2
\$\begingroup\$

I am attempting to use a N-Channel Logic Level MOSFET as a switch for a phone charger that runs on a 42V battery.

I have a resistor circuit with an LDR that currently outputs 3.3v when LIT with a torch and 0v when completely dark. Daylight outputs around 0.7V. The circuit works as expected for a few uses and then the MOSFET fails short and is now always on permanently (even when removed).

schematic

simulate this circuit – Schematic created using CircuitLab

My Guess: Could the load (5V Step Down Regulator/Phone Charger) be inductive, similar to a relay and need a flyback diode somewhere? Placing one on +/- of the charger has no effect on longevity, as well as one on the main +/- of the battery.

The current of the charger should be less than 200ma at 42V for a 5W charger. Could it be higher at startup (it's rated for 30A).

They seem to fail when it is switched on/off more frequently.

MOSFETs can be placed in parallel to split the load, would this help?

I have killed more than 10 MOSFETs so far trying different combinations

The MOSFET is FQP30N06L with this datasheet. The Phone Charger is this 5V Regulator

\$\endgroup\$
8
  • 1
    \$\begingroup\$ 1) please draw a proper schematic using the correct symbols (like for the MOEFET) etc. There's a schematic entry tool available when you edit your question. 2) details are important, the MOSFET is clear but "LDR circuit" and "5V regulator" are not, what are they? Include a schematic and/or datasheet. 3) My guess is that the 0-3.3 V you apply to the MOSFET's gate-source is the issue. 3.3 V isn't enough to fully turn on this MOSFET. That means it is not fully closed, doesn't have the low resistance it should have and power will be dissipated in the MOSFET. \$\endgroup\$ Feb 6 '20 at 7:33
  • 1
    \$\begingroup\$ ... maybe mounting the MOSFET on a heatsink will help but that doesn't address the real issue which is that you need to apply either 0 V (off) OR a much higher voltage like 10 V (on) to the MOSFET. Anything in between only partially switches on the MOSFET and then it might be damaged. Instead of a MOSFET maybe using a relay module that can work with a 3.3 V input be a better solution. Relays can only be on or off, there's no "in between" state like with a MOSFET. What makes you kill more than 3 MOSFETs before you decide that maybe you're not doing this right? \$\endgroup\$ Feb 6 '20 at 7:37
  • \$\begingroup\$ MOSFETs can be placed in parrallel to split the load, would this help? Yes they can and that could help but it is a "stupid" solution as again that would not address the real issue (Vgs too low). But if you connect many MOSFETs in parallel and don't make sure they're all at the same temperature and all have similar series resistance due to wiring, it can still go wrong. Then the load will not be share equally so the MOSFETs will be damaged all the same. \$\endgroup\$ Feb 6 '20 at 7:42
  • \$\begingroup\$ @Bimpelrekkie Thanks, I will try add a proper one. Added some more details to the question. It is heatsinked and nothing gets warm to the touch in any state, even when charging normally for extended periods of time, indicating its not dissipating very much power at all, even when at 3.3V. \$\endgroup\$
    – WaffleFPV
    Feb 6 '20 at 7:52
  • 1
    \$\begingroup\$ Would this really be damaging to it? What is a "small amount of time" to you is more than enough time to overheat and damage a MOSFET. If the "5V regulator" is a switching regulator (it would be a very poor design choice if you used a linear regulator instead of a switching regulator with 42 V in, 5 V out, that screams for a switched type) then when the MOSFET isn't fully on, a voltage develops across it. Then the 5 V regulator gets a smaller voltage, to compensate it starts to draw more current that means more dissipation in the MOSFET. \$\endgroup\$ Feb 6 '20 at 8:10
2
\$\begingroup\$

Your LDR is a very slow device that responds to daylight that also varies slowly .Sunrise and sunset can take an hour or more depending on season and lattitude .This means that your power mosfet will spend lots of time half on or half off .Under these conditions the power wasted and hence device heating will be bad . Cheap modern powermos devices are specified for switching .They can current crowd in analog mode .This internal current crowding can cause early failure due to internal hotspots.The gate should be close to 5VDC for charger on or close to 0VDC for charger off .This is also better for the charger because it gets clean 42V for on and zero for off .You can employ schmitt action by whatever means to switch clean and stop blowing fets.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.