Power Dissipated on a SMD Resistor

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
Hi.
I am new to pcb design and I have to choose an SMD resistor to use in a project.
Currently, I have a library available (I use Altium) with these resistors:
RES_0201 ... RES_0402 .. RES_0306 and RES_1206
I understand that the numbers indicate the size (length and width) .. but I have a couple of doubts when choosing.

I need a 10k resistor on which a maximum of 20mA will flow .. i.e. Pdiss(max) = 4 W.
So I need to find out if one of the resistors I mentioned above can handle that power without breaking and causing damage.

Following this document ... I notice that on page 2 in "MAXIMUM RESISTANCE CHANGE AT RATED DISSIPATION" the maximum power dissipated is 0.4W (performed by MCA 1206). Being less than 4W .. I suppose I can't rely on those 4 resistors I wrote at the top.
Is this correct?
Did I read the datasheet correctly?
If this is correct, then I should look up (for example on DigiKey) and use an SMD resistor that holds more than 4 W and import it into Altium. This is not a problem, but what I am interested in is if I have reasoned correctly or not.


In the meantime, I also take this opportunity to ask if 4W on an SMD resistor is too much or not... I have never designed a ppcb so I have no idea what high and low values are. In any case, there are SMD resistors avaiable on DigiKey that can handle much higher power than this .. so I think there should be no problem.

Thanks!
 

Ian0

Joined Aug 7, 2020
9,677
Why do you need to dissipate so much power?
A 2512 resistor is good for 1 Watt with no thought given to the pcb layout, or perhaps 3W with the appropriate amount of copper and thermal vias.
Surface mount resistors get disproportionately more expensive as they get bigger, so putting several smaller values in series might be an idea.
Also you need to check the maximum voltage: 10k x 20mA is 200V, and that is the absolute maximum for some SMD resistors.
 

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
Why do you need to dissipate so much power?
A 2512 resistor is good for 1 Watt with no thought given to the pcb layout, or perhaps 3W with the appropriate amount of copper and thermal vias.
Surface mount resistors get disproportionately more expensive as they get bigger, so putting several smaller values in series might be an idea.
Also you need to check the maximum voltage: 10k x 20mA is 200V, and that is the absolute maximum for some SMD resistors.
You are right. I realised I had confused myself when I read 200V.

I should put a resistor between the pin of an stm32 and the gate of a mosfet (that commands other things), but I probably got the resistor value wrong.
Let's still assume the case of 20mA maximum (otherwise you could damage the pin of the micro) .. I don't use a 10k resistor but a 1k resistor, so Pdiss(max) = 0.4 W.
So a 2512 resistor should be fine. Right?

Would it be risky to choose a 1206 resistor ?? .. since as stated on the datasheet it holds a maximum of 0.4 W (which is exactly what I could achieve if I had 20mA).
I know it is unlikely to reach that current value but it is the reasoning that interests me .. for educational purposes.
 

BobTPH

Joined Jun 5, 2013
8,813
The resistor will see nowhere near 20mA average. And a 1K resistor will slow your switching time so much you will burn up your MOSFETs. The usual resistor is more like 10-100Ω.
 

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
The resistor will see nowhere near 20mA average. And a 1K resistor will slow your switching time so much you will burn up your MOSFETs. The usual resistor is more like 10-100Ω.
The switching time is Rgate*Ciss(input capacitance) ??
In general, what R*C values could burn out the mosfet? Is it written in the datasheet?
 

BobTPH

Joined Jun 5, 2013
8,813
Depends on your switching frequency. You do not want it to be in the active region for long. If less than 10KHz, you can ignore what I said.
 

Thread Starter

kalemaxon89

Joined Oct 12, 2022
234
La tensione massima sul pin STM32 è 3,3 V.
Con un valore della resistenza di 1kΩ la massima potenza dissipata è di 0,01W.
[/CITAZIONE]
It's 0.01W because you are assuming a current = 3.3mA .. but what if the current is greater than that?
If, for example, the current at the gate reaches 15mA, then power R*I^2=0.22W would be dissipated.

But ideally no current should pass through the gate of the mosfet in DC.
(i.e. it should be really small)
am I wrong?
 

Ian0

Joined Aug 7, 2020
9,677
When a capacitor is charged through a resistor, then the power dissipated in the resistor depends only on the value of the capacitor.
Your calculations are giving you the peak power, but that power only lasts for microseconds. If the current is lower, the pulse lasts longer, so the average power dissipated is invariant of the peak current. Even a tiny resistor takes longer than that to heat up. It is not the power dissipated that determines whether the resistor fails, it is the temperature that it reaches.
 

MrChips

Joined Oct 2, 2009
30,720
It's 0.01W because you are assuming a current = 3.3mA .. but what if the current is greater than that?
If, for example, the current at the gate reaches 15mA, then power R*I^2=0.22W would be dissipated.

But ideally no current should pass through the gate of the mosfet in DC.
(i.e. it should be really small)
am I wrong?
How can the current be 15mA?
 
Top