Heating up of Charger IC

[Freshman]

I'm using this charger IC - MCP73842T (https://ww1.microchip.com/downloads/aemDocuments/documents/APID/ProductDocuments/DataSheets/21823D.pdf). The IC is driving my PMOS (https://www.vishay.com/docs/63262/sis413dn.pdf) gate about 2V with input voltage Vdd = 11.5V. The datasheet suggests a higher voltage (4.5V to 7V). With my charging current of 0.5A, the MOSFET is getting very hot.

What's the main reason my MOSFET is getting hot? Is the DRV pin of the Charger IC not driving my MOSFET properly?

Since the MOSFET part (https://www.digikey.in/en/products/detail/onsemi/NDS8434/244237) mentioned in the evaluation guide (https://ww1.microchip.com/downloads/aemDocuments/documents/OTH/ProductDocuments/UserGuides/51267b.pdf) is obsolete, I need to select another MOSFET and see if it helps.

I have selected this alternative MOSFET (https://www.vishay.com/docs/63239/sis407adn.pdf). Any suggestions whether it would help?

If not, please let me know the parameters which I need to check for selecting a new MOSFET to ensure it does not heat up.

[tunk]

Absolutely no expert on this, but datasheet says it's a
"linear charge management controller". An uneducated
guess is that the excess power is bled off in the mosfet,
maybe:
(11.5-3)*0.5= 4.25W

[Freshman]

Thank you @tunk. Will my alternate MOSFET part suit? What should I look for in the new MOSFET part?

[xvr]

Will my alternate MOSFET part suit?

It will get very hot (110°C above ambient). If you can live with a MOSFET at 135°C, you'll be fine.

What should I look for in the new MOSFET part?

On the package on which the radiator can be installed.

Or on different charger IC (not linear)

[squadchannel]

You need to understand linear topology.
Linear topology means that any difference between the input and output voltage is dissipated as heat.
It is "no surprise" to get hot. changing the FET to something else will not change that.
what you need is an IC that uses a switching topology.

[BillyO]

What kind of heat sinking are you applying to this?

Can you show a picture of your PCB layout and maybe one of the finished board too?

[Freshman]

I have not installed any heatsink.

So, you mean to say, the difference between input and output, (11.5V - 8.4V)*0.5A =  1.55W, will be dissipated in my MOSFET?

And regardless of whichever MOSFET I take, this is going to be the same issue?

[tunk]

A fully discharged li-ion is 3V, so more like this:
(11.5-6)*0.5= 2.75W

[BillyO]

I have not installed any heatsink.

So, you mean to say, the difference between input and output, (11.5V - 8.4V)*0.5A =  1.55W, will be dissipated in my MOSFET?

And regardless of whichever MOSFET I take, this is going to be the same issue?

Yes, but more like 3W for fully discharged cells.  The energy must go somewhere and that somewhere is the MOSFET.  The heatsink can be a large area of copper on the PCB where the dissipation pins solder to.  You can put this on both sides and connect with thermal vias to get extra cooling.  You can also reduce the input voltage.  Putting 2 rectifiers (1N4001) in series will drop the voltage down by 1.5V which will help, but you will still need some heatsinking.  Google "using copper pours as heatsinks".

[Freshman]

Thank you all for the suggestions. So, I understand that changing the MOSFET isn't going to work.

Proper thermal relief, heatsinks and better MOSFET package with good thermal properties might work.
Changing MOSFET isn't going to help.

Also, any idea if we have some charger ICs (probably switching charge controller ICs instead of linear) that I can try instead of this linear charge management controller (MCP73842)