film cap test

I could've sworn I have some Xicon polystyrenes at 0.1uF in a box but all I find currently available are 0.01 uF:

Have to remember to check what those big fatties are when I get home. Maybe they made them bigger at one time.

What's best these days for low D.A?

Those are great but chemically/physically/thermally fragile. Accidentally touch one with a soldering iron and they die.

Seems like all the work on film caps centers on high temperatures, for obvious reasons--the old faithful mylar (polyester), polypropylene, and (especially) polystyrene dielectrics won't stand SMT reflow. The best of the newer materials is polyphenylene sulfide (PPS). The newer, cheaper alternative is polyethylene naphthalate (PEN), but it has, like,

2 orders of magnitude worse soakage than PPS.

Polystyrene melts at around 100C, but is otherwise about the best dielectric out there--super low soakage, super low leakage, excellent linearity. Unfortunately it isn't out there much anymore.

The only outfit I know of that sells them these days is Xicon, and I've never used theirs. I should try them out--BITD polystyrene caps mostly came in

Teflon is also super good, but is a bit of a cottage industry at this point--there are tales told about the inability of suppliers (of the teflon film stock) to maintain quality.

Polyester is pretty good except for its weird cubic tempco curve.

So for the most part I use polypropylene.

Cheers

Phil Hobbs

1% tolerances at very high prices, but these 5% ones look like they might be OK, and they're sure cheaper.

"Doctor, Doctor, it hurts when I go like this!"

"So don't go like that."

Cheers

Phil Hobbs

Doctor, Doctor, it hurts when I breathe!

Actually, the best non-vacuum caps are SiO2 layers in ICs. That's why they are used in SAR ADCs.

They do tend to give you about 3" worth of lead on each end of the axial polystyrenes. I've never used them above a few 100s of kHz as it's my understanding they're kinda lousy in the MHz range even without the lead inductance. But down lower in e.g. integrator they're basically ideal caps

Big fatties!

A little smaller than a sewing thimble, IIRC. I think the fattest non-electro caps I have in my bin are some 1uF 1000 volt metalized polypropylene, those are getting into could-injure-someone-if-you-threw-it-at-them size.

How about PEEK? I did some quick googling and found thin film suppliers but no capacitors. Should be good mechanically and thermally but I don't know about the things that matter for a capacitor like dielectric constant, D. A., thermal coefficients, etc. PI (Kynar) might also be good, and Wikipedia actually said someone claimed to offer one but no details were available. If anyone wants to try making some I did find one place selling a pack of 15 different films for capacitor R&D that includes PEEK and PI, for a mere $2000 :-).

Dunno. There are a lot of things that you'd expect to make good dielectrics that don't. Glass caps, for instance, have soakage that has to be seen to be believed.

Cheers

Phil Hobbs

Well, the lack of available alternatives has something to do with that too. ;)

Cheers

Phil Hobbs

You've got to pick your dielectric film carefully if you want a linear analog ramp. Charge soak can be a problem for polyester films, and polycarbonate films aren't all that much better. Polypropylene films are quite a bit better, and teflon films are supposed to be even better, though the films aren't as thin, and teflon is quite a bit more expensive as well.

On a sunny day (Sat, 9 Jul 2022 23:26:05 -0400) it happened Phil Hobbs snipped-for-privacy@electrooptical.net wrote in <tadgsd$1eqe$ snipped-for-privacy@gioia.aioe.org>:

Mica? Used mica caps in RF years ago.

This about mica is maybe also interesting:

No point in discussing anything with Bill "Bozo" Sloman, the Australian troll. Bozo is an incessant liar who cannot be reasoned with. Its fiction never ends.

"the user has posted under the same name in other places, so not nym-shifting" (Bozo sucks at logic)

"the Mueller investigation was about Trump only because Trump made it so" (Bozo being Bozo)

"the concepts "male" and "female" are essentially social constructions" (Bozo is a textbook cannibal leftist)

Yup, horrible soakage but great at RF.

Cheers

Phil Hobbs

He's not a conscious liar. But he is in effect delusional.

"The easiest person to lie to is yourself."

Ignore him. Try.

I don't share John Larkin's delusions about the quality of his posts here. He thinks they deserve flattery. I'm unkind enough to notice that they don't.

This doesn't make me "delusional" no matter how much John Larkin would like this to be true.

John Larkin would know. He tells himself that he knows the truth about anthopogenic global warming when he is actually a sucker for half-witted denialist propaganda.

If we could get John Doe to ignore the entire sci.electronics.design user group we'd all be much happier. John Doe doesn't take advice. even John Larkin should have enough sense to have noticed.

See also these John Doe troll nym-shift names: John Doe snipped-for-privacy@message.header John snipped-for-privacy@post.header Judge Dredd snipped-for-privacy@post.header "Edward's Mother" snipped-for-privacy@post.header

How stupid is Troll Doe?

Troll Doe posting one of its vacuous insults at 05:39:20 UTC on 20 Mar

2022 with a grammatical error:

Then, at 05:55:56 UTC, 16 minutes and 36 seconds later, Troll Doe responds to its own post with a correction, but stupidly forgets that it sets a Followup-To: header to the "alt.test.group", resulting in its correction article posting only to "alt.test.group":

Troll Doe, mister " snipped-for-privacy@message.header", is so stupid it does not even remember it sets a Followup-To: header in its own vacuous insults.

Special thanks to corvid snipped-for-privacy@ckb.ird for pointing out the stupidity of Troll Doe:

The Troll Doe stated the following in message-id <sdhn7c$pkp$ snipped-for-privacy@dont-email.me:

And yet, the clueless Troll Doe has continued to post incorrectly formatted USENET articles that are devoid of content (latest example on Sun, 10 Jul 2022 12:14:35 -0000 (UTC) in message-id <taefrb$1cc84$ snipped-for-privacy@dont-email.me).

NOBODY likes the John Doe troll's contentless spam.

This posting is a public service announcement for any google groups readers who happen by to point out that John Doe does not even follow the rules it uses to troll other posters.

z5pNGmevyrwN