I am building a simple battery undervoltage protection circuit. It should stop conducting once the input voltage drops below a certain point. This should be achieved by utilizing a PMOS and setting the gate voltage with a resistor divider to select the critical input voltage at which the PMOS is not turned on anymore. Before building it, I designed the circuit in LTspice, which includes a specific model for the MOSFET in use: https://goodarksemi.com/docs/datasheets/mosfets/BSS84.pdf
Simulation shows satisfying results:
Green is the "output" voltage across the load resistor R3 and blue is the "input" voltage generated from V1 (in reality my battery). The gate voltage (red) is moving from ~800 mV to ~500 mV.
Since this was looking good, did a quick layout and lasered a prototype:
Ground is on top and the input voltage is connected on the bottom.
(please excuse my terrible soldering skills)
I tested various resistor values, that's why a different resistor is in the picture. The circuit behaves very different compared to the simulation. The gate voltage node is sitting at very high voltages, even if I drastically lower the resistor R2. The shown circuit uses a 3.9 MΩ as R1 and 100 kΩ as R2. However the voltage at the gate is 3.5 V.
I connected a multimeter and measured the current from gate to ground (I shorted the gate to ground that way) and it shows a current of ~3 mA. This kind of current obviously cannot come through the path of R1, so it must be related to the MOSFET. Looking through the datasheet, it mentions a Gate-Body-Leakage current of max. 5 µA - this is a whole magnitude off my measurement. Thinking that the MOSFET is broken, I replaced it, that didn't change anything.
Am I missing something here? What can cause such a high (leakage?) current through the gate?
Update: The circuit behaves as it should without the MOSFET. The diode test of the MOSFET pins yielded following results: DS: 0.6 SG: 0.97 GS: 0.92 DG: 1.3 - the others were open.
Solution: A never soldered MOSFET doesn't measure the same. This prompted me to check the maximum soldering temperature for this part - its 300°C for 10 s at 1.6 mm from the case. I soldered everything with 330°C. I now tried again with low temperature solder at 280°C and the circuit behaves like predicted.
Another main probable cause for the failure is ESD - I did not follow proper ESD techniques and MOSFETs without built-in ESD protection are very sensitive to ESD.
