Small explosion, did I do something wrong or was it defective?

[SeanL]

I actually managed to make it through 4 years of ECE without being told about proper grounding of an oscilloscope... Maybe because the ones in the lab were all isolated so there was no need for when we were in school. But still, a quick lecture would have been really nice. I always assumed they were isolated and learned the hard way recently at my first job! They also failed to give a lecture on ground loops. Two BASIC things that would not even take a whole 30 minutes to lecture on...

Once I had a little camp fire at my desk. An Assembly house got some smt electrolytic caps reversed and I am just glad I was able to blow it out. Another time a benchtop supply (new, btw) suddenly went off like someone lit 3 or 4 firecrackers and the factory smoke came out. After that I always wear my goggles on the first power test! When soldering, too. Molten solder flicks off of a stiff wire pretty easily...

 

[Sceadwian]

They didn't tell you about ground loops? Jeesh, that's negligent from an ECE Course =\

 

[Triode]

**broken link removed**

I know, it took me a while

 

[Hero999]

It IS odd since the 78xx series should have thermal and current limiting built in.

True, but it doesn't offer any protection against other silly things such as reverse polarity or a voltage injected into the Vout pin.

 

[Triode]

yeah, this was a short circuit, so it probably sent some current where it shouldn't have. Well, not technically a short circuit, but an unintended circuit.

 

[hv addict]

I got your back on this one Triode!

If you want to see and hear a good IC explosion you should be around when a 1200 volt 200 amp IGBT half bridge block gives up on life!

Your regulator IC gave up at a few watts, Power IGBT's give up at a few 10's of kilowatts!

haha i hear you on this one. see what happens when you put 5000A in a 500us pulse across a cm300 1200v 300A igbt for 1/2 hour straight Anyway, to answer your question, i have actually had a couple of linear regulators that actually exploded in my face but i found a similarity in all of them. they were all connected to a very large filter capacitor and i noticed that when they failed they went short circuit to the ground pin and the capacitor unleashed it's fury into the regulator and blew it to smithereens.

 

[Leftyretro]

haha i hear you on this one. see what happens when you put 5000A in a 500us pulse across a cm300 1200v 300A igbt for 1/2 hour straight Anyway, to answer your question, i have actually had a couple of linear regulators that actually exploded in my face but i found a similarity in all of them. they were all connected to a very large filter capacitor and i noticed that when they failed they went short circuit to the ground pin and the capacitor unleashed it's fury into the regulator and blew it to smithereens.

Actually the data sheet for many linear regulators discuss this problem of voltage on the output being higher then the input during powering off, and the damage that can result. The solution generally shown is a reverse diode installed across the regulators input to output terminals. This gives a alternate path for the discharge current from the output side of the regulator.

Lefty

 

[Speakerguy]

Yep, I know the LM317 data sheet says you need a diode if you exceed 25uF on the output. It also recommends a diode for the adjust pin bypass cap if one is used and it exceeds 10uF. Since I regularly exceed both those values it's a pain.

 

[Hero999]

It also depends on what's connected to the input.

If the input power source won't accept power in the reverse direction then the regulator is safe. I have an LM317 connected to a solar panel, with a built-in diode, charging 14Ah lead acid batteries. There's no diode on across the LM317 or on the output and it doesn't spontaneously self destruct every time the sun goes down.