Why is my MOSFET shorting?

Thread Starter

Newguys

Joined Sep 24, 2019
8
I am running a 12V 2A power supply to a 12V blower and using an IRF840 as a switch. https://www.vishay.com/docs/91070/sihf840.pdf
There is 1.2A running from Drain to Source. When I put 12V from the supply to the gate with a parallel line, the MOSFET becomes shorted and will continue to conduct after the 12V is removed from the gate. If I leave the 12V connected to the gate, it eventually starts smoking.
 

Thread Starter

Newguys

Joined Sep 24, 2019
8
Last edited:

Danko

Joined Nov 22, 2017
1,135
I am just running a wire in parallel with the positive of the fan and power supply and manually connecting and disconnecting it to my gate
You charged capacitance gate-source and it will discharge some minutes. In that time transistor gradually changes its state from ON to OFF, and becomes hot.
Right way is to use wire, connected to gate by one end. When you touch "+" by other end of wire, transistor will ON. After it you should touch "ground" for discharge inner capacitance and transistor will OFF.
 
Last edited:

Thread Starter

Newguys

Joined Sep 24, 2019
8
You can't leave a MOSFET gate floating as it can assume any arbitrary voltage.
Connect the 10k resistor between the gate and source.
Oh I didn't know that! I thought it was just a pull-up resistor for the transistor. I'll test that out when I get back to it in a couple hours.
 

Thread Starter

Newguys

Joined Sep 24, 2019
8
Please make sure that you wiring the right pins of IRF830 MOSFET.
That is the first thing I checked. I'm also using an 840, essentially the same except slightly better with less internal resistance. But just to make sure, looking at the mosfet from the front so( the heatsink is on the opposite side), I'm wiring left as the gate line, the middle to the - of the fan, and the right pin to ground.
 

Thread Starter

Newguys

Joined Sep 24, 2019
8
Do have a back-emf protection diode across the fan?
The 12V blower I'm using supposedly has one built on, but I'll also try attaching one when I try the 10k resistor. The MOSFET is rated for 500V and 8A drain to source so I am almost positive the problem is the gate to source.
 

ScottWang

Joined Aug 23, 2012
7,065
That is the first thing I checked. I'm also using an 840, essentially the same except slightly better with less internal resistance. But just to make sure, looking at the mosfet from the front so( the heatsink is on the opposite side), I'm wiring left as the gate line, the middle to the - of the fan, and the right pin to ground.
You can use a 3V/20mA LED and in series with a resistor(500Ω) to replace the fan to try again, and to see what will be happening?
 

Thread Starter

Newguys

Joined Sep 24, 2019
8
Update

I tried using both an EMF supression diode and the 10k ohm resistor, both together and seperately. Neither one had a noticable effect in the circuit behavior. I also switched the fan out for and LED and appropriate resistor. When the circuit is powered, the LED is dimed, when I send voltage to the MOSFET gate, the LED fully lights up, and stays fully lit even after removing the voltage to the gate. This picture isn't the best quality but maybe it will help. Ignore the stuff at the very bottom, it's a different circuit.
 

Attachments

Alec_t

Joined Sep 17, 2013
12,215
You appear to have the 10k resistor in series with the gate drive signal. The resistor should, instead, be connected between the gate and the source of the FET, to ensure that the FET switches off when the drive signal is absent.
 

Thread Starter

Newguys

Joined Sep 24, 2019
8
You appear to have the 10k resistor in series with the gate drive signal. The resistor should, instead, be connected between the gate and the source of the FET, to ensure that the FET switches off when the drive signal is absent.
Ohhhhh I thought by source they meant voltage source not source pin...thinking about it that makes much more sense and seems to be working as intended with both the LED and fan!! Now if I want to control it with a 3V source, is there an easy way to connect it and another resistor that would provide a sufficient voltage drop to the gate? I can't just wore it in series with a resistor because I have to keep the current at 1.2 amps for my fan I have one buck converter I could step up the voltage with if I really had to, but I would prefer not to use that. Thank you all so much for your help by the way.
 
Last edited:

Thread Starter

Newguys

Joined Sep 24, 2019
8
Nevermind, my original idea of using a 470R pull-up from 12V and connecting the transistor collector to gate, emitter to ground, base to pwm works. I just wired it incorrectly while testing!
 
Top