Capacitors at RF

So, "correlations" with what?

Different types of RF capacitors are aimed at different applications, so they /will/ be specified at different frequencies.

I'm using SMDs to minimise that issue. And I made up a measurement bridge to place the component as close to port 1 as possible.

True, but they developed work-arounds such as neutralising a stray inductance with a 'matching' capacitance and so forth. The strays and parasitics in the passives of 70 years ago were huge by modern standards, but managable given the low frequencies they used back then.

Try searching uk.d-i-y and you'll see he's still active under that 'original' name and a great many sock puppets in addition. His favourite response to everyone is "Even sillier than you usually manage...." Still, credit where it's due, at least he's not a foul-mouthed troglodyte like Phil. Just a massive nuisance and disruptive influence.

Classic capacitance bridges might work at 10 KHz, roughly. My wonderful Boonton analog c-meters are mostly 100 KHz, but I think there was a 1 MHz option.

High frequencies could have problems with part or fixture inductance.

That's a reactance of 170 uohms. 5 nH lead inductance would be more than that. Makes it tricky.

I want to know what a

Anything above maybe 10 KHz would be fine.

You can measure femtofarads or attofarads with gear that you probably have.

Spinning around backwards on your head, surrounded by weird poisonous critters, has that effect. Being descended from criminals and drinking Aussie beer compounds the brain damage.

That's a bad sign. The graph of measured c vs frequency should be flat to identify the capacitive component of the impedance. There should be a plateau between seeing shunt resistance (usually very low frequency) and the frequency where series R and L start to matter.

Measure that cap some other way as a sanity check.

I sometimes connect a 50 ohm sine function generator and a scope across an inductor or a capacitor to measure impedance. But I sweep the frequency to make sure I'm on a good slope region.

Square wave from the fungen is cool too.

TDR is fun for small caps too. You can usually see C, esr, and esl.

If the measured c value depends much on test frequency, something is suspicious.

I could have been clearer here. What I was referring to was ESR. ESR in caps is measured at a generally agreed frequency of 100kHz.

Er, nope! I think you have way too much (misplaced) confidence in me on that one, John! :-D I do know my limitations...

Femtofarads are easily measured.

Sine wave generator > cap in series > hi-Z scope input

The cap and the scope capacitance form a voltage divider. 1000:1 should be easily resolvable, and 15 pF/1000 is 15 fF. Signal averaging from an external trigger helps.

Add a fet probe or a fet opamp to see attofarads.

Phil is a long-time poster on sci.electronics.repair where he generally abuses people on the only subject he knows anything about: servicing audio equipment. Quite why he felt competent enough to contribute to a thread about RF is a mystery.

It is relevant, though. Most people here will be familiar with the fact that caps start behaving like inductors beyond their SRF, so one shouldn't be surprised that their measured values change with frequency.

That's why they're still sought-after collector's items. And increasingly valuable. I have a Boonton GDM for detecting resonant frequencies and it's now quite rare and hard to find with values accordingly. It still works fine and is *very* sensitive in the best possible way.

One can model a cap as pure elements: capacitance, shunt resistance, esr, esl, and maybe dielectric absorption bits. It's useful to me to separate and measure each. Once they get tangled, it's messy.

The c element is believable if impedance is inverse on frequency across a wide frequency range. Pick a point in the middle.

Draw a graph of z vs f and imagine the various zones.

He's almost always angry and foul and abusive. He must not be very happy.

Ditto Sloman. An endless stream of ritual insults is not a sign of good cheer, or good health.

And being ignored. No one looks to Australia for a lead in any field, so there are only 2 ways an Aussie can get noticed: excelling in sport or else being obnoxious to everyone else. They can't stand being ignored so if one's no good at sport, then throwing their own metaphorical faeces at others is the only remaining option to get noticed.

For *you* perhaps!

What B/W would the scope need to be? And what sine wave frequency?

Indeed. They're obviously deeply unhappy people. I'd guess Phil had had a few too many Fosters come bed time. Bill OTOH doesn't have alcoholism for an excuse for the crap he posts, though. It just comes naturally.... ;-)