Power Supply TO220 vs TO3?

[danial stocks]

  Now I realize, it is going
>to be difficult to pick a heat sink to do this because information on this 
>is
>difficult to find (does anybody know a good place to get this?).

This info is in pieces usually which is why it is difficult to find.. first, 
the ultimate thing which destroys the semi is actual chip temperature. If 
this isn't stated, then a max case temp will be stated. then you need to 
know the thermal capacity of the heatsink. this is usually specified in the 
place where you buy the heatsink. for instance, say you had a heatsink with 
a capacity of 5'c/W. This means that the temp of the heatsink is going to 
rise 5'C for every watt it has to dissipate. this is a 5'c rise per watt 
over the ambient temperature. So suppose you want a maximum permissible 
device case temp of 80'c, and you have to dissipate 10W in the device, and 
you reasonably expect that the ambient temperature may be a maximum of 50'C 
[take into account a hot day, + transformer heat etc inside the device] so 
we can tolerate a 30'c temperature rise. this means that our heatsink must 
be 3'c/W max, including the coupling between the case and h/s.. there is a 
thermal gradient between chip and case, case and heatsink, heatsink and air. 
you the engineer have control over the last 2. use of a TO3 makes for a 
lower thermal gradient over a TO220 to the same heatsink,thermal grease also 
helps.. I dont have the spec for the thermal gradient for to3 or to220, but 
the relevant databook should mention this.
Cheers,
Dan





A good rule of
>thumb is the bigger the heat sink, the more heat it will get rid of, but 
>also
>remember, that the poor little TO220 has a maximum disipation that you 
>cannot
>exceet (should be on the data sheet for the device).
>
>     -Jim
>
>Dave Wilke wrote:
>
> > Design and planning continues for my modular (mostly based on EFM's 
>stuff).
> >
> > An engineer friend dropped by and said my +/-12v PS design would simply
> > catch fire because I'd used TO220 7812 & 7912 regulators instead of 
>going
> > with TO3 packages.
> >
> > Should I be concerned?  What would the expected current draw be from 
>(let's
> > say) 18 modules of the VCA5b and deg2a sort?  Can I get away with 
>TO220's or
> > do I need to do a (third) redesign using TO3's?
> >
> > dave
> >
> > 
>_________________________________________________________________________
> > Get Your Private, Free E-mail from MSN Hotmail at 
>http://www.hotmail.com.
> >
> > Share information about yourself, create your own public profile at
> > http://profiles.msn.com.
>

_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.

Share information about yourself, create your own public profile at 
http://profiles.msn.com.