Can the Dielectric Permittivity of a Material Be Altered?

[Mr.Bomzh]

Hi , I would like to know if there is any material , a dielectric like the one used in a capacitor , whose dielectric permittivity could be altered , like increased or decreases , I'm not talking about older type capacitors which had plates on a rotor that could be adjusted so that one could change the plate area.
I mean more like a material which is a dielectric but whose dielectric permittivity could be changed by some interaction , electric, magnetic maybe ?

 

[Simon Bridge]

See:http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=1122961&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D1122961

 

[Jano L.]

Dielectric constant depends on temperature, so in principle by controlling temperature one can vary the dielectric constant. Does not seem to be too practical though. There are also artificial dielectrics, that can be manufactured with prescribed permittivity, but I think these are quite big and are used not in static fields, but in oscillating EM waves.

In principle, one could prepare material with desired permittivity $\epsilon$ by mixing powders of two kinds with lower and greater permittivity $\epsilon_1$, $\epsilon_2$, but the mixing ratio and compression used would most probably need to be found empirically.

 

[Mr.Bomzh]

Ok , let me take a wild guess as a question , is or would it be possible to make a dielectric whose permittivity amplitude is so large that using it in a capacitor for example I could go from a few nF to as high as a few uF and maybe more ? To achieve noticeable charge flow to and from the capacitor as it would be charging up and down ?

 

[Simon Bridge]

you want a relative permittivity of 3-4 orders of magnitude ... you could just look it up?
http://en.wikipedia.org/wiki/Relative_permittivity
... i.e. Barium Titanate @ 120C has $\small \epsilon_r=10000$.
That would make a 1nF cap into a 10000nF=10μF cap.

Also see conjugated polymers.

But you don't need to do that to notice the "charge flow" (i.e. electric current) as the capacitor charges and discharges.

Did you have a specific application in mind?

 

[Mr.Bomzh]

ok thanks , I see Simon , well the thing I had in mind is, I have a capacitor in a circuit attached to a power supply in series with a transformer , I would need to change the permittivity of that dielectric so that I could change the capacitance of the capacitor which would make the charge flow back and forth to the capacitor , in other words a capacitor which also acts like a charge pump , I hope you understand what I mean, like in a transistor you control the current flow with an electric field or a smaller current flow , here you control the charge flow with varying the dielectric so that the amount of charge would vary on the plates.

 

[Simon Bridge]

You can get that effect (a charge pump) more easily, and cheaper too, by other means.
(i.e. - the PSU is a "charge pump" ... or do you mean one of these?)
So what is the application?

 

[Mr.Bomzh]

well , thanks for the link , I wasn't aware of such devices , yet I see they too need a transistor to do the switching part , I imagined that if one could make the right dielectric material for such a purpose one could have a all in one capacitor charge pump which could be used in a circuit with a transformer as a power converter , that was kinda the application.
I mean that would be the simplest device in it's very basis requiring only two capacitors and a transformer.

 

[Simon Bridge]

You'd still need something to do the switching ... in your case, how does the dielectric material know when to change it's permitivity?

 

[Mr.Bomzh]

yes that is the thing, well I would need some switches that would change the permittivity of the material if possible at all of course, but then at least those switches wouldn't serve as the main current paths, in fact there would be no main current path at all , rather current formed from charges flowing back and forth , what do you think is the idea even sound?

 

[sophiecentaur]

You actually want a variable C Capacitor, I think.
If you could accept a small value for you capacitance then a Varactor diode would do it. You would need to operate at high frequency, though.

One point has to be made, though. If you want to get a voltage increase by changing the Capacity, that will fundamentally involve putting energy into the system. Work needs to be done to reduce the Capacitance when there is a given charge on the Capacitor. Halve the capacitance and you will need to double the energy stored. This energy has to come from somewhere.
E = CV²/2
and Q = CV
both apply. V changes inversely with C, for constant Q
In the light of this, are you sure you want to proceed?

 

[Mr.Bomzh]

Yes I know that if I want the same power transferred with smaller capacitance the frequency would need to go up and vice versa for larger capacitance.

So are you saying that if I would have this capacitor with variable dielectric , upon varying the dielectric constant I would get what ? Charge flow or voltage rise? Or both?
Im speaking about a circuit case not a free standing capacitor?

As in a single no circuit capacitor charged to some value , varying the dielectric would cause voltage increase decrease because the charges have nowhere to go so if the field is weakened they need to set up the voltage to increase it back to it's pre vaied condition right?
But does that happen in a circit also ? because in a circuit the weakened field makes the charges on one plate repel more as they are of the same polarity but instead of setting the field back to keepthemselves on the plates they now have a path to flow to , maybe even more , another empty capacitor to charge up , so would they flow away and the result would be a cap with less capacitance and less voltage or no?
I had this argument with a guy who studies math and electronics and he too didn't knew the answer to the case in a circuit , we both only agreed to the no circuit case.

Hope to solve this one , as I learned quite a bit from the last thread, thanks sophie.

 

[sophiecentaur]

Yes I know that if I want the same power transferred with smaller capacitance the frequency would need to go up and vice versa for larger capacitance.

So are you saying that if I would have this capacitor with variable dielectric , upon varying the dielectric constant I would get what ? Charge flow or voltage rise? Or both?
Im speaking about a circuit case not a free standing capacitor?

As in a single no circuit capacitor charged to some value , varying the dielectric would cause voltage increase decrease because the charges have nowhere to go so if the field is weakened they need to set up the voltage to increase it back to it's pre vaied condition right?
But does that happen in a circit also ? because in a circuit the weakened field makes the charges on one plate repel more as they are of the same polarity but instead of setting the field back to keepthemselves on the plates they now have a path to flow to , maybe even more , another empty capacitor to charge up , so would they flow away and the result would be a cap with less capacitance and less voltage or no?
I had this argument with a guy who studies math and electronics and he too didn't knew the answer to the case in a circuit , we both only agreed to the no circuit case.

Hope to solve this one , as I learned quite a bit from the last thread, thanks sophie.

I am just saying that Q=CV and E = CV²/2 will always apply. You get different results, according to how the capacitor is connected, of course but varying the capacitance will involve having a PD change or a flow of charge. Any PD change represents a change in stored energy and any flow of charge will involve some energy transfer somewhere else in the circuit. For this exercise to achieve anything (in the broadest sense), energy will need to be put into making this change of C. If you can't identify the mechanism (it may be a bit subtle in some circuits - like charging and discharging a battery) then you have to keep digging until you can find it. You clearly want to sort this out so stick with it! You should break off for the New Year festivities, perhaps.

 

[Mr.Bomzh]

Nah , the festivities are kinda boring and they have a very low capacitance in my eyes :D
Even though I will drink some vine and look at some fireworks, hopefully not the ones that come out of a test circuit.Anyway they both smell " made in china"

so here is the circuit , what do you think , in such a circuit changing the permittivity 180 degrees out of phase would achive charge flow back and forth ?
I attached a schematic I just did in paint, excuse me for the quality and the blurred lines, I guess the vine is stronger this year

 

[sophiecentaur]

What did you want the circuit to do?
I presume that you want to cary the C's and, thus, vary the charge at each end of the transformer primary, resulting in an induced AC on the secondary. The 'detail' of how you will vary the C's is missing, of course and that would be the difficult bit.
It doesn't appear to be achieving anything.
It's a non-starter as a 'power supply' circuit because it would be, necessarily, very inefficient, for the reasons I gave, involving the inherent power loss. The Power to drive it would all come from the circuit driving the diodes. There is no DC path from the (DC) Battery ("+") so it can't deliver any power. The power supply is just providing 'bias'. A power oscillator circuit would be more applicable for driving the transformer.

 

[Simon Bridge]

yes that is the thing, well I would need some switches that would change the permittivity of the material if possible at all of course, but then at least those switches wouldn't serve as the main current paths, in fact there would be no main current path at all , rather current formed from charges flowing back and forth , what do you think is the idea even sound?

If the energy for changing the permittivity does not come from the "main current path" where does it come from?

This is my concern here - you appear to be describing a perpetual motion machine!

Take care:
You have to do work changing the capacitance by changing the permittivity just like when you mechanically change the capacitance by sliding some plates around.
Do you understand this?

 

[sophiecentaur]

He seems to be ignoring this!

 

[Mr.Bomzh]

Ok this was a very rough picture of what I mean , possibly the constant +ve supply kills any charge oscillation that might occur if the capacitors would be varied in capacitance one by one , one after the other.

It might only look like I'm talking about a PPM machine because I am not aware of all the mistakes I could or have made , I definitely don't think in terms of PPM , I know enough physics to know that doesn't work.
But why perpetual motion ? I was just thinking that, well you can run a transformer with a current that runs back and forth and increases and decreases in strength periodically just like an AC sine or square wave , but the same could be done by moving charges which is the same current back and forth using the attraction of repulsion of the electric field , rather than a direct current path,
Is it just me or I kinda thought that changing or varying field both electric and magnetic can transfer energy too , by the way that is how the transformer works as it too doesn't have a current path nowhere through it's core.

Ok , can you tell me sophie what exactly I am ignoring because it seems I am not aware of it as if I would I wouldn't repeat it.

 

[sophiecentaur]

As I have told you, the energy can only come from the mechanism that changes the values of the Capacitors. You cannot produce power without putting some in and you have not specified how you will do it.

 

[Simon Bridge]

What he said - that's the pmm part.
It's the most important bit.

Example:
Say you rig a waterwheel to turn a crank that pushes a dielectric into and out of the space between the plates of a capacitor. This will make the capacitance vary periodically - when the wheel is in a water-flow.

Put it in a box labelled DVC (driven variable capacitor).
This box will have two electrical terminals for hooking to an electric circuit along with separate inflow and outflow attachments for power ... in this example they would be hose attachments.

Now the box can have anything that varies the capacitor in it ... whatever it is does not matter, the external configuration will be the same. (It may be that the power-flow stuff has a different physical shape though.)

What's important is how the capacitance varies with time and understanding that you need an external source of energy to drive it.

 

[sophiecentaur]

The nearest thing I can think of to this idea is a Piezo electric generator in which the atoms of an insulator are polarised as the crystal is flexed. Not a million miles from a change in dielectric constant.

 

[Mr.Bomzh]

Ok I get it sophie and Simon ,I wasn't thinking that the capacitance would just change itself, it is like a transistor a but for example , it has to have some base current in order to make the device conducting , a mosfet probably is more efficient as to make it conducting you have to apply electric field on its gate , put some charge on it.
I was thinking kinda similarly with the capacitor , if one could make a dielectric which permittivity could be changed with some applied voltage to it or some current passing through, even though can't imagine how a dielectric could pass current through itself.Anyway if one could use something like this to change the permittivity of the material it could work a little similar to a transistor , well the part were you can regulate the device.

Well yes you could always move the dielectric physically but that wouldn't be high tech anymore and probably inneficient.

 

[sophiecentaur]

@Mr.Bomzh
I am not sure what you are actually after here. You are presumably accepting that some energy is transferred whilst changing the Capacitance (however that is achieved). If the transformer is supplying energy to a load them the energy transfer in the changing C will need a net input. 1W out of the transformer will require at least 1W put into the varying C. I can think of no 'efficient' way of changing C (the mechanical one that you rejected seems as good if not better than anything else I could imagine).
So where does that leave us? A varying voltage on the transformer would give a method of detecting a changing C but that is a far cry from a method of generating electrical power. (And the exercise would be futile if you were to start with an electrical source to change the C; just supply the power directly. I think this thread has pretty much exhausted itself as you don't seem to have a goal for this thought-circuit.

 

[Simon Bridge]

Ok I get it sophie and Simon ,I wasn't thinking that the capacitance would just change itself, it is like a transistor a but for example , it has to have some base current in order to make the device conducting , a mosfet probably is more efficient as to make it conducting you have to apply electric field on its gate , put some charge on it.

So it would work exactly the same as one of those, but you don't want to use one?

I was thinking kinda similarly with the capacitor , if one could make a dielectric which permittivity could be changed with some applied voltage to it or some current passing through, even though can't imagine how a dielectric could pass current through itself.Anyway if one could use something like this to change the permittivity of the material it could work a little similar to a transistor , well the part were you can regulate the device.

OK - so where would the device be getting the electricity to do this from?

Note: you need to look more closely at the available variable dielectrics to see what makes them vary.

Well yes you could always move the dielectric physically but that wouldn't be high tech anymore and probably inefficient.

But the example does help you think about how the device will have to work.
Like I said, anything can be in the box.
You still need to work out how the capacitor needs to vary ... and how to get electricity to make it happen.

 

[Mr.Bomzh]

well it wouldn't work exactly like a transistor as there is no current flow through a capacitor, well a transistor takes the current through itself , on the other hand a capacitor can take way more abuse give the right safety margin , well i think you can imagine if this would work out , put such a capacitor or two of them depending on the circuit and drive a transformer , now you don't have to care about current shoot through or current lagging or whatnot , these things can destroy transistor but nothing would happen to the capacitor as t just accepts charge no mater how much of it unless driven above it's safety limit of course.

To be quite honest so that you would understand , of course i could use transistors and i will probably have to because a device like the one I'm speaking about isn't made yet and I'm not even sure if one can be made.
It's just that using caps with carriable capacitance to drive a transformer would seem interesting from some points , the only thing I'm not sure about is can such a dielectric be made and how efficient it would be as sophiecentaur already pointed that out , even though I can't find why it wouldn't be efficient , first of all the charges don't meet they are just driven by electric fields , which are varied using the dielectric between the plates now the thing that would drive the dielectric so to speak , why would that be inefficient as if we assume some charge would have to be given to the dielectric then that charge could be later removed like in a mosfet , the gate doesn't waste the charge in fact one has to drain it in order to make the fet shut off, at least that's what I know.

the electricity would have to come from the same supply as the main +ve into the transformer primary , of course this is the part I don't know because i have no idea how or even if at all it is possible to make such a varying upon request dielectric.

 

[Simon Bridge]

the only thing I'm not sure about is can such a dielectric be made

Your designs will not make sense until you have worked this bit out - you have been given a starting point.

I can't find why it wouldn't be efficient

things are neither "efficient" nor "inefficient". Efficiency is a scale - things are more or less efficient in comparison to other things. If you don't know how the capacitor would be made, then you also have no reason to believe that this approach is any more efficient than a mechanical one or just using transistors.

Shelve this idea until you have more information - maybe it is more efficient: you don't know - Sophie is basing comments on experience - and I suspect he is talking more of the entire design rather than just one component. He has more hands-on electronics and engineering than me so, with nothing else to go on, I don't see any reason not to take his word for it.

But you haven't answered the question: imagine the resulting device is a box with four leads.
Two leads go to the capacitor plates, and two leads provide power to whatever is changing the capacitance, the variable dielectric, an electric motor driving the varcap crank, whatever. Where is the energy to change the capacitance supposed to come from? You'll have to change the voltage at the leads to vary the capacitance ... how does the voltage know when to change? Won't you need some sort of switch? You'll divert some input perhaps?

 

[Mr.Bomzh]

Yes Simon , to answer the last question you asked , of course I will need a switch , I;m not afraid to use one but this switch will not be in the main current path , rather it would be used to provide whatever voltage/charge or whatnot needed to the dielectric so that the dielectric would then alter the field which would then make charge flow into the main circuit through the transformer primary.diverting input could be an option yes , but again i don't know as I have no material or method in sight.
Of course it is hard for me to say now how much the switch would have to do because i am not sure about the dielectric, well i guess shelving the idea for now until i have worked out the possibilitty of such a dielectric is a smart idea to do.

I kinda get what Sophie meant when he said about the inefficiency of such a design , i kinda think he was referring to the amount of charge that could be displaced in each cycle to do work on the transformer via current flow , also he could have thought about the mechanism itself which would power the caps to change their capacitances , yet i find hard speaking about that one at least for me as I don't know about that yet.
well one can later the capacitance mechanically but that for sure doesn't seem a advanced nor solid satte nor long life method.

well a simple google search won't help me much on things like these variable dielectrics , I guess I'll have to talk with some materials scientists or some other poeple which could have a clue on how could a dielectric be made to change it's constant and most importantly by what means ...
you guys posted a link to a material that changes the constant with temperature but that is a dead end I guess as it would be possible to make it change a few cycles or a few hz in second , as I can't imagine how could a material change in a temperature amplitude of 100c more than a few times a second and even that seems a lot.

 

[sophiecentaur]

. . . . . ..

I kinda get what Sophie meant when he said about the inefficiency of such a design , i kinda think he was referring to the amount of charge that could be displaced in each cycle to do work on the transformer via current flow , also he could have thought about the mechanism itself which would power the caps to change their capacitances , yet i find hard speaking about that one at least for me as I don't know about that yet.
well one can later the capacitance mechanically but that for sure doesn't seem a advanced nor solid satte nor long life method.

. . . . ..

I don't think you do because you still have not acknowledged that energy has to be put into the system. You talk of "main current path" etc but I was referring to the fact that you would be needing to drive this variable capacitor with all (all) the power that is to be delivered to the load.

I am reluctant to use the word "nonsense" in threads and reserve it for when I am starting to tear my hair out. I am getting to that stage now because you are just speaking in quadrature with Simon and me. You are missing out on the very basic problem with your idea.
You will either need a Mechanical device (squashing or moving stuff with some energy source - which is low efficiency) or an electrical device / oscillator/ amplifier which will need to supply all the power to change the Capacitance. Why on earth, would you want to introduce this capacitor thing when you already have a perfectly good transformer to giver you any output voltage you design it for?

We (I, at least) agree that it is possible to make charge flow in and out of a capacitor as you change its dielectric constant (=Capacitance) and that you could measure the effect. What more is there to say? It's a total bummer as a form of Power Generator.

P.S. Your idea of heating and cooling things down is about the most inefficient form of energy transfer you could imagine (think Puffa puffa steam engine).

 

[Mr.Bomzh]

I never said I think the heating , steam engine whatnot part is efficient or will work that was just a sidenote.

Hmm , this is were I have to admit I really don't know what are you talking about , not because you would be wrong rather I just can't understand your point, would you please explain what is the terrible part about changing that capacitance? If it is even possible which I don't know , well it is mechanically but is it in a solid state way possible.

Well ofcourse we have a good transformer but a transformer ain't going to "drive" itself , from a dc source it needs sopme switches to do that , or some charge flow which changes polarity and magnitude with time, that was all i was trying to do.

If for example, I'm not saying this will work that way just for example , if I could have a capacitor whose dielectric I can change by applying some given charge to it and that would alter the capacitance then why would that be inefficient or bad?
As Simon said if I have this magic box with two main capacitor leads the + and - and two other for changing the capacitance or maybe one , depends on the method one could achieve this, then If I have two such boxes in series with a transformer primary each box located at the end of a pirmary and I switch them back and forth one by one and use the charge or whatever I used to alter the ones capacitance to the other in the next cycle.
Ofcourse it's hard to speak about this as i don't have the mechanism for the needed varyable dielectric , if I would have one only then I could fully speak about the efficiency and other parameters.

So I guess that is what I need to know, also Sophie could you please explain me this one thing and keep your hair steady at the same time ? :D about the power incom as you mentioned,ok here is my take on it:

A transformer can only transfer energy from the primary to the secondary if there is a time and/or polarity varying current in the primary, right ? right.Normally we have semiconductor switches that switch the + to - and do that and flip the sides of the primary on each cycle so that the transformer would see not only amplitude varyation but also polarity variation , like in a H bridge smps.
My idea was to have two capacitors attached at each end of the primary and both tied to ground , but at any given amount if time only one of them charges up fully while the other is if not totally empty (the ideal case) then atleast with a much lower capacitance , and so by charging one cap and discharging the other we could achieve a constant charge flow back and forth, the power to movre these charges as the transformer would be under load would come from the electric field present at the capacitors , the one capacitor which would charge up at an given moment would suck the charges in aainst the back emf of the transformer , is that right ?
I ca't understand what is so wrong about this reasoning?

 

[sophiecentaur]

My hair is still firmly attached at the moment. haha
It seems you still haven't cottoned on to the energy deficit in all this. You only talk about charge and current. You don't even include a load (resistor) on the output of your transformer. Without the load, the whole exercise is pointless - you must agree.
The following is an attempt to show that Energy is lost to the load and has to be replaced.
First of all, how we choose to change the Capacitance of our Capacitor is not relevant.You seem to imply that changing the dielectric is, somehow, different from moving the plates. In terms of the Energy, there is no difference. It is just one extra link in the explanation which adds complication. I will use moving the plates in this instance because it's easier to discuss (one less practical difficulty).

1.The plates on the capacitor start charged and connected to a source of DC.

2. We move the plates apart (they are attracted to each other so work is done on them) and the Capacitance decreases. Charge flows out of them, through the transformer coil into the battery and a current flows through the load resistor due too the transformer action. The energy dissipated in the resistor is lost for ever from the circuit.

3. We let to plates come together and the plates do work as the Capacitance increases again. More current flows and energy is dissipated in the load.

4. Over the cycle of operation, work is put in and taken out of the plates. BUT less work is got out than had to be put in. This corresponds to a net amount of Energy dissipated on the resistor (plus other losses)and that needs to be put into the device during each cycle.

So your invention needs energy put into it, however you look at tit. For a moving plate capacitor, this could be done with a Motor or electrically / electronically, using an amplifier etc.. With your proposal, energy is needed, to power the resistor PLUS energy to make the dielectric change and this is totally unspecified because you don't propose a specific system.
One thing which is pretty certain is that any method that can exist for changing the dielectric constant will be very wasteful of energy because it will involve, heating, squashing or some other bulk action on the dielectric material. Like I said before, you seem to think there is something inherently better about feeding a transformer primary your way than in a conventional way, such as an oscillator amplifier or a generator. There isn't. It could only be worse, imo.

I do wish you would reply to the important points that are made here. I have already stated that you could 'detect' the change in capacitance. That much would work. The rest is, I hate to say it, --- a nonsense proposal. It fails on basic principles, on the availability of substances and also on the sheer quantity of energy that can be produced. Enough to detect but not to blow the skin off a rice pudding, as we say.