# Mosfet Power Dissipation

#### M0WW0M

Joined Jan 25, 2022
3
I have circut, that is using a microcontroller to switch a logic level mosfet.
This mosfet is driving a load at about 500mA and 5V, but i am not sure if i can get away with just the logic level.
The datasheet states a RDs(On) max of about 50mOhm so mit Power loss would be
$P = I^2 * R$
so with values: $0.5^2A * 0.05 Ohm = 0.0125 W$

Is this correct or am i missing something, because that seams rather low?
Thanks in advance for anyone who can help!

Edit: I am using the AO3400A Link to datasheet

Last edited:

#### dl324

Joined Mar 30, 2015
14,715
Welcome to AAC!

#### ElectricSpidey

Joined Dec 2, 2017
2,148
The calculations are correct, but I would take an actual voltage reading on the source-drain to confirm it.

#### M0WW0M

Joined Jan 25, 2022
3
The calculations are correct, but I would take an actual voltage reading on the source-drain to confirm it.
Perfect! Thank you both for the quick response. But as the Mosfet can dispate 1.4w it shoud be fine either way right. The power loss isnt a huge deal for me

#### ErnieM

Joined Apr 24, 2011
8,318
For a slow event the calculation is correct. However, if the FET is switching rapidly, say in a buck or boost inverter, the switching losses can predominate.

I've seen such devices used in LED controllers rated for 25W of 12V LED control. These set brightness by PWM so they are switching, but not that fast into a non-inductive load.

That said, for about the same cost when I built a similar controller I used an IRLR2905ZTRPBF rated 42A in a beefier D_Pak.

#### Irving

Joined Jan 30, 2016
2,809
Yes, for static or low speed (<1kHz) operation, but for accurate assessment of the losses you need to add switching losses and for power handling you need to able to check junction temperatures. The 1.4W refers to the absolute maximum total losses sustainable (not a recommended operating condition) on the spec'd heatsink which for an SMD device is typically 2500mm2 of 1oz copper eg 50mm x 50mm on the drain lead/tab, or 25mm x 25mm of 2oz. For a device mounted vertically on a negligible heatsink in free flowing air its about 1W and if mounted horizontally surrounded by other parts about 0.5W, all of which are >> your 0.0125W.

Switching losses can be calculated from eq 3, 4 and 5 here, but for this application they would be negligible.

Junction temperature can be calculated as Tjunc = Tamb + RΘja * Power. For reliability Tjunc should be <= 130degC and for RΘja for non-heatsink, horizontal mounting, in an enclosure/non-free-flowing air use the listed RΘja (here given as 125degC/W) * 1.5 or 2 as a good rule of thumb.

#### M0WW0M

Joined Jan 25, 2022
3
Thank all of you guys so much for the quick and detailed resonses! The circut is on for 30min once. Thats it. So switching isnt a problem.

p.s. I am quite new here. Is there a way to mark this as answerd or something similar?

#### dl324

Joined Mar 30, 2015
14,715
I am quite new here. Is there a way to mark this as answerd or something similar?
Once you get 10 posts, you should be able to edit the title. Adding [solved] will stop people from replying and summoning you back to a thread that you no longer care about.