PSU capacitor max.voltage question

[dhaen]

Unlike most other capacitors, electrolytic caps' voltage limit is about heating.
Each 'lytic has a leakage current that rises dramatically beyond it's voltage rating.
In times when big 'lytics were expected to be installed in valve equipment, the manufacturers gave 2 voltage ratings: Normal and Surge.
Normal was for continuous operation, and at this voltage, leakage (and therefore heating) was in the safe area.
Surge was for the period of time when the valves were heating up.
It was known that this time would not exceed 1 minute, therefore the rise in internal temperature known.

It would not be much work to test the leakage of a particular cap (in an enclosed place ), and therefore the temperature rise. This would make for assurance about long term reliability.

Warning: There are safety issues with electrolytic caps. Apart from the high voltages we deal with on this forum, missused caps can explode. Large caps should always be installed so they would blow into the equipment - not outwards. It's also important to allow space for the cap to blow. No use giving it a 5mm or 1/4" hole to blow through - it'll launch like a Saturn rocket.

[mad_z]

Not going to buid a racket - rather safe&reliable PSU - so I will rather go for 450v and 400v caps.

Thx for help!!!

Cheers

[Jennice]

Dhaen,
True enough, I don't want to hold an exploding electrolytic in my hands either (I guess I'd miss my hand if I ever tried it!)


Badman:
As for the best operating voltage of a capacitor, it is true that it stores less energy when operated lower than rated, but what makes them perform "worse" if operated below rated voltage? After all, the stored energy depends on the voltage and capacity; not on the rating alone. Obviously, a 200V / 1000uF can store more energy than a 20 V / 1000uF, but that should be due to the safe operating voltage (rating), as I see it.
However, the ripple voltage when used here as a PSU "storage" capacitor should depend on the ESR & ESL, and the current drawn from a given capacitor, regardles of it's "operating voltage" / "max voltage" ratio.


fdegrove & Per-anders:
As I wrote in my post (if you'd care to read before being rude, P.A.!!) I never worked with this gear, and since operating voltages are much higher than solid-state gear, the ripple voltages will have higher values too.


Jennice

[peranders]

Originally posted by fdegrove
I'd suggest you guys fight Viking batlles outside the classroom/forum?

Per-Anders, I know what you mean, but do we have to bite other people's head of?

Ripple voltage peaks will be much higher there than after the inductor.

Frank I missed the word "than" so I missunderstood the whole thing.

Of course ripple voltage will be more before the inductor but still, 260 VAC will give you 367 V and the amp must take at least 5% overvoltage, (better 10%) which gives you 386, or 403V.

I'd say that 400 V is close and to choose 400 V cap you must "know" how good it is. If you don't know anything I think the value _may_ be a bit low.

EDIT: Jennice, sorry I didn't mean to be rude.

EDIT2: Safety This is even more important at high voltage, as pointed out above.

[mad_z]

That's why(and to be sure) I'm planning to use configuration as posted below:
450V cap before choke, 2x400V cap betwen choke and resistor and 450V choke at end of chain.

Would this solve any possible problems of voltage-overload on caps in such PSU?

[fdegrove]

Hi,

Would this solve any possible problems of voltage-overload on caps in such PSU?

Those problems are not going to occur under any normal operating conditions.

If you change the cap(s) after the choke, you should also change the last one to 400V.

What you have is a CLCRC filter PS and with the full wave bridge ahead of it the maximum voltage you have there will be:

260 * 1.4 = 364 V.

So, keeping the 450V cap as the first filter cap is wise, after that 400V caps are best as it maximises capacitance and it still leave you with a fair safety margin under no- load conditions.

Hope that clarifies it,

[mad_z]

fdegrove - Thanks!!!

That's all my newbie-head wanted to know before assembling all that stuff altogether

Chers,

[badman]

Originally posted by mad_z
I don't have data of this power transformer but secondary for B+ are at about 250-260V.

Any help?

Oops! I should have reviewed the given info more carefully- I just saw 250-260V B+ instead of seeing that the secondary would be that...

Originally posted by fdegrove
Hi,



Yes...But what about safety margins?

Now do your sims at what cap values really are at operating voltages and feel as smug as you think you are.

Feeling kinda Freddish,

I don't recall being smug. If I were, I wouldn't point out my own goof. Sorry to be unclear MadZ, 400V caps are fully appropo. I was looking at is as if you were using little more than half your available energy storage on the caps (260/450) which would be a silly way to do it. If you really wanna go nuts.... try all poly caps for the PS.... Solen makes some that are appropo, and if you can deal with the funny layout of axial parts, www.apexjr.com has some lovelies VERY cheap.

[fdegrove]

Hi,

I don't recall being smug. If I were, I wouldn't point out my own goof.

Badman,

Don't take it personal, it wasn't directed at anyone in particular...

It just amazes me from time to time how such a simple question can end up so misread, hence my little jest about it.

Going polyprop all the way would be nice, there's the Solens and also the ASC which I personally prefer.
Either way it's not exactly cheap but well worth it IMHO.

Cheers,

[mad_z]

Polypropylene - you mean sth like this in PSU would be OK?