Understanding transistor Ptot in data sheet

[SwedishWings]

Hi,

referring to the attached image, I wonder how total power dissipation is defined in each case; "Tamb < 25C" and "Tc < 25C" respectively. I suspect that Tc is referring to usage with a heat-sink , but it makes no sense as the TO92 case is not well suited for heat sinks.

Thanks in advance!

 

[AlephZero]

Hi,
referring to the attached image, I wonder how total power dissipation is defined in each case; "Tamb < 25C" and "Tc < 25C" respectively.

Ta = ambient temperature, Tc = case temperature.

I suspect that Tc is referring to usage with a heat-sink, but it makes no sense as the TO92 case is not well suited for heat sinks.

Really?
http://uk.rs-online.com/web/p/heatsinks/7124320/
http://www.rapidonline.com/electronic-components/to-92-heatsink-77178
Or make your own from a small metal plate held onto the transistor with heat-shrink tubing!

It's unlikely a heat sink would keep the case temperature down to 25C, but the data sheet should give the derating factors for higher temperatures, as in http://www.onsemi.com/pub_link/Collateral/P2N2222A-D.PDF

 

[berkeman]

Hi,

referring to the attached image, I wonder how total power dissipation is defined in each case; "Tamb < 25C" and "Tc < 25C" respectively. I suspect that Tc is referring to usage with a heat-sink , but it makes no sense as the TO92 case is not well suited for heat sinks.

Thanks in advance!

Were those numbers for the TO-92 package version, or maybe for a different package option (like metal can)?

 

[jim hardy]

There's a "Thermal resistance to case" and a different "Thermal resistance to ambient".
Many transistors have their collector bonded right to the metal case for good heat transfer.
As you observed - with a heatsink you'd add 'to case' thermal resistance to heatsink's thermal resistance to get thermal to ambient...

I like the metal can (TO39?) for troubleshooting.

The one on the left had its base wire melted off inside, emitter too if i recall. We attributed that one to human error not aging. Probably a 'scope ground probe got plugged into the wrong test point.

 

[SwedishWings]

Were those numbers for the TO-92 package version, or maybe for a different package option (like metal can)?

Yeah, i think i might have mixed up the data sheets, that would be the TO-18 case. But are there any heat-sinks for TO-18?

 

[SwedishWings]

Ta = ambient temperature, Tc = case temperature.

Really?
http://uk.rs-online.com/web/p/heatsinks/7124320/
http://www.rapidonline.com/electronic-components/to-92-heatsink-77178
Or make your own from a small metal plate held onto the transistor with heat-shrink tubing!

It's unlikely a heat sink would keep the case temperature down to 25C, but the data sheet should give the derating factors for higher temperatures, as in http://www.onsemi.com/pub_link/Collateral/P2N2222A-D.PDF

I totally missed the fact that there are heat-sinks for those casings. Thanks!

 

[SwedishWings]

There's a "Thermal resistance to case" and a different "Thermal resistance to ambient".
Many transistors have their collector bonded right to the metal case for good heat transfer.
As you observed - with a heatsink you'd add 'to case' thermal resistance to heatsink's thermal resistance to get thermal to ambient...

Thanks for clarifying, think i understand now. But does the Tc based Ptot refer to _zero_ thermal resistance to 25C? That would be purely theoretical, right?

I like the metal can (TO39?) for troubleshooting.

Great idea, never thought about it =)

 

[SwedishWings]

It's unlikely a heat sink would keep the case temperature down to 25C, but the data sheet should give the derating factors for higher temperatures, as in http://www.onsemi.com/pub_link/Collateral/P2N2222A-D.PDF

What does "Derate above 25°C - 5.0 mW/°C" actually mean? Is it the SOA that falls with 5mW/C?

 

[jim hardy]

But does the Tc based Ptot refer to _zero_ thermal resistance to 25C? That would be purely theoretical, right?

Zero thermal resistance case to ambient, yes. If you could keep the case at 25 deg you could dissipate 3 watts.

Take a look at the thermal resistances given here for TO-39 case
http://www.st.com/web/en/resource/technical/document/datasheet/CD00003223.pdf

compare to the plastic case...


try the numbers and see if both conditions get you to max junction temperature at max power dissipation.

 

[AlephZero]

I would take the Ta values as meaning the device was on a circuit board in air, without any heat sink.

If you have a heat sink, you should know the thermal resistance between the case temperature and the air. For example the RS link gives the thermal resistance as 60 degrees C per watt.

So if you were dissipating 0.5W with that heatsink and Ta = 25C, Tc would be 25 + 60 x 0.5 = 55C.

The derating parameter tells you how the power rating decreases as the case temperature rises. If the max power rating is 1.5W at Tc = 25C and the derating is 12mW/C, at Tc = 55C it would be 1.5 - 0.012*(55-25) = 1.14W.

If you increase the actual power dissipated, Tc will rise and the max power rating will fall. When the actual power = the max power, that is the operating limit for the transistor with that type of heatsink.