Film capacitor insulation resistance spec.

I would "guess" there is a typo and it should say TC (time constant) > 400 seconds?

Hmm... I would just read it as IR (insulation resistance) * C > 400 seconds. Seems simple enough.

George H.

Ok. Thanks. My boss will probably want me to call AVX and get a number. I will report back to group.

Some guy here once posted measured time constants of some film caps. The numbers were in years, or possibly centuries.

If you have a charged cap that has a big negative temperature coefficient, when the temp increases C decreases, CV is conserved, so the voltage goes up. Stored energy increases.

So does the charged cap have a higher specific heat than a discharged cap?

Hmm, does the Q*V energy couple to the temperature of the cap? Where does the energy come from? A neg. TC means the cap expands as it warms up... pulling the "plates" apart. I guess that has to come from the thermal energy... I was going to say, "No, specific heat is the same". But I've talked myself into yes.

George H.

It may be wrong to assume CV is conserved. There's quite a bit of literature on dielectric constant changes for various films over temperature ranges, but I haven't come across any that explain what happens to a stored charge during a change in temperature or dielectric constant. It may be lost as heat. Another fun question is does storing a charge affect the melting/glass/transition temperatures of dielectrics?

Anybody here ever make a parrafin wax electret? I havan't, yet.

It says IR C > 400 seconds. That means that the insulation resistance times the capacitance (the self-discharge RC time constant) is > 400 seconds. So if your cap is 2.2uF then R > 181M ohms.

In practice this kind of spec tends to be an artificially 'bad' limit set by the maximum test time they want to allot, test equipment leakage or something like that. The insulation resistance of a film cap (that hasn't been damaged or contaminated- like a reworked SMT film cap that has split apart and got moisture inside!) is likely much higher than the guaranteed spec- by orders of magnitude- probably

--sp

I'm pretty sure charge is conserved... it would be a pretty strange world if it weren't and so far, I've not see any physicist say it isn't conserved. So CV is constant.

But let's consider another example. A capacitor is made with some dielectric with a large relative permittivity, say 4, giving a capacitance of C. Charge the capacitor up to voltage V. So the charge is CV and the energy is CV^2. Now, remove the dielectric. The charge is still CV but it is also C'V' where C' is C/4, so V' must be V*4. The energy is now 4*CV^2.

Clearly if this is correct, energy was added by removing the dielectric. Since the only source of energy was force times distance, there must be a force exerted on the dielectric as it is removed.

D delta E = Integral F(x) dx 0

Yeah I think we worked all that out in freshman physics. There was some lab where you pull a sheet of dielectric out from between the two plates of a parallel plate capacitor. (The big problem with the lab was that pulling out the sheet rubbed against the plates and knocked some extra charges off.... "In theory" the lab was great :^)

George H.

I'll rephrase my statement. It may be wrong to assume the stored charge is available later, as a charge. We know capacitors are not 100% efficient and there are weird effects like hysteresis where high voltage caps are stored shorted out.

Like peeling apart staticy layers of plastic, there is quite a bit of force involved in separating the plates. Are there any type of olden day high voltage generators based off peeling capacitor plates apart in a loop or someting like that?

I think I see where you are going with this, but I think you have still misstated it. I think you mean the energy put into the capacitor may not all be available which is true. This is measured by the dissipation factor and is well known. The hysteresis effect you mention does not prevent the charge from being available, but it may make it harder to utilize. It is also not so significant.

That's an interesting idea. Put low voltage charge on two closely spaced plates. Separate them and let the charge be drained off to storage. Bring plates close together again and transfer charge back to capacitor. Repeat.

If you simply separate the variable capacitor and the storage capacitor they can be separated by a diode eliminating the need for movement between the two. The only thing that moves would be the ground side of the variable capacitor (or the dielectric if the variability is created by moving the dielectric in and out of the cap). Another diode allows a voltage source to supply current to the variable capacitor when it the capacitance is high.

In essence this is analogous to a voltage doubler but with a large range of capacitance the voltage increase can be much more than doubled with a single stage.

Using MEMS, this could utilize very tiny vibrations to step up voltages and generate power.

van de Graaf generator.

Cheers

Phil Hobbs

snip!

This isn't new, of course. Look up 'electrophorus'. The Wimshurst machine and pelletrons rely on the same effect too.

Jeroen Belleman

Wimshurst machine:

Joe Gwinn