78XX series voltage regulator - NP capacitors

Refreshing an earlier thread on this topic, I have a new experience.

Returning to electronics in retirement, I was faced with repairing a mixing panel at a local community radio station. The fault was hum on the audio. Replacing the main p/supply electro caps fixed most of the problem. (A 7815 & 7915 delivers the +/- for the op amp chips.) I then fitted an electro cap across the minus rail at the chip itself, and a NP cap across the + rail. Both 100 uF/35v. Turned on, and bam! The 7815 blew apart. The cause was the NP cap I'd just installed (because I had one in the junk box). Replacing the 7815 and removing the NP cap thankfully restored normal operations. All good and quiet.

Question - what is it about NP caps that caused this dramatic end to the 7815? I thought NP caps were a universal replacement for traditional electrolytics?

Negative on that.
NP caps are ok for decoupling in small values but not for power supply filtering.
I can't give you the theory on it but I've never seen NP caps used on LM78XX regulators for
the output filter cap. I've only seen electrolytic.
There may not be a reason why it blew but from an electronics 101 standpoint , the only logical
explanation is that the cap was shorted and you didn't know it when you installed it.
I would bet almost any amount of money that you didn't test every single cap with a capacitance meter
before installing them. ( I only do this on high priority prototypes at work if they have a cap meter)
If there is no room for failure on a project you pull out all the stops and test everything .
Obviously nobody does that for home hobbyist projects and very few people ever test any caps, period.
Technically the LM7815 should survive with direct wire short from the output to ground . It should get hot
and shutdown. It may be just a coincidence and the NP cap didn't cause it.
Are you really ready to try to reproduce the issue by removing the electrolytic and putting an NP cap in
there to see if you can blow it up again ?
This may sound nuts to you but when I really want to get to the truth that's how far I'll go.

The capacitor was probably defective.

But, the only place I've seen nonpolarized electrolytics used in in speaker crossovers. ...If you don't need a nonpolarized capacitor I'd only use one in an emergency.

DVDdoug:
The capacitor was probably defective.

Doesn't make sense!

If the 7815 blew (not the capacitor), then it was either faulty, interested backwards (but that should have been obvious) or a 7915!

All a capacitor could do would be to short the output - the regulator should current and heat limit.

I always put NP caps on both input and output of regulators along with bulk capacity. Reason is for high frequency bypass. Not all regulators with the same number are the same, you need to check the data sheet to be sure. I also put a diode across the regulator so if the power is interrupted it will discharge any capacitors on the board into the power source, and not through the regulator, this tends to damage them. This was a real problem years ago when they were made with an older technology. This response is to help you get started in solving your problem, not solve it for you.
Good Luck & Have Fun!
Gil

Working NP caps will not damage the regulator. (assuming the capacitor is not a super cap or very large value)

I'm assuming by NP you mean non-polarized, if so, where did you find a 100µF /35V? was it a film capacitor?

So what might have happened..... My best guess is the NP capacitor was faulty (shorted) and the input voltage of the regulator was on the high side. I realize that in theory the 7815 is thermally protected and short circuit protected.

Thanks everyone, good ideas all.

The non-polarised capacitor came from my local electronic hobby shop (Jaycar), so a regular item. To be honest I picked up a selection of caps in a hurry and meant to get all electros. But thought "what the hell" and installed it. When I say blew the 7815, it literally blew half of the plastic body away, with sparks.

Reading the reply thread, I checked on the cap and sure enough it's low resistance. Not dead short, and it's difficult to tell the exact resistance because the effect of the meter circuit. But it seems likely that it's defective under powered conditions. The input side to the regulator is 35v, so quite a punch and maybe beyond the ability of the 7815 to thermally protect itself (despite being heat sunk).

Reading more about NP caps, I think in theory they'd be fine to use around a 78XX, but in practice I think I'll be sticking with electros or tantalums.

Bottom line - replaced the 7815, fitted small electros down the rails near the op amp chips and it's quiet as a mouse. All hum gone.

Oh, and a PS - yes, not all 78XX are made equal. Thanks for the tip about checking the data sheets!

The overload and thermal protection circuits will
work correctly once the 78XX is powered up. But,
if power is applied with the fault condition already
in place there may be no guarantees!
Herb