LM78xx regs - what the ?!?

[NewUser]

I'm building a dual voltage p/supply using a 7812 and a 7824. Thought I'd see how they cope first with an output short circuit. I found while they recover immediately after the short is removed, their method of coping during the short leaves a lot to be desired IMO.

With the 7824 reg mounted on a 75mm x 25mm x 3mm (3" x 1" x 1/8th") al. heatsink, an output short causes a .8A current draw at the (27vdc) input of the reg, and within a minute the entire heatsink is incredibly hot, my guess is 120 deg C+. With the heatsinks mounted on metal studs to space them off the plastic enclosure, I'm worried that if a short occurs for hours (eg while I'm away from home) the studs would conduct the heat as well and possibly cause the enclosure to melt or even catch fire.

Is this how they are supposed to work or did I buy cheapo poorly designed regs?

 

[Hank Fletcher]

I'm no electromagician, but wouldn't all you need be an appropriately rated fuse? Or do you have some sort of application in mind that necessitates shorting the leads on the regulators?

 

[bananasiong]

You need a PTC.

 

[Roff]

78XX is designed to protect itself from shorts, not to protect the world from idiots (no slur implied).
It is up to the user to provide appropriate fusing and/or heat sinking, and to mount the parts on materials what won't melt, or worse, catch fire at low temperatures.

 

[NewUser]

The PTC suggestion sounds like the way to go. As for the others - as mentioned, a short causes an input current of 0.8A. I need around 1A from its output (which is, I think, 50% little less than its normal max. current output). So if I add a 1A input fuse it would never blow. While I didn't check output short circuit current (will do though at the next opportunity), my understanding of these regs is that, upon an output short, the output current would not be enough to blow a 1A fuse, this being one of the ostensible features of the reg - no output fuse required.

>> Or do you have some sort of application in mind
>>that necessitates shorting the leads on the regulators?
I call it "preparing for the unexpected". While I don't design to cause short-circuits, in an imperfect world, and particularly with field-connected devices, it is an event that should be anticipated and allowed for, at least IMHO.

So while noone has directly anwsered my question, I assume that it is indeed normal for a 78xx reg to get extremely hot on a short-circuit? Please, a simple yes or no though!

 

[Roff]

According to the Fairchild datasheet, short circuit current is typically around 250mA. Even at this level, the power dissipation of the higher voltage parts will be very high, assuming a very "stiff" supply voltage. A 7824 with a 27V supply will dissipate (27*0.25) 6.75W, so the sustained short-circuit current could very well be even less. A 7805 might only dissipate as little as (8*0.25) 2W.
What brand of 78XX's do you have?

 

[rjvh]

So while noone has directly anwsered my question, I assume that it is indeed normal for a 78xx reg to get extremely hot on a short-circuit? Please, a simple yes or no though!

yes it is normal (and not only for 78xx regulator) to get extreemly hot in case of a short circuit

 

[Hero999]

Th short circuit current is normally lower than the maximum current the regulator can supply. When the current draw exceeds about 1.5A, the power dissipation is excessive or it gets too hot, the LM7824 will go in to shut down mode and limit the current to a safe level which can be as low as 500mA.

Just make sure that there's nothing that can be damaged near the heatsinks and they can't be touched by the user.