feel like if I understood I could make the necessary changes.

Please help.

jacob84401:
I have made this soft latching switch to control the power to my project but it is way to fast. I have to flick the switch as fast as I possible can to make it work. I have tried changing some of the resistors and the cap which does seem to make some change but not enough.

Ultimately I don't understand how the circuit works. I feel like if I understood I could make the necessary changes.

Please help.

When the circuit is OFF ...
Both Mosfets are Off / Open
The 10K / 100K resistors charge up the CAP.

When the Push Button is closed ...
The voltage in the CAP turns ON the Lower Mosfet.
The lower Mosfet turns on the Upper Mosfet
The Upper Mosfet now applies +V via the (right most) 10K R to the Gate of the lower Mosfet
The circuit is now Latched ON, allowing the Push Button to open.

When the circuit is ON ...
Both Mosfets are On / Closed
The lower Mosfet discharges the CAP via the 100K resistor.

When the Push Button is pressed the Lower Mosfet is turned OFF by the discharged CAP
When the Lower Mosfet turns OFF, the (left most) 10K R turns off the Upper Mosfet
The circuit is now latched off, allowing the Push Button to OPEN

Repeat , over and over ...

EDIT:
Verify: one resistor is 100K and the other two are 10K.
Verify: 1 microfarad capacitor.
Did you use the IRF7319 Dual Mosfet or discrete Mosfets?
The Mosfets need to Turn On and Off very quickly.

Maybe, a larger 2 uf CAP would help hold the GATE Voltage stable
until the signal propagates through both Mosfets and the circuit LATCHES?

Use a momentary push button, not a On/Off toggle switch.

When turning ON, I think you can hold the push button down forever.
Initially, there is a RACE condition because the (right most) 10K R + Load
are temporarily trying to discharge the charged CAP !
But hopefully very quickly, both Mosfets will turn ON and then
the (right most) 10K R will keep the lower Mosfet Gate ON and
keep the CAP charged = LATCH, even though now the lower Mosfet is ON
and it is trying to discharge the CAP through the 100K R.
Only when the push button is released will the charged CAP start discharging.

When turning OFF, I think you can hold the push button down forever.
Initially, there is a RACE condition because the (right most) 10K R
is temporarily trying to charge the discharged CAP !
But hopefully very quickly, both Mosfets will turn OFF and then
the (right most) 10K R + Load will keep the lower Mosfet Gate OFF and
keep the CAP discharge = LATCH,
even though now the lower Mosfet is OFF and
the 10K + 100K R are trying to charge the CAP.
Only when the push button is released will the discharged CAP start recharging

But if you have a load that's capacitive its not going to work very good because it will try to keep the bottom Mosfet turned on. you'll need to add a diode like the second picture.

mrsummitville:
.
Did you use the IRF7319 Dual Mosfet or discrete Mosfets?

Yes I am using the IRF Dual. I would also assume my load will be capacitive.

jacob84401:
Yes I am using the IRF Dual. I would also assume my load will be capacitive.

Does the Soft Latch function properly with just a 10K resistive load?

Why do you have large capacitance downstream of your on/off switch?
Doesn't your load discharge that capacitance, very quickly?

Hi,
Purpose built IC.

Tom...... :slight_smile:

Use a 22 µF capacitor in the original circuit.

LOl I would go back and read the whole page where this came from.
It tells how it works it tells why you may need to do it like post 2 It's from there page it tells you if there is a load that's capacitive. And there's lot's of those type loads that you need to block that with a diode and add that 10K before the diode.