power rating of smd resistors

[engr_joni_ee]

I am wondering about the power rating of the smd package resistors for example 0603, 0805 and 1206. I have seen somewhere that the power taring is a function of tolerance. For example resistance tolerance 1% resistance points to higher power rating as compared to 0.5 resistance tolerance which is less power rated.

How can I find the rule or a table in which power rating of some common packages are given with tolerance.

 

[KlausST]

Hi,

How can I find ...

--> datasheet
It tells you (at least: me) everything you need to know.

****
I´ve never heard about a "tolerance" rule.

But for sure, every resistor has a temperture drift. So let´s say a resistor has 200ppm/°C of drift.
Then the value will change by 1% at 50°C change, and 2% at 100°C and 3% at 150°C
and for sure you may dissipate more power when you allow for more temperature change.

This needs no table. This basically is what "temperature drift" means.

Klaus

 

[engr_joni_ee]

I guess you mean if the resistance tolerance is 200ppm/°C.

At 50 degC it will be 10000 ppm/50°C.
Percentage = 10000/10^6 * 100
Percentage = 1 %

At 100 degC it will be 20000 ppm/50°C.
Percentage = 20000/10^6 * 100
Percentage = 2 %

and so on. This shows that tolerance is temperature dependent.

But this is also true that smd resistors are available in more then one power rating. For example 1206 is available in 250 mWatt and also in 500 mWatt. What the difference between them other then power rating ?

 

[KlausST]

Hi,

no.
The tolerance is NOT temperature dependent. --> The reisitor_value and the resistor_error is temperatue dependent.

Read about terminology of
* resitor tolerance
* resistor temperature drift / temperature coefficient

I guess it´s high time to open a resistor datasheet and read some informations provided by resistor manufacturers.

Klaus

 

[stenzer]

Hi,

the power rating highly depends on the used technology i.e. a thick film resistors can usually handle more power than a thin film one.

But as Klaus mentioned, have a look in the datasheet for example you may also consider the power derating vs. temperature. The start of derating temperature also differs between different manufacturers (e.g. 70°C vs. 85°C). You might also have a look on the pulse power handling capabilities if included in the datasheet.

BR

 

[KlausST]

....
pad dimensions, air flow, ambient temperature, orientation (H/V), altitude ... uptime, expected lifespan....

So not only the resistor itself, but also the PCB, case and your own requirements play a role.

Klaus

 

[cupoftea]

Hi,
Yes, there is a situation here, because a general 0805 resistor say, is said to be for 100mW.
But when i worked for a big Euro lighting co, we were told never to use them if power was >50mW.
And koa speer resistor website says nothing about the amount of copper pour that you need to give you the
"spoken" datasheet power rating.
And say, if you use a 2512 resistor on its minimum land pads, then it is not generally good for 1W.

At the end of the day, you need to thermal test at max ambeint and check the solder joint temperature
is <100degc. But thats very time consuming, to do that for all your smd resistors, so you have to
realise and get some decent copper pour area.

I mean, why not buy a load of 220 ohm 0805 resistors...and make yourself a wee 4.7V, 2W power supply...
then put the ten 0805 resistors across the 4.7V, each with a little more copper pour and thermal vias,
and copper weight...and see how hot they each get?

With SMD resistors, i garantee that you will be truly and utterly amazed at how much difference just a little bit extra of
copper pour makes to their temperature.

 

[engr_joni_ee]

How much margin is better in power rating ? If my calculation says that I would need 10 m Ohm shunt resistor across which 200 mWatt power will be dissipated them what should be my choice ? To find a resistor with 250 mWatt rating or to find a resistor with 500 mWatt rating ?
Is that true that power rating increase with the physical size of the resistor ? like 0603 is less power rated than 1206 package ?

 

[KlausST]

Hi,

Indeed power rating is not what I am looking for.
There are chip resistors with higher power rating, just because they allow them to get hotter.
But each 10°C of temperature rise results in half of lifespan. 20C --> 1/4, 30°C --> 1/8, 40°C --> 1/16

Thus I rather look for temperature rise and uptime.
For my long term industrial applications, on circuits with almost 100% 24/7 uptime I don´t allow for more than 50°C. Thus even for 200mW of real power dissipation I use a 1W rated resistor.

I have PCBs here where 1W rated resistors with less than 400mW of real power dissipation look like "burned" (including the PCB) after 20 years of operation.
The continous (relatively) high temperature makes the solder to gets grey and then open circuit.

Klaus

 

[cupoftea]

I have PCBs here where 1W rated resistors with less than 400mW of real power dissipation look like "burned" (including the PCB) after 20 years of operation.
The continous (relatively) high temperature makes the solder to gets grey and then open circuit.

Thanks, it makes you wonder how long a 2512 on minimum pads would last, with 1W in it, and on 24/7, and with minimal extra cooling copper on the PCB, and 1 0z copper, and limited free air, and 30degC external ambeint. I bet it would be burned up inside of a year....its solder joints gone open.