Captret - Perpetual Light with Dead Batteries

[gravityblock]

gravity,
I think he was trying to say if the battery is connected +/- to cap it will only fill up to the voltage on the battery. A camera uses an oscillator to fill up the cap. and your right it is the same thing as a stun gun or taser. But that is not what he is saying.

A voltage multiplier is an electrical circuit that converts AC electrical power from a lower voltage to a higher DC voltage by means of capacitors and diodes combined into a network.

When the battery is connected +/- to cap then the breakdown voltage of the dielectric isn't exceeded, and the cap will only fill up to the voltage of the battery. When the battery is connected in the captret configuration, then the capacitor's dielectric breakdown voltage is exceeded and conduction occurs just like a diode (this is the virtual or phantom diode Dr. Stiffler has already mentioned). Don't forget the LED is also a diode. This allows the capacitor to be charged slightly higher than the battery voltage. The captret isn't anymore OU than a voltage multiplier by means of capacitors and diodes. Don't kill the messenger.

GB

[ibpointless2]

I don't know what capacitors you guys are opening but my are two strips of aluminum separated by paper soaked in electrolyte. and the whole thing is again wrapped in another piece of paper soaked in electrolyte. So they outer case is aluminum too so it is just acting like another plate.

So i don't know people are so worried about them, the outer case is nothing more than another lead just like the + and -.

So, how can something so simple give a higher voltage then whats in the battery. A flash in a camera has a series of components that do it, the catret is only one thing.

A captret is nothing more than a 3 plate capacitor that does things very different.

[gravityblock]

I don't know what capacitors you guys are opening but my are two strips of aluminum separated by paper soaked in electrolyte. and the whole thing is again wrapped in another piece of paper soaked in electrolyte. So they outer case is aluminum too so it is just acting like another plate.

So i don't know people are so worried about them, the outer case is nothing more than another lead just like the + and -.

So, how can something so simple give a higher voltage then whats in the battery. A flash in a camera has a series of components that do it, the catret is only one thing.

A captret is nothing more than a 3 plate capacitor that does things very different.

A captret is an equivalent circuit of a half-wave multiplier.

An equivalent circuit refers to the simplest form of a circuit that retains all of the electrical characteristics of the original (and more complex) circuit.

GB

[rave154]

Hi, all

i rigged up a circuit...comprising my function genmerator driving a dual flip flop output to two mosfets ( A & B ), when A is on B is off and vice versa.

connecting from the positive of the cap ( the power source) through the LED...to the mosfet and onto the negative of the cap.

each mosfet has its own LED and uses the same cap as a power supply.
The cap is precharged and then the battery is removed and the circuit is turned on.

using 1.432 Hz and a 50-50 duty cycle... the cap was initially charged up to 4.6V..when the pwm is turned on, the LED's go full bright but within 1/3rd of a second go very dim and the votage on the cap has dropped to about 1.3V.

However... i then re-wired the circuit as it is in the attatchment.

As you can see, each LED & mosfet now uses each half of the captret as its power supply..but only 1 half is ever "Loaded" at any one time ( due to the flip-flopped nature of the circuit )...

again charged to 4.6V...when turned on the LEDS go full bright and it takes about 8 seconds to dim down to the same level of brightness as in the other circuit...meanwhile.the voltage goes down to only 3.6V.and drops only about 0.01V per 1-2 seconds.

[rave154]

re:- last post,

i now have left the battery connected across the capacitor ( as a top up ).and have an ammeter in series from the +

in captret mode the LEDS are nicely bright, not full but pretty good..and its drawing 3.07mA from the battery (@ 4.6V)

in normal mode, the LEDs are brighter than above, but drawing 107mA @ the same battery voltage

so,

normal = 4.6 x 0.107 = 0.4922 watts

captret = 4.6 x 0.00307 = 0.0141 watts = almost 1/35th of the wattage

however, as stated, the LEDs are a bit brighter in normal mode.

i will re-wire the circuit to use resistors as loads so i can put the scope on them and try and get a more quantitative measurement of the output ( any suggestions for this btw? )

[rave154]

continuing,

i have replaced both LEDs with two 3.3K ceramic coated resistors.

if we use Ohms law to find what the current should be in a standard circuit we get

I = V/R = 4.6 / 3300 = 1.393 mA

running the circuit in normal mode..the current draw appears as 1.41mA...pretty damn close

re-arranging the circuit to be in captret mode ( as described in my earlier post )...current draw appears as 0.37mA

putting the scope across one of the 3.3K resistors(while still in captret mode) gives the image attatched..

settings are, timescale 0.2 mS , 1V per Div

readings are, peak-peak voltage 3.69V, frequency 1.44kHz, RMS voltage 1.38V

doing some quick eye-ball clac's based on the scope shot, it would appear that we have almost 3V continuous across the resistor (minus the lil curved part)..so

I = V/R = 3 / 3300 = 9.09e^-4

P = I * V = 9.09e^-4 * 3 = 0.00272 watts

power IN = 4.6 * 0.37mA = 0.0017 watts

so power IN = 1.7mW and power out = 2.72 mW ???????

ya'll let me know what to make of these findings

[gravityblock]

continuing,

i have replaced both LEDs with two 3.3K ceramic coated resistors.

if we use Ohms law to find what the current should be in a standard circuit we get

I = V/R = 4.6 / 3300 = 1.393 mA

running the circuit in normal mode..the current draw appears as 1.41mA...pretty damn close

re-arranging the circuit to be in captret mode ( as described in my earlier post )...current draw appears as 0.37mA

putting the scope across one of the 3.3K resistors(while still in captret mode) gives the image attatched..

settings are, timescale 0.2 mS , 1V per Div

readings are, peak-peak voltage 3.69V, frequency 1.44kHz, RMS voltage 1.38V

doing some quick eye-ball clac's based on the scope shot, it would appear that we have almost 3V continuous across the resistor (minus the lil curved part)..so

I = V/R = 3 / 3300 = 9.09e^-4

P = I * V = 9.09e^-4 * 3 = 0.00272 watts

power IN = 4.6 * 0.37mA = 0.0017 watts

so power IN = 1.7mW and power out = 2.72 mW ???????

ya'll let me know what to make of these findings

The captret may have a different internal resistance than a capacitor connected in the normal way, and you may need to take this into account. Also, you took power IN measurements with the LED, then took power OUT measurements after replacing the LED with a resistor? Are you sure the math is correct? I'm only suggesting things for you to think about.

GB

[rave154]

hi,

power in was measured via ammeter in series from positive of battery & battery voltage...wether using resistors or LEDs

[ibpointless2]

@electricity

The captret 1.1 you made doesn't have a complete connection to allow the LED to turn on when in start mode. The negative of the battery never gets hooked up, so the LED never lights. The negative of the battery is only connected when in charge mode. Unless you're using the case to send to negative, but i don't know, am i missing something?

For you design do you always have to get the led to flash first or can you do without?

[Guruji]

captret

I put a diode on either side of the led to better protect the light emitting diode from reverse spikes. This reflects them backwards away and I think it is this that is going back to the battery. When I added the diodes they didn't drop the voltage at all. In fact the signal got clearer on the scope now. The led is bright now and the voltage is stable.

I started with 9.30 volts and now it is stable at 9.22 volts with about half to 3/4 brilliance on the led. The led reads 2.62 volts steady and is rated for 3.7 volts. I think I'm gonna add more ceramic caps to the top end in parallel to see if I can get a better swing to the signal. The ceramic caps seem to be very tight seeing they are rated for 1k volts.

I'm running at about .2 uf on the ceramic caps each side (2 in parallel) and they are labeled 104m. The package said they were 100000 pf 1kv.

Oh I forgot this is whith the new stuff I just bought. The Electrolytic cap is a whopping 1000uf 200v and I got a potentiometer with 100k ohm sweep. It is running better then the "captret" single can version and I think it can push more for less now with no degraded performance over time like the single can showed.

I'll let you guys know what the new diagram looks like after I update the drawing so that others can replicate this...

And here is the new diagram:

Hi Jbignes this circuit is a bit different now. The negative is not connected to the capacitor body.

[Jbignes5]

Yes...

Hi Jbignes this circuit is a bit different now. The negative is not connected to the capacitor body.

Trying to get a simulation of the effect the captret is showing. I don't think I got it right yet but it does use the led at higher output without killing the battery. I think I need some diodes in the circuit to direct the flow still. Since there are virtual diode being made between the pos , neg and the casing. The orientation of the diodes is tricky and I haven't had time to concentrate on the design much with the holidays coming up.

I havent given up on the idea yet but with my circuit at least it does not loose voltage on the battery so far if it is tuned right and it does not use the capacitor in the wrong way either making it safer. There should be a pair of diodes in between the Electrolytic and ceramic caps but getting the right direction of the diodes is frustrating to say the least with me not being able to concentrate on the design. I also believe there is a certain amount of induction from the case to each plate so that is still missing in my design. Once the inductors are included that will give us the amplifying factor we will need to cause a movement in the other side of the circuit. The inductors would be in series with the ceramic caps and then connected to the negative rail.

[Artisan]

Captret with Esaki mod Progress

It's been over 9 days since I have had the captret circuit running with the modified Esaki circuit cited in earlier posts. The PP3 battery is now showing 9.03 volts, and the 3mm superbright LED is the same brightness. All components are still running cool, and I see no defect in my 47uF cap.

Over the weekend I have successfully ran a 28 LED bulb off the same circuit, but from 2 4700 uF caps and an 18V supply (2 PP3 in series). The bulb was shining intensely.

I am currently doing further testing with this.

Thanks again for sharing this discovery ibp2.

Best regards

[ibpointless2]

It's been over 9 days since I have had the captret circuit running with the modified Esaki circuit cited in earlier posts. The PP3 battery is now showing 9.03 volts, and the 3mm superbright LED is the same brightness. All components are still running cool, and I see no defect in my 47uF cap.

Over the weekend I have successfully ran a 28 LED bulb off the same circuit, but from 2 4700 uF caps and an 18V supply (2 PP3 in series). The bulb was shining intensely.

I am currently doing further testing with this.

Thanks again for sharing this discovery ibp2.

Best regards

How fast is your LED blinking?
Do they blink so fast that they seem to not blink?

If its the same brightness for this long then thats very exciting! Even if it does blink. I made the one you shown me and had it blinking but blink so fast that it was hard to tell if it was blinking, sort of like a strobe light. I might revisit it, I've got a interesting idea that could boost the catret even more.

And the defect to the capacitor was a plain lie. The captret doesn't harm anything, the case is just another lead just like the + and the -.

[Artisan]

re: Progress

How fast is your LED blinking?
Do they blink so fast that they seem to not blink?

If its the same brightness for this long then thats very exciting! Even if it does blink. I made the one you shown me and had it blinking but blink so fast that it was hard to tell if it was blinking, sort of like a strobe light. I might revisit it, I've got a interesting idea that could boost the catret even more.

And the defect to the capacitor was a plain lie. The captret doesn't harm anything, the case is just another lead just like the + and the -.

----------------------

Hello ibp2,

Yes, the blink rate is too fast to notice. Although it all depends on the transistor used, it will blink once the avalanche threshold has been reached. Brightness in my little superbright has not diminished during this time.

The new circuit has both 4700uF caps/captrets connected in parallel (sharing the same o, + and -). I have also placed some inductors in series to the load which seem to greatly increase the discharge/drain time when the battery is disconnected from the circuit. The LEDs stay lit for longer before requiring a 'top-up' from the battery.

Powering LED arrays in this manner has sparked my interest as they are normally wired in series to each other. The larger the array, the larger supply voltage required.

I would be interested any input to making the captret even more efficient!

[gravityblock]

It's been over 9 days since I have had the captret circuit running with the modified Esaki circuit cited in earlier posts. The PP3 battery is now showing 9.03 volts, and the 3mm superbright LED is the same brightness. All components are still running cool, and I see no defect in my 47uF cap.

Over the weekend I have successfully ran a 28 LED bulb off the same circuit, but from 2 4700 uF caps and an 18V supply (2 PP3 in series). The bulb was shining intensely.

I am currently doing further testing with this.

Thanks again for sharing this discovery ibp2.

Best regards

Below is a quote from e2matrix on the captret thread found at overunity.com

As it is I've got a flashlight that will run over a year which uses a 9 volt battery and a 5mm LED. That's running non-stop over a year. Most 5mm LED's are very low power diodes which consume very little for the effect they give which is a fairly dim amount of light.

As you can see, 9 days isn't saying much.

GB