Fans of various high-voltage experiments often encounter a problem when it is necessary to use high-voltage capacitors. As a rule, such capacitors are very difficult to find, and if you do, you will have to pay a lot of money for them, which not everyone can afford. In addition, our site’s policy simply will not allow you to spend money on buying something that you can make yourself without leaving your home.
As you may have guessed, this material We decided to devote our attention to the assembly of a high-voltage capacitor, which is also the subject of the author’s video, which we invite you to watch before starting work.
What do we need:
- knife;
- what we will use as a dielectric;
- food foil;
- a device for measuring capacitance.
Let us immediately note that the author of the homemade capacitor uses the most ordinary self-adhesive wallpaper as a dielectric. As for the device for measuring capacitance, its use is not necessary, since this device is intended only so that in the end you can find out what happened in the end. Everything is clear with the materials, you can start assembling a homemade capacitor.
First of all, cut off two pieces of self-adhesive wallpaper. You need about half a meter, but it is desirable that one strip is slightly longer than the other.
The resulting sheet of foil is cut into exactly two parts lengthwise.
The next thing is to place one piece of wallpaper on a flat surface, on which we carefully place one piece of food foil. The foil should be placed so that there is a gap of about a centimeter along the three edges. On the fourth side the foil will stick out, which is quite normal at this stage.
Place a second sheet of wallpaper on top.
Place a second sheet of foil on it. Only this time we make sure that the foil protrudes from the side opposite to the previous step. That is, if the author had the first piece protruding from the bottom, then this time it should protrude from the top. Separately, it should be noted that the foil sheets should not touch each other.
Now we remove the backing from one edge and glue our capacitor.
If you plan to build a laser, an accelerating tube, an electromagnetic interference generator, or anything else of the like, then sooner or later you will be faced with the need to use a low-inductance high-voltage capacitor capable of developing the Gigawatts of power you need.
In principle, you can try to get by using a purchased capacitor and something close to what you need is even available for sale. These are ceramic capacitors such as KVI-3, K15-4, a number of brands from Murata and TDK, and of course the beast Maxwell 37661 (the latter, however, is of the oil type)
Using purchased capacitors, however, has its drawbacks.
- They are expensive.
- They are inaccessible (the Internet, of course, has connected people, but carrying parts from the other side of the globe is somewhat annoying)
- And, of course, the most important thing: they still will not provide the record parameters you require. (When we are talking about a discharge in tens or even a few nanoseconds to power a nitrogen laser or obtain a beam of runaway electrons from a non-exhausted accelerating tube, not a single Maxwell can help you)
Using this guide we will learn how to make a homemade low-inductance high-voltage
capacitor on the example of a board intended for use as a driver
lamp dye laser. However, the principle is general and with its
using you will be able to build capacitors in particular (but not limited to)
even to power nitrogen lasers.
I. RESOURCES
II. ASSEMBLY
When designing a device that requires a low-inductance power supply, you need to think about the design as a whole, and not separately about capacitors, separately about (for example) the laser head, etc. Otherwise, the busbars will negate the benefits of the low-inductance capacitor design. Usually capacitors are organic integral part similar devices and that is why the example would be a dye laser driver board.
Blessed is the do-it-yourselfer who has sheets of fiberglass and plexiglass lying around him. I have to use kitchen cutting boards, sold in the store.
Take a piece of plastic and cut it to the size of the future diagram. 
The idea of the circuit is primitive. These are two capacitors, storage and peaking, connected through a spark gap according to a resonant charging circuit. We will not deal in detail with the operation of the circuit here; our task here is to focus on assembling the capacitors. 
Having decided on the size of the future capacitors, cut pieces of aluminum angle to the size of the future contactors. Carefully process the corners according to all the rules of high-voltage technology (round all corners and blunt all edges). 
Attach the leads of future capacitors to the resulting “printed circuit board”. 
Mount those parts of the circuit that, if not assembled now, may interfere with the assembly of the capacitors later. In our case, these are connecting buses and a spark gap. 
Please note that the low inductance when installing the arrester is sacrificed for ease of adjustment. IN in this case this is justified, since the self-inductance of the (long and thin) lamp is noticeably greater than the inductance of the spark gap circuit, and besides, according to all black body laws, the lamp will not shine faster than sigma*T^4, no matter how fast the power supply circuit is. Only the front can be shortened, but not the entire impulse. On the other hand, when designing, for example, a nitrogen laser, you will no longer attach the spark gap so freely.
The next step you need to cut foil and possibly laminate packages (unless the capacitor size requires the use of a full package format, as is the case with the storage capacitor on the board in question.) 
Despite the fact that lamination ideally occurs hermetically and breakdown along the edges should be excluded, it is not recommended to make edges (dimension d in the figure) less than 5 mm for every 10 kV of operating voltage.
Edges measuring 15 mm for every 10 kV voltage provide more or less stable work even without sealing.
Select the size of the leads (size D in the figure) equal to the expected thickness of the stack of the future capacitor with some margin. The corners of the foil, naturally, should be rounded.
Let's start with the peak capacitor. Here's what the blanks and the finished, laminated lining look like: 
For the peak capacitor, a laminate with a thickness of 200 microns was taken, since a voltage surge of 30 kV is expected here due to “resonant” charging. Laminate required amount covers (in our case, 20 pieces). Place them in a stack (with the terminals alternately in different directions). Bend the leads of the resulting stack (if necessary, cut off excess foil), place the stack in the slot formed by the corner contactors on the board and press it with the top cover. 
Fetishists will secure the top cover with neat bolts, but you can simply wrap it with electrical tape. The peak capacitor is ready. 
The assembly of a storage capacitor is no fundamentally different.
Less work scissors, since the full A4 format is used. The laminate here is chosen with a thickness of 100 microns, since it is planned to use a charging voltage of 12 kV.
We collect them in a stack in the same way, bend the leads and press them with a lid: 
A kitchen counter with a cut handle looks, of course, evil, but does not interfere with functionality. I hope that you will have fewer problems with resources. And one more thing: if you decide to use pieces of wood as the base and lid, you will have to seriously prepare them. The first thing is to dry it thoroughly (preferably at elevated temperatures). And secondly, seal it hermetically. Urethane or vinyl varnish.
This is not a matter of electrical strength or leakage. The fact is that when the humidity changes, the wood will bend. Firstly, this will disrupt the quality of the contact and lengthen the discharge time of the capacitors. Secondly, if, as here, a laser is supposed to be mounted on top of this board, it will also bend with all the ensuing consequences.
When bending the leads, do not forget to lay them over an additional layer of insulation. Otherwise, in fact: the plates are separated from each other by two layers of dielectric, and the leads from the plate of opposite polarity are separated by only one.
Let's see what we got. Let's use a multimeter with a built-in capacitance meter.
This is what the storage capacitor shows. 
And this is what the peak capacitor shows. 
That's all. The capacitors are ready, the topic of the guide is exhausted.
However, you probably can't wait to try them out. We complete the missing parts of the circuit, install the lamp, and connect it to the power source.
This is what it looks like. 
Here is an oscillogram of current taken with a small ring of wire directly connected to the oscilloscope and located near the circuit that powers the lamp. True, instead of a lamp, the circuit was loaded with a shunt. 
And here is an oscillogram of a lamp flash, taken with an FD-255 photodiode aimed at the nearest wall. The diffused light is quite enough. It would be more correct to say “more than.” 
You can scold poorly-produced capacitors for a long time and look for the reason why the discharge lasts more than 5 μs... In fact, a flash lamp dumps a bunch of megawatts and even the light scattered from the walls drives the photodiode into deep saturation. Let's take the photodiode away. Here is an oscillogram taken from 5 meters, when the photodiode is not looking exactly at the light bulb, but slightly to the side of it. 
The rise time is difficult to determine accurately due to noise, but it can be seen that it is on the order of 100 ns and is in good agreement with the duration of the half-cycle of the current.
The remaining tail in the light pulse is the glow of a slowly cooling plasma. The total duration is under 1 μs.
Is this enough for a laser based on a punisher? This is a separate question. In general, such an impulse is usually more than enough, but it all depends on the dye (how pure and good it is), on the cuvette, illuminator, resonator, etc. If I manage to get lasing on one of the commercially available fluorescent markers, then there will be a separate guide on homemade laser on dyes.
(PS) I had to add another 30 nF to the main storage capacitor and it was really enough. The pipe, a photo of which can be found right there in the “Photos” section, worked even better than from a two-maxwell GIN.
In general, a discharge time of 100 ns is by no means the limit for the described technology for creating capacitors. Here is a photo of the capacitor with which the air pumping nitrogen laser works stably in superradiance mode: 
Its discharge time is already beyond the capabilities of my oscilloscope, however, the nitrogen generator with this capacitor effectively generates already at 100 mmHg. allows you to estimate the discharge time at 20 ns or less.
III. INSTEAD OF CONCLUSION. SAFETY
To say that such a capacitor is dangerous is to say nothing. An electric shock from such a container is as deadly as a KAMAZ truck flying towards you at a speed of 160 km/h. This capacitor must be treated with the same respect as a weapon or explosive. When working with such capacitors, use all possible safety measures and, in particular, remote switching on and off.
Predict everything dangerous situations and it is simply impossible to give recommendations on how to avoid falling into them. Be careful and think with your head. Do you know when a sapper's career ends? When he stops being afraid. It is at that very moment when he gets on friendly terms with the explosives that his head is blown off.
On the other hand, millions of people drive on the roads with KAMAZ trucks and thousands of sappers go to work and remain alive. As long as you are careful and think with your head, everything will be fine.
T-shirt capacitor
This type of capacitor got its name due to the similarity of the shape of the plates to the “T-shirt” package.
The inductance of this capacitor is greater than that of the capacitor described above or the candy capacitor, but it is quite suitable for use in CO2 or GIN. It is difficult to start the dye and is not suitable for nitrogen.
The materials you will need are the same as in the guide above: mylar film (or laminating bags), aluminum foil and tape/duct tape.
The diagram below shows the dimensions of the main gaps.
L - dielectric length
D - dielectric width
R - outer radius of the capacitor
The gaps from the edges of the dielectric are 15 mm. On the side where the contact strips of the plates come out, there is a 50mm indent. These indentations are made as minimal as possible for maximum capacitance at given L and D of the dielectric. Please note that these gaps are selected for 10 kV. (I doubt it makes sense to make this type of capacitor for higher voltages, so I won't write formulas here to recalculate offsets and gaps for other voltages)
The distance between the terminals of the plates is 30mm. This gap is also taken as minimal as possible for 10 kV. Increasing this gap will make the leads too narrow - the inductance of the capacitor will increase.
Manufacturing
The tank capacitor is ready. You can install it on your laser, GIN or other high-voltage device.
The electrical capacity of the globe, as is known from physics courses, is approximately 700 μF. An ordinary capacitor of this capacity can be compared in weight and volume to a brick. But there are also capacitors with the electrical capacity of the globe, equal in size to a grain of sand - supercapacitors.
Such devices appeared relatively recently, about twenty years ago. They are called differently: ionistors, ionixes or simply supercapacitors.
Don't think that they are only available to some high-flying aerospace firms. Today you can buy in a store an ionistor the size of a coin and a capacity of one farad, which is 1500 times more than the capacity of the globe and close to the capacity of the Earth itself. big planet solar system- Jupiter.
Any capacitor stores energy. To understand how large or small the energy stored in the supercapacitor is, it is important to compare it with something. This is a bit unusual, but visual way.
The energy of an ordinary capacitor is enough for it to jump about a meter and a half. A tiny supercapacitor of type 58-9V, having a mass of 0.5 g, charged with a voltage of 1 V, could jump to a height of 293 m!
Sometimes they think that ionistors can replace any battery. Journalists depicted a future world with silent electric vehicles powered by supercapacitors. But this is still a long way off. An ionistor weighing one kg is capable of accumulating 3000 J of energy, and the worst lead-acid battery is 86,400 J - 28 times more. However, upon return high power behind a short time The battery quickly deteriorates and is only half discharged. The ionistor repeatedly and without any harm to itself gives out any power, as long as it can withstand it connecting wires. In addition, the supercapacitor can be charged in a matter of seconds, while the battery usually needs hours to do this.
This determines the scope of application of the ionistor. It is good as a power source for devices that consume short-term, but quite often more power: electronic equipment, flashlights, car starters, electric jackhammers. The ionistor may have military application as a power source for electromagnetic weapons. And in combination with a small power station, an ionistor makes it possible to create cars with electric wheel drive and fuel consumption of 1-2 liters per 100 km.
Ionistors for a wide range of capacities and operating voltages are available for sale, but they are quite expensive. So if you have time and interest, you can try to make an ionistor yourself. But before giving specific advice, a little theory.
It is known from electrochemistry: when a metal is immersed in water, a so-called double electrical layer is formed on its surface, consisting of opposite electric charges- ions and electrons. Mutual attractive forces act between them, but the charges cannot approach each other. This is hampered by the attractive forces of water and metal molecules. At its core, an electrical double layer is nothing more than a capacitor. The charges concentrated on its surface act as plates. The distance between them is very small. And, as you know, the capacitance of a capacitor increases as the distance between its plates decreases. Therefore, for example, the capacity of an ordinary steel spoke immersed in water reaches several mF.
Essentially, the ionistor consists of two electrodes immersed in the electrolyte with very large area, on the surface of which a double electrical layer is formed under the influence of applied voltage. True, using ordinary flat plates, it would be possible to obtain a capacitance of only a few tens of mF. To obtain the large capacities characteristic of ionistors, they use electrodes made of porous materials that have a large pore surface at small external dimensions.
Sponge metals from titanium to platinum were once tried for this role. However, the incomparably better one was... ordinary activated carbon. This charcoal, which after special processing becomes porous. The pore surface area of 1 cm3 of such coal reaches thousands square meters, and the capacity of the double electrical layer on them is ten farads!
Homemade ionistor Figure 1 shows the design of an ionistor. It consists of two metal plates pressed tightly against the “filling” of activated carbon. Coal is laid in two layers, between which there is a thin separating layer of a substance that does not conduct electrons. All this is impregnated with electrolyte.
When charging the ionistor, a double electric layer with electrons on the surface is formed in one half of the carbon pores, and in the other half with positive ions. After charging, ions and electrons begin to flow towards each other. When they meet, neutral metal atoms are formed, and the accumulated charge decreases and over time may disappear altogether.
To prevent this, a separating layer is introduced between the layers of activated carbon. It may consist of various thin plastic films, paper and even cotton wool.
In amateur ionistors, the electrolyte is a 25% solution table salt or 27% KOH solution. (At lower concentrations, a layer of negative ions will not form on the positive electrode.)
Copper plates with wires pre-soldered to them are used as electrodes. Their working surfaces should be cleaned of oxides. In this case, it is advisable to use coarse sandpaper that leaves scratches. These scratches will improve the adhesion of the coal to the copper. For good adhesion, the plates must be degreased. Degreasing of the plates is carried out in two stages. First, they are washed with soap, and then rubbed with tooth powder and washed off with a stream of water. After this, you should not touch them with your fingers.
Activated carbon, purchased at a pharmacy, is ground in a mortar and mixed with electrolyte to obtain a thick paste, which is spread on thoroughly degreased plates.
During the first test, the plates with a paper gasket are placed one on top of the other, after which we will try to charge it. But there is a subtlety here. When the voltage is more than 1 V, the release of gases H2 and O2 begins. They destroy carbon electrodes and do not allow our device to operate in capacitor-ionistor mode.
Therefore, we must charge it from a source with a voltage no higher than 1 V. (This is the voltage for each pair of plates that is recommended for the operation of industrial ionistors.)
Details for the curious
At a voltage of more than 1.2 V, the ionistor turns into a gas battery. This is an interesting device, also consisting of activated carbon and two electrodes. But structurally it is designed differently (see Fig. 2). Typically, take two carbon rods from an old galvanic cell and tie gauze bags of activated carbon around them. KOH solution is used as an electrolyte. (A solution of table salt should not be used, since its decomposition releases chlorine.)
The energy intensity of a gas battery reaches 36,000 J/kg, or 10 Wh/kg. This is 10 times more than an ionistor, but 2.5 times less than a conventional one lead battery. However, a gas battery is not just a battery, but a very unique fuel cell. When charging it, gases are released on the electrodes - oxygen and hydrogen. They “settle” on the surface of the activated carbon. When a load current appears, they are connected to form water and electric current. This process, however, goes very slowly without a catalyst. And, as it turned out, only platinum can be a catalyst... Therefore, unlike an ionistor, a gas battery cannot produce high currents.
However, Moscow inventor A.G. Presnyakov (http://chemfiles.narod.r u/hit/gas_akk.htm) successfully used a gas battery to start a truck engine. His considerable weight - almost three times more than usual - in this case turned out to be tolerable. But the low cost and the absence of harmful materials such as acid and lead seemed extremely attractive.
Gas battery simplest design turned out to be prone to complete self-discharge in 4-6 hours. This put an end to the experiments. Who needs a car that cannot be started after being parked overnight?
And yet, “big technology” has not forgotten about gas batteries. Powerful, lightweight and reliable, they are found on some satellites. The process in them takes place under a pressure of about 100 atm, and sponge nickel is used as a gas absorber, which under such conditions acts as a catalyst. The entire device is housed in an ultra-light carbon fiber cylinder. The resulting batteries have an energy capacity almost 4 times higher than that of lead batteries. An electric car could travel about 600 km on them. But, unfortunately, they are still very expensive.