How to eliminate voltage spike from MOSFET switching?

Thread Starter

jaricurry

Joined Nov 9, 2018
7
Hi, I'm designing a circuit that with a pir sensor that is amplified by two op amps. The signal is than sent to a transistor which drives an LED. There is also an RC timer. The problem is that when the mosfet turns off it retriggers the op amp. With a BJT this does not happen. I would like to use a mosfet to conserve amps. Is there a solution. I have tried a RC scrubber tied into the drain with zero success. Thanks.Untitled.jpg
 

Sensacell

Joined Jun 19, 2012
3,447
If your load is an LED, it's unlikely that you are generating any spikes to snub.

What I do see is the opamp output being basically shorted to GND when its output goes high- driving the 10 uf cap through the diode.
Than might be upsetting things.
 

Thread Starter

jaricurry

Joined Nov 9, 2018
7
Do you have decoupling caps on the opamps?
yes I do have a 1 uF cap decoupling the op amps

If your load is an LED, it's unlikely that you are generating any spikes to snub.

What I do see is the opamp output being basically shorted to GND when its output goes high- driving the 10 uf cap through the diode.
Than might be upsetting things.
This is the best RC timer I could configure without placing a capacitive load on the op amp
 

Sensacell

Joined Jun 19, 2012
3,447
Add a series 4.7k resistor with the diode to limit the peak current when the opamp goes high.
Then the opamp won't see a momentary short.

But I don't think that's the problem.
 

Alec_t

Joined Sep 17, 2013
14,313
Wouldn't it be better to configure the FET as a low-side switch?
With a supply of only 3.3V, a FET with a rated Vgs(thr) in the 2V-4V range and a drain load of 10k I'm surprised the LED gives a satisfactory light output.
I appreciate that you are trying to minimise current consumption, but the use of such high value resistors in the circuit makes it prone to erratic operation as a result of interference pick-up. You may need to screen it well.
 
Top