This is a spin off from the Dollard thread. In it there was discussion of mutual inductance and complex quantities. long story short it had me think of something and I gave a quick try.

A transformer core made up of capacitor plates. It's a very strange thing. I'll state that there is a lot more to it than just randomly putting the parts together, there are some very critical and complex ratios between capacitance and inductance in this particular geometry.

RF frequencies have odd effects differing from DC voltage and low frequencies as well. DC Pulsed is most effective.

I made a couple of these cores, one used some magnets I had laying around and the other was made from a ferrite bead cut in half like a bagel and then two copper plates affixed with an oil soaked wax paper dielectric, this one is only 3/4" dia yet ended up with a 10000uF capacitance

There is no difference between the magnets and ferrite core, prefer the ferrite core so that copper plate may be used.

Thought I'd share and see if anyone else has tried such a thing.

I'll be working a bit more with this and seeing if I can generate some equations to design an efficient core.

Can you measure that again please....

Here's the thing, both my meters tap out at 10000uF and this exceeds it. All it is, is two copper disks 0.75" dia with a 0.25" hole in the center, ground and polished flat, vellum paper soaked in oil as the dielectric.

Hi madhatter, Tesla did a similar thing, or more or less the same thing, explained in this book,
http://ia700302.us.archive.org/16/it...00martiala.pdf

Chapter XXI page 101, MOTOR WITH A CONDENSER IN THE ARMATURE CIRCUIT

Maybe his findings can help.

Thanks Farmhand, I've got that book and will be going thru it with a hard look at that.

I also need to re-do the arrangement. instead of the plates being separated by the dielectric alone there will need to be a hi-MU material between it as well. the magnetic field is between the plates due to the charge movement.

I'll make another core that's multiple layers and use some sheet MU metal that's sandwiched between the dielectric.

mu metal will be interesting!

Sorry but the figure of 10,000 uF for a small washer cap is stuck in my head so I had to pull out the formula and see.


Area of a washer is Pi(R sq – r sq) or Pi( .75 sq - .25 sq) which yields 1.57 sq in.


Using an online capacitance calculator with the dielectric constant for mineral oil and plate separation of .001 in. I get a capacitance value of 787 picoFarads which is in line with what I would have guessed.


Either I missed something or there is a problem with the measurement you made earlier. Just trying to be helpful. I am very interested in what you are doing.

Well that little test one didn't last long, arc'd over and fried at about 40 volts. It's was very odd though, couldn't get a value on it

finished the stacked unit tonight, 3 copper plates, 6 mu-metal sheets and 6 oil soaked vellum paper dielectric sheets.

It's a 2"x2" square, with a 0.6 x 0.6 center opening. wound 22ga mag wire on one leg with about 20 turns. pulsing the capacitor core with 1v dc will light a 5v 45ma LED. I can also light the LED with an RF signal of 600Hz and .75v on the plates.

need to run some more tests, also need to build a pulse circuit for the generator.

Ohh the capacitance on this new core is ~3.0nF more than likely due to the mu-metal material separating the plates.

There is also the thought of looping one lead of the winding back to one of the plates and seeing if it will resonate on itself.

Very nice Out-of-the-Box thinking I like it. Maybe Transpacitor? or Capaciformer? If you look at the basic electronic components there just aren't that many so to have a completely new one could potentially really open a lot of new doors. While it may still have values like inductance, capacitance, resistance and so on who knows what effects might be found with a new integrated combo.

It's different that's for sure, not sure if it's a new idea though. Went thru Tesla's notes and his use of the capacitor in the core was different in that the core was axial to windings.

The magnetic field in a transformer core is bound geometrically to the core, the magnetic field in between capacitor plates runs circular. The question then is, would the capacitors field have mutual inductance to the windings when paired to a core? Is there and amperage draw in the capacitor circuit when putting a load on the windings?

Uploaded with ImageShack.us

new core.

This is one interesting and fun little project. The peak voltage occurs when the frequency on the plate matches the length of wire wound on the core. an increase occurs when the start lead is connected the center capacitor plate. The signal is fed to the outer plates, the entire circuit is one wire.

feeding a 0.6V signal at 1.48Mhz 0dBm the winding will output 0.917Vdc

waiting for the oscilloscope to arrive to do more tests.

So far I've been able to determine that the signal frequency to the plates will induce resonance in the windings when the fq matches the winding resonance based on length.

I can not get an amp draw reading on the plate signal source, still not sure if there is a current draw off the plates when inducing the windings.

It seems like the same effect could be achieved if the inter-turn capacitance was induced independently of self inductance. there is a magnification of voltage due to increased windings, what I'm not clear on is current. The phase seems matched as well, the oscilloscope will help with that when it shows up, hopefully soon. the RF volt meter reads fine for now.

going to try a few different type of windings and see what the results are. have a couple ideas. The math is starting to form in my head based on my notes so far.

Such a simple arrangement, yet it's highly complex in how it operates. very very sensitive to capacitance and electrostatic fields. Also seems to react to or pick up the CD -capacitive discharge- of ignition systems in passing vehicles.

By placing my hand near the transformer the RF voltage will increase about 300mV.


Has me also wondering what type of effect could be achieved with a tube arrangement. That of course goes into my original research work, a modified multipactor electron sink/source. FTL with ES wave signaling, this may be the parametric layout needed to signal from ES to EM without losses.

I hope Eric follows this, as there are many similarities to his work.

As you know, I have done a lot of research on multipactors as well. It is my favorite thing. I have recently come up with something I call Farnsworth Unleashed. I would like to throw some ideas your way and get your opinion in private message or email. Interested?

my email is infra.rods@gmail.com

Orion

I've got lots of notes to compile into a spreadsheet and start extracting useable calculations.

I'll fill in here summation as I go along and will hopefully have more detail later tonight.

noticed an interesting thing, its' not winding layers or number of turns that gives the voltage but surface area of the windings in the first layer, subsequent layers do not increase the output. very strange.

the last winding will be a bifilar type as it'll take the most effort.

HF source voltage to the plate has more current output on the windings where as a LF source has higher voltage.

have not tried spaced windings yet to test inter-turn capacitance.

some very interesting neon bulb data to compile. will post a bit later..

going to feed it with a pulse coil next...

Uploaded with ImageShack.us

Power supply is via a DC convertor. 9.75V @ 1.5A. The supply has an AC leakage of 4.5V @ 0.3mA when measured against a ground on either lead from the PS. No AC present when tested from both leads.

output measurements as follows.

22.20 VAC
46.58 VDC

attaching a 12v 21W 1156 bulb to the PS side changes the voltage thusly:

21.88 VAC
49.82 VDC

this circuit will charge a capacitor to 70V, lights neon bulbs.

tried to measure amp draw on the PS side when charging a cap and neon bulb and there is no measurable draw. putting the bulb on the PS leads shortens the time it takes to charge a capacitor.

the ground lead is responsible for the voltage, no ground lead and voltage drops signifigantly.
disconnecting the lead from the negative terminal to windings and ground lead the V points measure 4.54VDC and 1.97VAC, putting the bulb on the PS side the voltage increases steadly but slowly.

The circuit draws strongly from the earth ground, a 16sq/ft foil surface wrapped in 3mil poly acts a virtual 'antennae' ground and voltages are lower by about 15VDC and ~18VAC.

interesting to note, it's difficult to measure the voltage, as it is dependent on load. just the terminal ends reads a maximum of 48V putting a capacitor across the terminals the voltage increases to 70V however it will light a 120V neon bulb.

the windings are .017" 25awg enamel magnetic wire parallel wound series connected bifilar. total of 17 turns.

just adding a few notes...
total wire length is 10'

this is an unbalanced circuit, as the capacitor plates are odd numbered there is going to be a miss match in the dielectric charge, this in-turn pulls from the ground. with the ground being on the load side the load is now in the circuit and needs to be balanced as well.

70v is the maximum so far at 60Hz.

need to adapt a number of capacitive equations and develop some capacitive inductive equations to model this.

core field strength is ~480 uT

Steinmetz capacitive M inductance equations with given results predicts a 0.8 T field to achieve the voltages given.

Very interesting, May I make a suggestion?

Separate the circuit for giggles. Create a tank circuit out of the primary winding and a separate inductor. Drive this at resonance. Now place a DC current through the secondary winding. Place your scope probes across the secondary. Now you can see if the effect of charging and discharging the capacitive plates affects the transformer operation. If it changes the permeability of the core, you will get a parametric oscillation, and you may see a wiggle on the scope probe of 2x the frequency of the tank circuit created by the primary and separate inductor.


Just a thought.....

I know that the charge cycle on the plates effects the windings, in a sense this is a tank circuit as the primary windings are the plates and the secondaries are the windings around the core.

got a scope tonight, going to setup and start again tomorrow.

got the scope up and running yesterday, odd results. One of my probes is not working as it should, need to fix that.

putting a DC charge on the plates then overlaying it with an AC signal is very effective. Still having a dickens of time getting an ampre draw reading on capacitor plates when in operation. I'll try a couple other things today, last resort is to go get one of those wall plug meters and use that.

One big issue is near field capacitance, and the voltage spike. The latest prototype hums at 120Hz or twice the plates. a cap and neon bulb will resonate based on the std parameters although it's when the bulb is just about to flash there is a MASSIVE voltage spike that the meter chirps a warning on uS before the flash. the scope picks it up with the probe laying about 12" away.

I'm also not comfortable with the floating differential measurements, as the ground significantly increases voltage.

plugging away, well the cores don't last long. shorting and arcing over to the windings is becoming a problem. need to move up to borosilicate glass and probably silicone or ptfe insulated wire.

back to work...

new core design,interesting result. Taking the B field of a capacitor and optimizing a layout to work in this strange circular field I came up with a log periodic array. I put together a simple 2 sector prototype, not calc'd just rough. It's the most effective layout so far.

did a quick layout in cad and will need to work on the calcs for the series. a number of parameters to account for, size, frequency, pulse duration, field intensity.

Eric if you're reading this it falls directly into your post on the log periodic. What I'm wrestling with though is the time variant or invariant electrostatic field movement that induces the magnetic field in a capacitor.


I'm still surprised by the simplicity of this and that is does indeed work. OK simple is relative, there is a surprising amount of calculations needed to make it efficient, but it's still simple enough.




Uploaded with ImageShack.us

9' of 0.017" wire wound in the star pattern, log periodic spacing. ends up with 4 turns per pattern. one wire lead out.

output is 28 VDC.

it's seems to be a type of electrostatic generator, going to build a bank and see if I will increase the voltage.

A mineral oil bath added 5 volts to the efficiency.

experiment and research, well during the research phase of this it's developed into where this phenomenon is known as 'displacement current'. This is where is gets complex, very complex. The missing 'J in Maxwell's modified Ampere law.

Hi madhatter,
Thanks for all the updates on this interesting topic.
If you can post the odd photo, every now and then, so that I can get a better idea of the setups,that would be great.

PS: I am dunking my "replications" into the air core of excited coils and try to "see what gives".

Attached Images

4aluplates_between_ringmagnets.JPG (69.1 KB, 23 views)

been busy working away on the calculations and equations. So far I'm at where something doesn't add up, no surprise really.

The voltages I'm getting are much higher than the math suggests, more research is needed.

I also wonder about lenz's law in this case as well.

Hi madhatter,
thank you for the photos.
So you place the flat star winding in the electric field between two capacitor-plates ?
Which basically gives you the following 5-layer sandwich:
Is my understanding correct?

Interesting. This reminds me of something I noticed, I think watching some TV documentary on something relating to religion/churches/cathedrals and being cynical of all the "innocent" religious/masonic symbolism and all that propaganda type stuff, they showed a pattern on a floor which I thought through reason of my views on the use of symbolism might make an interesting coil. After some google image search and narrowing it down to the terms "Notre Dame floor" or "Labyrinth Notre Dame":











[edit] Centre of the square cornered one

http://3.bp.blogspot.com/-3Nqy3-Wm9G...Q/s1600/31.jpg

Yes, it's a flat steel washer. I'm going to replace that with plexi tommorow as I want to remove any ferromagnetic fields.