LM317 thermal resistance and heatsink selection

[Elerion]

Looking for thermal resistance in the LM317 datasheet from TI, I found something which doesn't make any sense to me.

https://www.ti.com/lit/ds/symlink/lm317.pdf
Page 4

There're two TO220 packages, which seem differ in the thickness of the tab.
One of them is 4.2 ºC/W junction-case, which is close to the typical 5ºC/W assumed by many people. But the other one is just 0.1 ºC/W ??



Can it be that LOW ?
I think that either I'm missing something, either there's a mistake in those figures.
Anyone knows?
Thanks

 

[crutschow]

Yes, that thermal-resistance value does seem to be abnormally low.
I suspect a typo.
For example this regulator (page 2) shows a thermal resistance of 3°C/W for that package.
I would use the value for the LM317 KCT package.

 

[ronsimpson]

I thought going to the LM337 data might give more information. No help.

 

[Elerion]

The onsemi datasheet shows 5ºC/W.
This is the figure I'm using.
Strange that TI datasheet is wrong for this classic IC.

 

[crutschow]

Strange that TI datasheet is wrong for this classic IC.

Typos get into even the best written documents.