why did my regulator blow up

[amateur_25]

Thermal limiting if pushed can the regulator blow up and produce a spark across the terminals
or have shorted somthing out? Correct me if wrong the 7805cv has both thermal and output protection?

[Psi]

Thermal limiting if pushed can the regulator blow up and produce a spark across the terminals
or have shorted somthing out? Correct me if wrong the 7805cv has both thermal and output protection?

The 78xx/79xx series has both current and thermal limiting but you can still blow them up.

One way to kill them is to push their output higher than the input (using some other power source).
Also an overly large capacitor on the output can do it too, as it will feed backwards when the input power is turned off.

[Mechatrommer]

Also an overly large capacitor on the output can do it too, as it will feed backwards when the input power is turned off.

i think thats why datasheet recommends the reverse diode between output and input and its drawn as BIG diode.

[amspire]

Thermal limiting if pushed can the regulator blow up and produce a spark across the terminals
or have shorted something out? Correct me if wrong the 7805cv has both thermal and output protection?

Also the thermal limiting in these regulators is only designed for allowing the regulator to survive a short term overload. It is not designed for regular use, as when it is in thermal limiting, the junction temperature can be 10 degrees above the absolute maximum junction temperature. I wouldn't bank on a 7805 lasting too long if it was regularly taken into thermal limiting.

If you wanted a reliable design, and say you wanted to get 1A from the 7805, the maximum input voltage I would allow for a TO220 package device would be 13V if it was on a big external heatsink or 8V if it was on a PCB heatsink. If there is no heatsink, you cannot run a 7805 reliably at 1A output. 400mA max with 8V input perhaps.

If I was regularly going to overload or short out the 7805, I would add an extra current limit circuit before the regulator, or use a different regulator with a proper current limit.

Richard.

[amateur_25]

I've always used 100uF on the input and 10uF on output. is this ok?

Could you explain what u mean by proper protection?

[Ajahn Lambda]

I've always used 100uF on the input and 10uF on output. is this ok?

Those are fine values for general applications, yes.  However,  I would probably want to look at the voltages on a 'scope, on both sides of the regulator just to be sure.  Just my experience, but I've had some dirrrrrty input voltages cause stupid problems.


Do you have a schematic of the circuit you're trying to supply?  Or at least, could you describe it?

Could you explain what u mean by proper protection?

Condoms.      j/k


Basic circuit protection should always be used, e.g. circuit breakers, fuses, PTCs (poly-fuses), etc., simply to prevent secondary effects, such as physical damage and fire.


The other commonly-implemented protection scheme, specific to this type of device, has been mentioned, namely (reverse-biased) diodes.  I'm attaching a schematic (from an LM317 data sheet) that shows how they're to be connected in the circuit.  Be sure to get the polarity correct.


EDIT:  I just noticed that the part number for D₂ [1N4002] overlaps the C_O connection to ground.  Please disregard that; the D₂ part number should have been located ABOVE the diode for better clarity.  Silly co-op students. 

[amyk]

You actually got a spark across the terminals?

Do you have a huge inductor somewhere in your circuit?

[Rerouter]

yeah i am thinking the same thing, you wont get a spark between 2 terminals unless:
1. there are 2 solder dags so close to touching that a swift movement makes them touch
2. or you had a large inductor in what you where powering or in your circuit and when it hit some limit as the voltage dropped the inductor gave it one hell of a belting,

[amspire]

From the original post, it sounded like the regulator was cooked, it shorted and then the short probably blew out making a flash near the terminals.

That is why my earlier comment was centered on the thermal aspects of using the 7805, and not the load capacitance problems.

Richard.

[AlphZeta]

Shorting the input and the output pins (e.g. having the protection diode backwards) can also damage voltage regulators as 78xx can't sink current.

[amateur_25]

From the original post, it sounded like the regulator was cooked, it shorted and then the short probably blew out making a flash near the terminals.

That is why my earlier comment was centered on the thermal aspects of using the 7805, and not the load capacitance problems.

Richard.

God damn it I was afraid of that.  I think that what happend.Because the transformer was screeching for abit then it went. Then all a sudden the regulator blew up. It even cracked the die of the atmega328p. It definately wasn't there before because that mega was only just over a week old. I brought from rs brand new. That is quite a spectecular blow out.

[T4P]

From the original post, it sounded like the regulator was cooked, it shorted and then the short probably blew out making a flash near the terminals.

That is why my earlier comment was centered on the thermal aspects of using the 7805, and not the load capacitance problems.

Richard.

God damn it I was afraid of that.  I think that what happend.Because the transformer was screeching for abit then it went. Then all a sudden the regulator blew up. It even cracked the die of the atmega328p. It definately wasn't there before because that mega was only just over a week old. I brought from rs brand new. That is quite a spectecular blow out.

Pictures , that would make it all spectacular .

[amspire]

A good rough guide is that whenever a component is getting too hot to touch, it is probably getting too hot for reliable operation. At very least, you need to check the data sheets and the circuit currents to see of the part is overloaded.

Also anytime you start smelling anything, something is too hot - turn  off straight away and find the hot part. There was probably a smell from the 150 deg C regulator before it blew.

When silicon gets too hot, the semiconductor silicon turns into a lump of plan metal, and you have a dead short.

If the power to the circuit if high enough (like you have a unfused Mains transformer), something will eventually blow. In your case, it sounded like a bonding wire to the silicon blew like a fuse.

Unless the load on a regulator is down in the tens of mA , it probably needs a heatsink.

Richard.

[IanB]

Because the transformer was screeching for abit then it went.

If you hear funny or abnormal noises, stop or unplug the power immediately. Not so long ago I accidentally made a short in the circuit I was testing. As soon as I did it the transformer changed to a different sound and I knew something was wrong. It took me less than half a second to disconnect the wire...

[w2aew]

A schematic of your circuit surrounding the regulator will help us determine if there is a design problem that lead to the regulator and uC demise...