capacitor transformer

Hi,

Is there any application for using two electrostatically coupled capacitors to make a transformer? Would changing the ratio of plate area and plate seperation from the primary to secondary capacitor plates be able to control the voltage input and output ratios? This transformer would only be able to work with AC input just like an inductor based transformer, the main benefit I can see is that it works electrostatically so may not need an iron core to hold the EMF, and should be good for high frequency?

cheers, Jamie

The Colpitts oscillator make good use of this effect. For general use though, capacitors (unlike inductors) are pretty crap at storing energy. For reasonable powers you either need shed sized caps or terrifying working voltages (or both). regards john

You mean one of these ? Only good up to certain powers though.

Graham

Here's an interesting overview.

Graham

I'm not sure I totally understand your question. But, I think you are talking about the modern piezotransformer. It is a resonant circuit based on inductance and capacitance, which has to be driven at or near it's resonant frequency to get any output. It is largely capacitive however.

It's method of transferring energy from the input to the output is MECHANICAL in nature however, the ceramic piezo body moves as the electrical input varies in applied voltage, which induces a change in output of the output voltage.

It is more like an electrical to mechanical energy converter on the input side. This mechanical energy is applied to the output side, where it is converted back to electrical energy.

Step up voltages range from 10 to 200 and the output voltage is only limited by the dielectric thickness that separates the transformer elements. If the absolute maximum output voltage rating is exceeded, the device shatters from voltage arcing. It is probably possible to build these units to produce 200 kV (or more), but the present manufacturers of the devices only make them in configurations that limit the output voltage to around 6 kV p-p.

There are 2 basic types of devices, one is rectangular and is the more common Rosen type, used in backlight inverters and personal ozone generators (also known as piezo air purifiers). The less common round type is easier to drive and has more output power, but is virtually unavailable as it is being developed in labs at the moment.

If these are the devices you are looking for, there is a large collection of piezotransformer technical materials posted on the web (18 MB, zipped). Let me know if you want the URL.

Regards,

M

The role of mechanical force in a magnetic transformer's core isn't obvious, I can see where you'd have trouble imagining a metaphor.

Consider how levers and gears transforms force/displacement ratios. Now think of a capacitor as an energy transducer. How does force act on the plates of a capacitor? How could you put a lever between the plates of two capacitors to make a transformer?

Teaching piezo-transformers theory before magnetic transformer theory would lead to better appreciation of magnetic transformers as mechanical and not merely electrical devices.

Scott

--
**********************************

DIY Piezo-Gyro, PCB Drill Bot & More Soon!
http://home.comcast.net/~scottxs/

POLITICS, n.
A strife of interests masquerading as a contest of principles.
The conduct of public affairs for private advantage. - Ambrose Bierce

There is no giant behind the devastation of the world?only a shriveled 
creature with the wizened
face of a child who is out to blow up the kitchen because he cannot 
steal his cookies and eat them, too. - Ayn Rand
**********************************

Inductance doesn't have anything to do with it, unless it is the fictional "motional inductance", a term I've read used in modeling watch crystals, which can have thousands of "Henries". This "motional inductance" refers momentum as if it were flux in an inductor so spice can deal with the component.

Don't you think magnetic transformers are MECHANICAL in nature? Go take an iron transformer apart, delaminate then relaminate the core so you can move the "I" section away from the "E" section. I did this with a

200W transformer. You can hardly pull it off. If you put cardboard between, you'll notice the MECHANICAL FORCE making sawdust of it.

Electrostatic and magnetic forces themselves can be modeled as a compactified dimension in 3-space. Charge (Coulombs) measured in meters displacement.

You can model a magnetic transformer similarly.

I'd like to see a 200kV transformer! I took a cheap piezo disk, etched the electrode surface (see my web page) and found occasionally I could get twice the voltage out I put in. If I tried putting over 50-100V (IIRC) on the disk, around 5W, it would begin glowing blue from corona on the surface, and eventually fracture, glowing bright blue at the fracture sites. They can use piezo disks in ionizers too.

Where?

Scott

--
**********************************

DIY Piezo-Gyro, PCB Drill Bot & More Soon!
http://home.comcast.net/~scottxs/

POLITICS, n.
A strife of interests masquerading as a contest of principles.
The conduct of public affairs for private advantage. - Ambrose Bierce

There is no giant behind the devastation of the world?only a shriveled 
creature with the wizened
face of a child who is out to blow up the kitchen because he cannot 
steal his cookies and eat them, too. - Ayn Rand
**********************************

If by 'transformer' you mean a gadget that transfers power efficiently between source and load of different impedance, there's no direct capacitive equivalent of a transformer.

The closest thing is a charge-pump inverter. You need at least two capacitors: charge them in parallel and discharge them in series, and you get the equivalent of a step-up transformer. There are lots of ICs that do this for you. There's a newish part, TI or LTC maybe, that adapts the number of caps dynamically to make a variable-ratio inverter, which allows efficiency to be kept up, sort of like an automatic transmission.

John

No, I don't normally think of a conventional copper and magnetic transformer as functioning as the result of a mechanical movement of the coils or the core. Perhaps this is wrong, if so, thanks for the information.

I do imagine the coils as being fixed and the magnetic field pushing against the core...and in doing so, transferring the electrical energy into magnetic energy. The secondary senses the build up of magnetic flux, and of course, it tried to neutralize the build up, so it removes energy from the magnetic energy stored in the core.

Again, this might be wrong, or perhaps understated is a better word.

I agree 100 percent however that you could model a transformer as an electrical to magnetic to electrical conversion...or, as an electrical to mechanical to electrical conversion and get the same results.

You are either talking about extracting voltage from a piezo sensor, or I think you aren't aware of the piezotransformer specifically.

All piezo materials output a voltage when you stress them mechanically. The common application is quartz crystals and piezobased sensors. Yes, you would have problems getting anywhere near the multiple kV output from these.

However, a piezotransformer is a specialized beast, even when made in small versions, it outputs 6 kV p-p, and if operated with a high value resistive load, they can exhibit gains of 500 or more. In this case,

100 volts in would give 50 kV out.

The piezotransformers currently manufactured will shatter if operated with high output voltages however as the ceramic material will break down and arc.

If a piezo manufacturer chose to make a device that was physically larger, it could produce higher voltages.... The possibility probably exists to mount them in a vacuum tube enclosure to minimize the possibility of high voltage arcing around the device and to provide a means of extracting the high voltage output.

Take a look at the many technical papers that deal with the piezotransformer specifically. You will see that the thickess of the materiel and physical dimensions determine the output voltage gain. These factors are discussed in great detail in many of the documents in the piezotransformer data archive.

The archive can be found at:

Click on the PiezoXfrmrs.zip link and save to disk.

I am the collector of the material contained in the archive, which all came from www searches over the years. Some is recent and some is older. Joel had access to the web space, and agreed to post it. There are pictures, a couple of thesis papers, many theory of operation papers, some practical examples of non-typical piezotransformer usage, various sources of piezotransformer vendors and my Spice model (which works well in Spice).

Enjoy the reading.

M

Scott,

I did take a look at your website.

You used a piezo sensor to do your experimentation...........

I think you would see quite a bit of difference if you duplicated your work with an actual piezotransformer designed and optimized to put out higher voltages and power levels.

I agree that even piezotransformers made today won't go to mega kV outputs, but only because no one designed a device to withstand the higher voltages and because no one ever enclosed their high voltage capable device in a vacuum in order to extract the highest possible voltages without corona arcing problems.

You might want to order one to the TI evaluation modules that come with multiple layer PZtransformers. The archive has the full technical specs for the EVM module. You can also obtain them on ebay from steminc, steminc uses ebay as a means to sell samples. If you want a unit without specs, you can find them in used ccfl inverter boards or in 'piezo air purifiers' sold on ebay.

I would comment on your driver. In a properly designed piezo transformer, your driver would fail immediately. In real life, the voltage on the output of cmos gate exceeds Vcc and Vdd by some number of tenths of volts. This exceeds the capacity of the chips output as Vcc and Vdd typically define how much voltage you can apply to the input. For this reason, it is common practice to use a higher output voltage transistors set...which can withstand the higher voltage swings. Yes, you can clamp the PZtransformer to prevent the gate inputs from being driven to far negative or positive...but, in doing so, your efficiency go to hell. See the TI UCC3977 driver spec sheet in the archive for better info on this.

Enjoy the reading and please let me know if you do additional work with PZtransaformers.

M

The "electrostatic coupling" is just more capacitance. So the whole thing, circuit-wise, reduces to an array of capacitors, which doesn't make a very useful transformer.

If the circuit reduces to a big cap connecting the source to the load, sure you can get current to flow through it. But it's not a transformer.

John

There's that little thing about putting the power source in series with the load that cause the most trouble.

Scott,

Scott,

I've read your DIY page a couple of times now and I have to admit I skipped over alot of 'worthless' technical material on piezo transducers in my quest to learn more about piezotransformers. Reading your page today, I think I might have discovered the missing link.....so to speak.

Piezo transducers do come in BIG sizes, for instance the hull of ships below the waterline and the nose cone of submarines are both transmitters and receivers for sonar. A physically large piezo device is just what we need to make a really high voltage piezotransformer because of the physical spacing between the terminals provides a higher breakdown voltage.

The breakdown voltage is essentially the limiting factor in the voltage output amplitude, so a piezo material with a longer dimension between the electrodes should enable higher output voltages.

I know some piezo sensors are large and already come with metal electrodes bonded to the material, mounting structure and attached leads. I'm wondering if a box stock piezo sensor can be used to make a piezotransformer. The primary difference between a sensor and a transformer is the presence of a part of the crystal that is polled in a different direction. If I take a piezo sensor and remove part of the electrode and then mount it between 2 plates on the end axis, I can apply a dc voltage and change the polling for a portion of the crystal.

If I can get a third electrode bonded properly to the newly polled axis, the newly built device should exhibit the properties of a piezotransformer, with voltage gain and a resonant frequency!

I have to admit, despite all the research I did on piezotransformers, I never thought about making them myself from a piezo sensor.

I might just have a crack at this. I already have a significant knowledge of lasers, which can be used to quantify the physical movement of the piezo transformer with a varied voltage applied.

I'm psyched.

Are piezo sensors available cheaply or free?? I know piezo actuators are horribly expensive:>:

Thanks for the nudge in this direction.

M

In the simplest case I am really just thinking of a capacitor that has 4 sheets seperated by dielectric, and the first and third sheet make one capacitor and the second and fourth sheet make the second capacitor, and this is all rolled up like a radial electrolytic cap, but with 4 terminals instead of 2, corresponding to two seperate capacitors that are electrostatically coupled.

If you can run AC into the "primary", ie. charging and discharging it at an average current of 1Amp over a 50volt range, then should you also not be able to extract a large portion of this power from the "secondary", if it is electrostatically coupled very well to the primary?

So a capacitor transformer isn't good at storing energy compared to an iron core transformer, but it is good at moving energy, so if you keep the primary charging and discharging quickly that may work well?

cheers, Jamie

I forgot about AC coupling a load. :) (much simpler!)

cheers, Jamie

Its not desired the core move or vibrate, but since it isn't infinitely stiff, the energy of the flux it is conducting must cause it to deform. Looking at the electron spins, atomic bonding and crystal lattice, it looks like a composite electro-mechanical system, not just a perfect electronic component.

Charge is charge. If it isn't changing relative to your reference frame, it appears as an electric field. If its changing relative to your frame, its appears to be a magnetic field. Or something in between.

Thanks, sounds familiar.

Scott

--
**********************************

DIY Piezo-Gyro, PCB Drill Bot & More Soon!
http://home.comcast.net/~scottxs/

POLITICS, n.
A strife of interests masquerading as a contest of principles.
The conduct of public affairs for private advantage. - Ambrose Bierce

There is no giant behind the devastation of the world?only a shriveled 
creature with the wizened
face of a child who is out to blow up the kitchen because he cannot 
steal his cookies and eat them, too. - Ayn Rand
**********************************

Somewhere I read about a Russian EMP HERF gun used to zap perimeter security devices, which featured a crank that wound up a spiral spring on a ratchet. When the trigger was pulled, the spring unwinds, accelerating a hammer and hitting a substantial piezo-stack.

After that I surmise (the article didn't say) the pulse, probably in the microsecond regime, either fires a spark gap or GaAs-type impulse device to create a ~10 Ghz impulse.

Googled, but couldn't find it.

Of course. Also consider some VanDeGraaf generators have "equipotential" conductive wire rings going up the length of the belt tube, to equalize and average the voltage gradient so hot zones wont form if dust begins to accumulate. And high voltage caps, the self-healing ones, have layers of aluminized mylar floating between sheets of dielectric. Not only do the equalize and prevent inconsistencies/inhomogeneity of the dielectric from forming a hot spot to break down, but even if a cosmic ray happens to ionize a path, the aluminum vaporizes if an arc forms, reducing the field in that spot, thus the cap is "self-healing".

So it might help to deliberately reduce the resistance of the ceramic, if possible without ruining the crystal structure. Perhaps poly rather than monocrystaline material?

I read one article that described that. I was thinking (being ignorant of transformer design) that you'd have a donut shaped pattern on a disk. The large outer ring is the high-capacitance, low voltage primary, that mechanically, vibrates with a higher mass and lower velocity than the center. The center has low capacitance, high voltage secondary, that is vibrating with a high velocity and lower mass.

I suppose you could bake and repoll the ceramic. Don't know if the epoxy could stand the baking.

I did a double-take when I measured more voltage out than I was putting in! I worked with high-Q saw filters that had 20dB loss, and didn't expect to see it in a cheap disk.

I don't know what you mean by "piezo sensors", but the buzzer-disks you can get those disks for a $1/ea at Electronics Goldmine, All Electronics, et.

Scott

--
**********************************

DIY Piezo-Gyro, PCB Drill Bot & More Soon!
http://home.comcast.net/~scottxs/

POLITICS, n.
A strife of interests masquerading as a contest of principles.
The conduct of public affairs for private advantage. - Ambrose Bierce

There is no giant behind the devastation of the world?only a shriveled 
creature with the wizened
face of a child who is out to blow up the kitchen because he cannot 
steal his cookies and eat them, too. - Ayn Rand
**********************************

A tube full of transformer oil, like a car coil, would probably be the cheapest and most effective unless its for some kind of satellite or expensive application.

Of course, transformers are power devices and require protected drivers. Piezo-gyros and are low-power applications, but they, like accelerometers, if they got a good hit, could knock a chip into latching up without protection diodes.

Thanks for the info, I let you know if I do.

Scott

--
**********************************

DIY Piezo-Gyro, PCB Drill Bot & More Soon!
http://home.comcast.net/~scottxs/

POLITICS, n.
A strife of interests masquerading as a contest of principles.
The conduct of public affairs for private advantage. - Ambrose Bierce

There is no giant behind the devastation of the world?only a shriveled 
creature with the wizened
face of a child who is out to blow up the kitchen because he cannot 
steal his cookies and eat them, too. - Ayn Rand
**********************************

Hey Scott, I'll email you so we can take this dicussion out of the public domain.

TTFN.

M