The SRF for similar constructions will depend mainly on lead spacing and body size. The material limits the dip ~ ESR.
If you can get ahold of an old Philips data book, this trend becomes clear, as charts can be more easily compared. A collection of Panasonic pdfs may serve the same purpose.
RL
If it conducts.. It's got inductance.
I see.... I don't know!
I would guess it is all OK as long as you keep the temperature well below the maximum rating. And this will be determined mostly by the copper area rather than packaging as usual. But who knows? If the DF increases with temperature perhaps there is a safe operating area style limitation too, you could get hotspots that suddenly start to dissapate more heat and burn up.
The nice thing with the ceramics I've looked at when it was important is that the ESR tempco is negative. Never found one positive.
What about DF?
The problem I found with polyester is exactly that... it heats up a little, which heats up more, pretty soon the dielectric decomposes and becomes resistive, etc... Good way to burn a hole in the things. Polypropylene doesn't do this, probably something to do with it not containing oxygen (as ester groups, which are slightly polar).
Maybe it would only happen to the ferroelectrics like X7R? I'd already skip those because they're too lossy for something like this. C0G is just about indestructible though... supposedly it's basically pure TiO2? With no Curie temperature to pass, they could keep on running and not care...
Tim
"Tim Williams" kirjoitti viestissä:hffskl$1c0$ snipped-for-privacy@news.eternal-september.org...
Taiyo Yuden lists ripple vs temp rise at few frequencies For 1206 X7R 330nF they say about 1 amp at 100 kHz and 2 amps at 1 MHz at dT
20 C.Now you have wrong attitude towards Z5U and likes:) Y5V tolerates quite big ripple currents. Also temperature/capacitance graph looks so steep that they may actually self-regulate around
70 C or so. ie. -80% at 90 C. Never tested that thought. Year ago I built 10 amp 24/12 V switcher with some 10 uF Y5V at output. 100 kHz and few volts of ripple didn't heat them at all. No clue about current, but 330 uF low esr electrolytic got hot if but there Those capacitors are pretty hard to solder with iron. Touch with iron-> quiet "CLICK" sound from cap -> shorted. Designed it with forty caps but gived up after fourth or so... Thats why the ripple. Luckily second LC filter suppressed it quite nicely.-ek
Hmm, well here's a typical part by TY
Heh, but that assumes the current has somewhere to go. I want to use them for resonance, so the only thing that will change is the frequency. Impedance rises some so V goes up a little and I goes down a little, but most of the I is still there. It's not actually a big deal that capacitance and therefore frequency changes, but if it rises by 1/sqrt(10%) = 3.2 times, that's an awful lot.
That's an awful lot even for a switching power supply. You have to do nasty things to your feedback loop to keep things stable under those conditions.
A few volts? Pffbt, I want a few hundred. ;-)
Tim
You can probably pick one up for USD 5k. They were USD 20k new. You'll need the test fixtures which can add up. I find the tweezer probe to be the most useful.
"Tim Williams" kirjoitti viestissä:hfm5lt$hmn$ snipped-for-privacy@news.eternal-september.org...
Hmm. Apparently it is only on their web page
-ek
Ah, that's handy. Too bad it doesn't also show mounting (infinite 1 oz. PCB??). Well, there's "The Application Soldering Method: R", but I don't know what that is.
So let's see, at 100kHz, ESR is like 1.2 ohms, so 0.2A would drop about
0.24V and dissipate around 0.29W. 20C temp rise implies 69.4 C/W thermal resistance. So I guess 1/4W isn't a bad guess after all. If better heatsinking (infinite 2 oz. board maybe?) and more temp rise is acceptable, I could run at double the frequency = half the ESR = sqrt(2) more current, and double the temp rise = sqrt(2) more current, for a total of 0.4A each. Let's see, I was looking at about 1uF, so that'll be a hundred, or 40A total... nahh, still not enough. Still needs about three times more current, which would be nine times more dissipation = lots of melted ceramic.Tim
factor=20
Volts?=20
Nobody is stopping you from putting on your thinking cap. =20 Where does the heat come from? What is the failure mode?
Have something to add? Share your thoughts — no account required.
Ask the community — no account required