At what temp do voltage regulators shut down?

Thread Starter

daviddeakin

Joined Aug 6, 2009
207
Most voltage regulator ICs have built-in thermal protection. But does anyone know at what junction temperature a device like the LM317 actually begins to shut down?

The safe operating temperature is usually stated as 125°C, but my own measurements of heatsinks suggests the shutdown happens at a much higher junction temperature. Are there any semiconductor designers around here who can answer this?
 

crutschow

Joined Mar 14, 2008
34,285
According to this (page 4), the thermal design shutdown temperature is 150°C. So it would be advisable to not operate in the thermal shutdown mode for long periods if you want best reliability.
 

#12

Joined Nov 30, 2010
18,224
it would be advisable to not operate in the thermal shutdown mode for long periods if you want best reliability.
True. However, I did see a garage door safety beam (light) using a 78L05 chip in current limit mode. It seemed to be as hot as a match head! That is my opinion of bad design.
 

alfacliff

Joined Dec 13, 2013
2,458
and a few brands of coin op "crane" games with a 5 volt regulator with 35 volts on their input and a very small piece of circuit board for a heatsink. they fail quite regularly till corrected.
 

Thread Starter

daviddeakin

Joined Aug 6, 2009
207
Thanks for the link.
I allowed a 317T to go into shut down (no heatsink), which happened at 2.18W. The datasheet claimed a junction-to-air thermal resistance of 50°C/W, so I estimated the shut down happened at 50*2.18 + 25 = 134°C. Not out of the realm of plausibility.
 

#12

Joined Nov 30, 2010
18,224
Some of them seem to think the high limit for die temperature is a perfectly valid operating condition. Very literal translation: You wrote that number in the datasheet, so I can use it for my design criteria!

Or the engineer that read an LM317 would limit its current when the output is overloaded..so he thought the LM317 would limit the current in the (helper) pass transistor. That's when I learned that thermoplastic rectifier diodes make a crackling noise when they explode. :D
 
Top