Keep MOSFET "ON" for a certain time

Thread Starter

pplg

Joined Nov 19, 2019
14
Hello everyone!

I'm currently designing a simple circuit which is working as I want but I'd like to make sure that I haven't done any dumb mistake. That's why I'm asking here.

I'd like a signal to switch ON a MOSFET in order to activate a relay. This signal is usually low (0 VDC) but sometimes it becomes a pulse train of +3VDC at around 3Hz. That's when I want my MOSFET to be ON. But ALWAYS ON until it stops oscillating and goes back to 0VDC. I don't mind if it stays 2 or 3 seconds ON when it has already stopped oscillating.

So here's my design. It works, but again I'm not sure I did it right.

I used a 1nF capacitor in order to keep the gate high when the pulse wave is at low state and a diode so it can only discharge through the 100 Ohm resistor and the gate impedance (around 200 MOhm).

Is this right? Is there a better (or simpler) way of doing it?

THANKS!!!
 

Attachments

crutschow

Joined Mar 14, 2008
25,433
Is this right?
Basically.

The MOSFET needs to be a logic-level device that will fully turn on at about 2.3V (Vgs max threshold of 1V or less).

Using a Schottky diode for the input diode will give you a little more gate voltage.

If you let it gate capacitance discharge through only the gate resistance, the MOSFET could stay on a very long time.
Add a 10megohm or so resistor from the gate to ground.
 
Last edited:

jpanhalt

Joined Jan 18, 2008
10,246
How do you turn the gate off? Sure there will be some leakage, but if you put a large turn off resistor (calculate the RC time constant) it might work better.
 

Thread Starter

pplg

Joined Nov 19, 2019
14
Basically.

The MOSFET needs to be a logic-level device that will fully turn on at about 2.3V (Vgs max threshold of 1V or less such as here).

Using a Schottky diode for the input diode will give you a little more gate voltage.

If you let it gate capacitance discharge through only the gate resistance, the MOSFET could stay on a very long time.
Add a 10megohm or so resistor from the gate to ground.
Thank you crutschow! Yes, you're right! I noticed that. Using another resistor as you pointed will be all I need.

:)
 

Thread Starter

pplg

Joined Nov 19, 2019
14
How do you turn the gate off? Sure there will be some leakage, but if you put a large turn off resistor (calculate the RC time constant) it might work better.
It discharges through the gate, but VERY slowly. It takes like 15 seconds to turn OFF again. I'll use another resistor as you guys told me :)

Thanks you both!
 

TeeKay6

Joined Apr 20, 2019
572
It discharges through the gate, but VERY slowly. It takes like 15 seconds to turn OFF again. I'll use another resistor as you guys told me :)

Thanks you both!
The diode also has a reverse leakage current that for a 1N4007 (or a Schottky) could be as great as the MOSFET gate leakage.
 

jpanhalt

Joined Jan 18, 2008
10,246
Years ago, I used a similar circuit. Of course, it will delay turn off. A problem you may see is that because of the slow turnoff, the relay may chatter as the mosfet will pass slowly from being on to linear to being effectively off. In short, you may want a snap action. Adding a comparator with hysteresis will do that for you.
 

Thread Starter

pplg

Joined Nov 19, 2019
14
Years ago, I used a similar circuit. Of course, it will delay turn off. A problem you may see is that because of the slow turnoff, the relay may chatter as the mosfet will pass slowly from being on to linear to being effectively off. In short, you may want a snap action. Adding a comparator with hysteresis will do that for you.
Wow mate! That's exactly what I DON'T WANT to happen! I haven't had the opportunity to check it yet but the relay is going to be connecting a circuit with mains voltage so it can not chatter for sure... I get the idea of adding a comparator with hysteresis but I don't know if you're referring to a specific IC or a combination of components?

Thank you!
 
Top