So I was reading the datasheets of some transistors, and I couldn't help noting that despite them being power transistors, they didn't specify their internal resistance nor their thermal conductivity. Now maybe I'm stupid and their on-resistance is negligible, their switching time instant, but it seems to me that a TO-220 package, like the TIP41C, is meant to dissipate power. Likewise with a TO-3PN package. There are also no power dissipation derating curves to be seen.
I tried searching online, but I'm not getting anything useful.

Anyone know what the thermal conductivity of transistors is?

Thanks!

 

Look for the specification: Thermal resistance, junction to case.

 

So I was reading the datasheets of some transistors, and I couldn't help noting that despite them being power transistors, they didn't specify their internal resistance nor their thermal conductivity. Now maybe I'm stupid and their on-resistance is negligible, their switching time instant, but it seems to me that a TO-220 package, like the TIP41C, is meant to dissipate power. Likewise with a TO-3PN package. There are also no power dissipation derating curves to be seen.
I tried searching online, but I'm not getting anything useful.

Anyone know what the thermal conductivity of transistors is?

Thanks!

It would be helpful if you link to one or more of the datasheets that you are having trouble with. Otherwise, we're shooting in the dark. You don't even bother to tell us if these are BJT or MOSFET or whatever type of transistors. We are not mind readers.

 

Look for the specification: Thermal resistance, junction to case.

Umm, I knew that... why does the datasheet from mouser not have that info?...
Where did you get your datasheet from?

It would be helpful if you link to one or more of the datasheets that you are having trouble with. Otherwise, we're shooting in the dark. You don't even bother to tell us if these are BJT or MOSFET or whatever type of transistors. We are not mind readers.

TIP41C
D4515

 

The answer lies in the reality that these are very old and obsolete parts with only a small sales volume and not recommended for new designs. It takes effort and resources to obtain and maintain datasheets and the less money the follow-on suppliers of parts no longer made by the original manufacturer the better. They may not even have the information themselves and are not willing to devote resources to obtaining the information.

I have a 1972 Power Data book (hardcover) from TI. I'll look and see if there is more information there. Problem is that TI tested and stood by those numbers for the parts that it made. There is no telling if the follow-on suppliers can make the same guarantee.


ETA: On p. 5-352 of the TI's The Power Semiconductor Data Book, it has:

Rθjc	Junction-to-Case Thermal Resistance	1.92 °C/W
Rθja	Junction-to-Free-Air Thermal Resistance	62.5 °C/W

I can't help you with the other part. Sorry.

 

A well-sold IC will have a second source that resembles its characteristics. Such datasheets may be incomplete.

 

A well-sold IC will have a second source that resembles its characteristics. Such datasheets may be incomplete.

There is huge difference between a second source and a legacy source.

 

Umm, I knew that... why does the datasheet from mouser not have that info?...
Where did you get your datasheet from?



TIP41C
D4515

It's a sad reality that data sheets often don't have the specific information you might want -- even information that you would think would be pretty broadly needed, like the thermal characteristics of a power transistor.

If the data sheet lacks that information, you have a couple choices. The best choice is usually to use another part that has the data you need to make good design decisions. That may or may not be a practical option, depending on your situation. If you are wedded to these parts for some reason, and are willing to take some risks, look at the data sheets other similar parts in the same case and see how much those parameters vary by. If they are all the same, then pick the worst one and perhaps derate it a bit for safety.

 

The answer lies in the reality that these are very old and obsolete parts with only a small sales volume and not recommended for new designs. It takes effort and resources to obtain and maintain datasheets and the less money the follow-on suppliers of parts no longer made by the original manufacturer the better. They may not even have the information themselves and are not willing to devote resources to obtaining the information.

I have a 1972 Power Data book (hardcover) from TI. I'll look and see if there is more information there. Problem is that TI tested and stood by those numbers for the parts that it made. There is no telling if the follow-on suppliers can make the same guarantee.


ETA: On p. 5-352 of the TI's The Power Semiconductor Data Book, it has:

Rθjc	Junction-to-Case Thermal Resistance	1.92 °C/W
Rθja	Junction-to-Free-Air Thermal Resistance	62.5 °C/W

I can't help you with the other part. Sorry.

The ratio between those two numbers tells you that for any amount of power you care to name over about 1W you better damn skippy have that part mounted to a substantial heatsink. If a design has to function in an environment where the ambient (Free-Air) temperature is nominally 50 °C, then 1 watt in free air will mean a junction temperature of 112.5 °C. Normally a semiconductor manufacturer will want you to keep the junction temperature at or below 125 °C. This actually allows a bit of safety margin but not much. Running hotter than that will almost certainly shorten the lifetime of the part.

They mount GPUs on blocks of metal with fans for exactly this reason.

 

Why not FJP13009?

 

Why not FJP13009?

I dunno. What is that?
Oh - it is another transistor. So, I don't have any better resources than you for finding this kind of stuff. You have to ask the suppliers for the information you need, or you can look at the specifications for TO-220 packages and use that generic data for all parts in that particular package.

 

Wikipedia has approximate thermal resistance values.

TK9J90E is TO-3PN package, thermal resistance is specified in the data sheet. This is a thermal resistance value close to the Wikipedia value.

Since the package was selected according to the amount of heat generated during IC development, it is a good idea to design with a margin for this thermal resistance.

 

Going back to your TIP41 data sheet, the info you wanted is there, just in a different form. Under 'absolute Max ratings' it says Ptot, total dissipation at case temp of 25degC is 65W, and underneath Tj, Max Junction temp is 150degC so Tjc = (150-25)/65 = 1.92degC/W, which is typical for a TO-220 package.

In general most modern TO-220 devices are good for around 1.5W with no heatsink at 25degC ambient that's roughly a Tja of 60degC/W, which is fine for short term testing, though if you want it to have a long & fruitful life a small clip-on heatsink is beneficial.

 

I can't help you with the other part. Sorry.

I didn't expressly mention it in my original post because I didn't expect you to have data for an obscure part like D4515. But thanks for trying.