bipolar aluminum caps

Never used them... I made my own years ago by sticking two Al electro's in series. I'm guessing they would stink at high frequency. George H.

Tants make good bipolars ;)

NT

But then they tend to lose the magic smoke.

For a while. There's engineering lore that dry tants are OK at -10% of rated voltage, or something.

I didn't know that. Long ago I stuffed a regular lytic wrong and it worked fine... at -23v on a 16v cap. I don't think I'd use that in production though :)

NT

Talk to Murata. IME they have among the best "figure of merit" caps when it comes to high relative dieletrci constant and capacitance loss over applies voltage. Their TechHelp folks are quite knowledgeable.

I don't trust electrolytics much and only use them if I really have to. Especially on hi-rel design.

Passive power filters for audio were a major market but with the advent of active speakers that segment has shriveled up. Many people don't even have a stereo with huge speakers anymore. Other than some diehard boomers.

I don't know dry from wet tant's, but I had a 35 V (4.7uF) through hole tant last for years (intermittent use) when put in backwards in 5V regulator.

George H.

My dim recollection is that a bipolar electrolytic is equivalent to back-to-back polarized electrolytics. That is, you get twice the ESR & ESL, and some weirdness around 0-volts. I'd be nervous if the current was appreciable. However they've been used for many years in lots of applications, mostly in moderate-level coupling applications where distortion wasn't an issue. Probably as reliable as a standard electrolytic.

I guess I could do something like this...

---------+

I'd still add a diode across each cap. That -20/+80% spec range could make a bit of a mess otherwise.

Cheers

Phil Hobbs

If R=0, the top cap is just low-sided to the -6 supply and the bottom cap bypasses that, which I should do anyhow.

Do people still say "duh" ?

A long time ago, just after a shake test, we found out that high-C ceramic materials are either piezo electric, or flexible enough to change capacitance with vibration. Either way, they made dandy microphones.

Since the circuit in question had DC paths to ground on both sides of the cap, we replaced it with a pair of back-to-back tantalums with the far side biased through a high resistance to some negative voltage (probably

-12V) that was already on the board. It worked pretty good for a panic- driven kludge.

Sumpin' to think about.

How much capacitance do you need?

Don't use them much, but 25 years ago, the nonpolar capacitor for vertical sweep in a MacPlus was a major replacement item. There were film-capacitor replacement parts, with molded axial-to-radial fixtures.

Tens of uF at 20 volts maybe. There's a lot of current drive available, so a tantalum would be risky.

X7Rs are available in that sort of capacitance/voltage in 1210s. They're a bit pricey but some aren't that bad (C/V wise).

If you have to ask its price, and you get an answer: it's a dry tant.

Dry tants are usually rated for "5% or 3V, whichever is higher [in magnitude]", reverse. More than that is definitely your own gamble.

Don't know about wet, but I can't say I like the idea of boiling sulfuric acid gel splattered in my face!

Aluminum electrolytics tend to put up with quite a lot of abuse (including bipolar operation*, gasp!), as long as it's done slowly (so the oxide layer can reform).

Electrolytics,m over limit, tend to fail with a time constant of ~months. YMMV.

Tim

An excellently valid solution, and note that tantalums are precise enough (~10%) to avoid the problem mentioned a few posts above (reversal for mismatched caps in series).

Electrolytic caps (chemically speaking, this includes wet and dry tants) are basically semiconductor* junction diodes with a fuckoff huge junction area. Reverse bias is a natural win for them!

Tim

Or aluminum polymer (which, as with others: two in anti-series, with bias).

If you needed energy storage, C0G would be great. It's just preposterously expensive, in any large value...

Tim