Crapacitors

that is pretty bad, but not a murata issue, vishas has a short doc on MLCCs

Yep, it's a matter of high dielectric constant materials. Standard for Z5U[*], unexpected for X7R-types. The subject carpacitor(tm) was X5R, but looks more like Z5U than X7R.

I'm looking at a few mfrs' offerings to see if anyone's dielectric is better...

(*)(e.g.

It's like a tantalum cap rated for X volts, with recommendation to never use it at X volts.

(I actually use tantalum caps at rated voltage *if* there's not much charging current available. Otherwise, X/3 is about right.)

Yuck. Graph straight off the ceramic's D-E curve (compare: B-H curve), the derivative that is. Only thing they don't show is hysteresis, which is bound to be comparable.

They need to start putting some air gap into those suckers!

Tim

-- Deep Friar: a very philos> > 4.7uF 6.3V X5R

Yep, it's a matter of high dielectric constant materials. Standard for Z5U[*], unexpected for X7R-types. The subject carpacitor(tm) was X5R, but looks more like Z5U than X7R.

I'm looking at a few mfrs' offerings to see if anyone's dielectric is better...

(*)(e.g.

-- Cheers, James Arthur

What we're talking about is whether you can, in real life situations, actually use an X volt rated cap at X volts.

Have you appointed yourself to be a net-cop on-topic thread-drift enforcer, or do you just try to find a reason to whine about anything that I post?

Got anything useful to say about capacitors? Are you even interested in capacitors?

Geez, you are a crabby old git.

df

pdf)

Well as you know the X5/7 spec is just the tempco. Anything goes with the V spec I guess. buyer beware and all that....

Thanks for the 'heads up'

George H.

Now when i tried to look at the PDF, my Adobe 9.0 wanted an "upgrade" by adding Japanese font. I said no, and got essentially blank pages. So, i retried, said yes and everything was there. Drawings, graphs, etc have NOTHING to do with Japanese, and only English was used. WTF??????????????? Crapadobe.

I've got this audio amp that runs off 15 volts. But I've told people you can stick upto 40V into it. (as long as it doesn't over heat.) The IC's are good to 60V (I think, LM675?) but I've only got 50V tants as bypass C's. I should do a mod to 100V tants. Someone will want more V.

George H.

.pdf

Yup, the X/?/? notation mostly tells you Class I, II, or III dielectric. I thought perhaps different mfrs would have different secret sauces. If so, I haven't found it yet.

Pg. 9, figure 4 of the Yageo link is frightening. A 50V Y5V has 90% C- loss at 20WVDC.

Getting a MUCH-higher woltage-rated part helps preserve capacitance a smidgen. Using a physically larger part helps a lot.

Sure. I thought Joerg might be interested for his transducer-cap selection. He'd want high capacitance, low-voltage, and small package. That should maximize everything we normally hate.

CAPACITANCE LOSS vs. DC VOLTAGE, 0603, 4.7uF, mfr=TDK

DC-Bias/V X5R, 35V JB, 6.3V X5R, 10V

--------- -------- -------- -------- 0V 0.0% 0.0% 0.0% 1.25 -3.7 -4.4 -16.2 2 -10.2 -11.6 -38.2 2.5 -16.6 -23.0 3.15 -24.5 -31.5 -58.8 4 -35.4 -41.4 -68.0 5 -45.9 -54.0 -75.6 6.3 -56.9 -81.8 8 -67.5 -86.5 10 -74.8 -89.6 12.5 -80.9 16 -86.0 25 -91.6 35 -94.2

Just FYI, here's a comparison of some TDK parts in various dielectrics.

These data are from TDK's Component Characteristic Viewer,

JIL

What you're interested in is personal conflict, not capacitors.

I have previously mentioned using caps as the active devices in parametric amplifiers and as the nonlinear elements in pulse-sharpening NLTLs. I don't want to repeat myself too much.

Hey, post some circuits using cap nonlinearity.

No, actually, John suggested a parametric oscillator and several other applications.

e.g.

So I can build a resonant 33-60Hz VFO that only needs a 0-6V command voltage (assuming I can find a 4.7H coil), and you say that's a BAD thing?

Think of all the cool microwave techniques that can now be brought down to base band!

Or if you really want to annoy your neighbors, you could sweep a resonant load across your powerline...