Electrolytic caps and heat

[Screech]

Electrolytic caps,They have liquid in them, yeah?

So does placing them near hot heatsinks or inside a warm/hot incloser make them hot and the liquid leak out?

In some cataloges I've sean caps with teperature ratings and only a few thousand hours of life (probably explanes why my 300 watt soundstream amps only lasted a few thousand hours. You could fry an egg on it).

How long will they last in a warm to hot incloser?

All other types of caps are fine in hot teperatures?

I want to use upto 1000uF caps , but they only come in electros.
Are they no good in hot enviroments? What do I use then?

what the hell is the capacitors ripple current?
i got lots of caps in my circuit.
will more then 1000 milliamps of ripple current, destroy 1 amp regulators, diodes etc?

 

[Nigel Goodwin]

For a start you should use higher temperature electrolytics, standard ones are only rated at 85 degrees C, but 105 degree ones are now commonplace.

As with most things, the hotter you run them, the shorter their life!. You should design to keep things as cool as possible!.

 

[Styx]

With electrolytic caps and temerature two things need to be watched.

1) you dont cook the electrolytic and fook it up, ie take a 85C cap to 120C they will oooz and smell

2) WATCH THE VOLTAGE RATING!!!!!
the capacitance stays pretty stable (within the tolerance) BUT the voltage rating comes down!!!. I have some 35V electrolytic and at 90C their rating down to 25V, since it is on a 15V rail that headroom is fine!!

you dont watch the de-rating of voltage and well BOOM!!!!!!


Equally the ripple-current rating will come down. Since current will heat the cap via its ESR.



Now as to what ripple-current is. Think of a diode-bridge with a resistive load and a capacitor TANK for smoothing. There will be some DC-link ripple that follows the 50Hz from the mains.

That ripple as far as the capacitor is concerned is charge building up its voltage and charge decreasing it voltage, ie current-flow, which in turn it a ripple == current ripple


The current-ripple must not exceed the rating of a cap - overcurrent caps are worse then over-voltaging caps!!! BOOOOOOM.


We have banks of electrolytics closing on 10mF in a DC-link for a 150kW generator, it is sitting in a ambient of abt 60C and they are fine, rated at 600V for a 350V link room-temperature


As to your question

what the hell is the capacitors ripple current?
i got lots of caps in my circuit.
will more then 1000 milliamps of ripple current, destroy 1 amp regulators, diodes etc?

only IF they are in series with the capacitor - very unlikely so no they will not see this current-ripple as the capacitor see's it, they will see some but due to the voltage ripple w.r.t. the load impedance

 

[Pyroandrew]

Just to clarify, Electolytic caps do not have liquid in them. the liquid that oozes out is the dielectric that has melted. Has enyone placed an electolytic in reverse bias. It goes bang after a while and spits out horrible brown smelly liquid. what happens is the cap heats up and when the dielectric melts the metalic foil comes into contact and causes a short, that is what causes the bang. Similarly, in normal bias, if you heat the cap up, i.e. by putting it next to a heatsink, the same thing happens.

 

[audioguru]

Just to clarify, Electrolytic caps do not have liquid in them.

Well, I think it is more like a jelly than a liquid, but it is still wet. Manufacturers of regulators (National's LM2931) say that when the electrolyte freezes then its capacity is reduced.

 

[Screech]

Just to clarify, Electrolytic caps do not have liquid in them.

Well, I think it is more like a jelly than a liquid, but it is still wet. Manufacturers of regulators (National's LM2931) say that when the electrolyte freezes then its capacity is reduced.

I've found that the isulative paper inside the cap is damp just like the moist towlets from KFC.


Thaks for your lenghtly answers Styx.

Thanks all.