Demonstration of an electrolytic capacitor charge (excursus for those interested in electronics ;-)

[Andreas Thaler]

I believed that you meanwhile had finished this experiment?

Best,
Reginald (Digi Club)

I need more expert opinions on this.

 

[Andreas Thaler]

@Bushcat

Thanks for the practical information!

 

[Andreas Thaler]

I regard electrolytics that are over 15 years old to be dangerous because they dry out, some faster than others

If you have an old flash unit, I suggest replacing its electrolytic.

But all flash units that are older than 15 years would then have to be disposed of because replacing the flash capacitors is not worth it or not possible?

I have a large collection of flash units that I activate every few years, not every month as recommended, and they all work without problems.

 

[Bushcat]

Electrolytics are usually rated by operating hours at their rated temperature. Historically, that's been about 5000 hours at 105C or 65C or whatever. Roughly, every 10C reduction in operating temperature doubles the life. The "life" is a blended concept that includes factors like leakage. For something like a consumer flash capacitor, it spends most of its life around ambient temperature, leakage that would be unacceptable in other equipment may not be an issue anyway.

Some electrolytics are simply made better than others. This tends to include flash capacitors (as opposed to standard capacitors, doing duty as flash capacitors).

So: a flash capacitor spends very little time actually on, or anywhere near its rated performance. It's usage cycle is, vaguely close to how one would maintain the dielectric anyway. So I'd wait for it to fail: they tend to last forever, until they don't.

But as albada said, in other gear like audio and computers, I'd replace anything that looked vaguely iffy because it can, indeed, dry out and not perform correctly, or leak and wreck the PCB. I might replace anything off-brand anyway, because of capacitor plague between 1999 and 2007, due to a higher-than-expected failure rate of aluminum electrolytic capacitors from manufacturers who had "acquired" an incorrectly formulated electrolyte from a Japanese company. And the huge amount of fake stuff, of course.

 

[NedL]

Using that capacitor, the best thing in my opinion is a neon bulb oscillator. You already have 3 neon bulbs, just need three 1 Megohm resistors and three small capacitors.

wow that brings back some memories. Agree that's very cool.

 

[koraks]

6.5 kilohm, 20W

A 20W bleed resistor? A normal bleed resistor for a 350V cap would be in the 100k ~ 1Meg range. A 100k bleed resistor at 350V would need to be a 2W or 5W type as it'll dissipate around 1W. Such bleed resistors are extremely common in old tube gear where these voltages are also very common. They are *never* as low impedance as you suggest. Nobody is going to overdimension a power supply massively just to power some bleed resistors. It's putting the horse behind the cart.
For a temporary bleed resistor you only use to discharge a cap before working on it, OK. But a permanent bleeder..nope.

 

[Bushcat]

Well yes, indeed, and covered in the remainder of the post. But positing a fully-charged 350V, 800uF cap which someone wants to discharge in 15 seconds, and in that type of circuit where the bleed resistor is not switched in at power-off, the maths is around 6.5 kilohms, 20W, because current through it is effectively "peak" all the time the circuit is on. As you (and I) say, in the real world, use 500k or 1 meg and discharge over an hour. It's all in the previous post.

 

[koraks]

Yep, sorry; missed that - you're right!