the MOSFET not fully turning off? Unexpected 1.1V instead of 0V

[adamskyyy]

I'm working on a circuit using an **IRLML6402 P-channel MOSFET** to control power to a **Capacitive Soil Sensor V2.0**. The schematic is attached.

### **Expected Behavior:**
When the MOSFET is supposed to be off, I expect the **Vcc** pin of the soil sensor to be **0V**.

### **Observed Issue:**
Instead of 0V, I measure **1.1V** at the **Vcc** pin when the MOSFET should be off.

### **Key Circuit Details:**
- **Source (Pin 2)** is connected to **+3.3V**.
- **Gate (Pin 1)** is controlled via a **2.2kΩ pull-down resistor (R3)** and a **4.7µF capacitor (C4)**.
- **Gate-Source voltage should be 0V when off**, but the MOSFET seems to still conduct slightly.

### **Questions:**
1. **Why is the MOSFET not fully turning off?** Could there be leakage current causing this?
2. **What could be the reason for the unexpected 1.1V at the sensor's Vcc?**
3. **Is the gate pull-down resistor (2.2kΩ) too weak, or is the capacitor affecting turn-off time?**
4. **Should I try a stronger pull-down resistor or another MOSFET with a lower threshold voltage?**

Any insights would be greatly appreciated! Thanks in advance.

 

[Diver300]

How much current does the soil sensor take at 1.1 V? At very low voltages like 1.1 V the sensor may take so little current that the slightest bit of leakage could cause that voltage.

I suggest that you try with 10 kOhm between Vcc of the sensor and ground.

 

[Tony Stewart]

The condition for Vgs =0 is called Idss.
Your sheet says this max leakage -1 to -25 uA @ -20V from 25'C to 70'C
Using V/I=R this is equivalent to 20 M to 1 Meg resistance.

Yet the schematic says capacitance sensor and the above is resistance.

So where is your AC pulse current measurement? This will not do it and if the FET capacitance is more than dry soil, that is a bad design. A CMOS Schmitt inverter Osc. would be adequate with a C analog bridge.

 

[Tony Stewart]

Please learn how to read datasheets and define how the design should work with specs.

 

[rjenkinsgb]

He's using an active "capacitive soil sensor V2.0" module that has an analog output when powered.

eg.

https://www.aliexpress.com/item/1005004708790625.html

Pardon our interruption...

 

[ahsrabrifat]

You can test with a slightly lower value of R3. For example, 1K.

 

[Tony Stewart]

It it does not behave as expected, it may be damaged or miswired or the sensor has some storage voltage with low leakage decay time, in which case is not a problem if it does decay and there is no input current.

The sales page contains no info to describe how it works. The typical sensor does not show any schematic but it does have a 555 timer and an SOT-23 3 pin device that may be a 3V LDO to provide the sensor output. 0, 3V.

It is a moisture controlled frequency using the 555 clock when the capacitance detects water that inversely controls frequency with moisture content.


Can you see anything on the interface pin (data) under dry/moist conditions or in the info that came with the product and tell us?

Nothing you have showed in your #1 post schematic is helpful until you decide what you need or want to do with that and tell us.

If you just want a power switch, Check that you have the correct fpins connected, measure current and voltage to verify the sensor 1st, then your external switch 2nd. If you are uncertain how test a transistor switch, say so with test data.

.

 

[Tony Stewart]

If the sensor is CMOS and has a high impedance that produces 1.1 V with the FET leakage spec and a 10 Meg probe, verify the leakage with just the probe on drain. I think your choice of FET is too small an RdsOn which accounts for high leakage. Please measure and report results.