Using ultracapacitors for high current source of inductive power

[Spencer DePue]

Hello all, I almost gave up on my fusor project knowing that i couldn't develop a sufficient power supply. ( It always broke..) I am always looking for alternative ways to power stuff, and ultracapacitors have seemed to attract my interest. These cola can size caps can store ridiculous amounts of energy. Some are have a capacity of thousands of farads, and usually store kilojoules worth of energy for only two or more of them. Of course they do not discharge this energy immediately like their lower capacitance higher voltage cousin, the electrolytic capacitor, Ultracapacitors can release tens of amps a second, maybe more. Sorry for the rambling , but here is my main point.

I know that in order for IEC fusion to work , the power supply needs high voltage (20,000-60,000 volts) and milliamp current to prevent the electrostatic grids from overheating.

However, big machines such as tokamaks use multiple forms of energy input ( Ohmic heating, RF heating, Inductive transfer of energy, neutral beam injection etc..)

Fusors only use electrostatic compression. So why not use 120 volt 15 amp 60 hertz Alternating current to start a plasma without fusing it, and than hooking ultracapacitors to an inner toroidal coil thus magnetizing it and ohmically heating/ compressing the plasma? After all it is power that does fusion. With a plasma start up at 1800-2000 watts, and magnetic compression at a couple thousand watts per second, wouldn't this work?

[Jeroen Vriesman]

IEC fusion doesn't work with "temperature", it is the electric field of the virtual cathode which attracts the positive ions towards the center.

Having a "pre-heated" "plasma" just creates some noise I think, maybe even less fusion.

[Spencer DePue]

This is not traditional inefficient fusor fusion. There is a reason why fusors produce nanowatts of energy, it's because of the mere concept of electrostatic confinement, using this method alone has not seemed to do us much good in the prospect of fusion energy has it? The only method I can see, and I regret saying it, is big machines such as tokamaks that cost billions of dollars for the projects entirety. Tokamaks require countless amounts of heavy duty electrolytic capacitors for their power guzzling regimes. Compare a cola can size ultracap with 3000 farads, and a couple hundred electrolytics in parallel to match this. We can make tokamaks less expensive with the use of alternate technologies. Ultracaps use nanotechnology.

The whole point of pre-creating the plasma is for a host for the ultracaps to do their work. tens of thousands of kilojoules of work to be exact. ( For 1 ultracap cycle of which it can do this 1 million times) Putting 40 kilojoules for a couple minutes in a coil of thick wire is bound to create a strong magnetic field. This approach also eliminates the complexity out of designing complex circuits. I can't see how this wouldn't do the job of creating better than average fusion yield for an amateur reactor.

[Carl Willis]

Hi Spencer,

IEC fusors, being a variant hollow-cathode geometry, make decent low-pressure plasma sources, and certainly magnetic confinement is an accepted route to detectable fusion. Whether or not your hypothetical, turduckenesque composite of a toroidal pinch coil inside a fusor would result in detectable fusion (i.e. would it "work") depends on too many factors to meaningfully speculate. Questionable unit dimensions ("watts per second", "can release a couple thousand amps a second") point to some possible physics misunderstandings.

First, though, I question your rationale for redeploying in this novel direction. Your rationale seems to be merely that you can't get an HV supply to work! The HVPS challenge--technical, budgetary, whatever--seems pale in comparison to the effort of designing and implementing the hybrid creature you have described. Lots of people can help you with the HVPS, and there are many ways to get one that works on a budget (I have given out several), but you have to share the details of your efforts.

-Carl

[Spencer DePue]

Allow me to correct myself, joules per second equals the watt, simple misstatement. And I never said a couple thousand amps per second.. I said an couple thousand watts per second (mistake) translating to couple thousand joules per second. First of all, one ultracap has 3000 farads capacitance, and is rated for 2.7 volts. Put two in parallel and you get 6000 farads, and 21 kilojoules of stored energy. If you don't believe how awe-inspiring these capacitors are check the maxwell ultracapacitor datasheets.

This is not anything close to my first post about the pyroelectric crystal. That was a misunderstanding of physics... This is based on information given by product datasheets.

As I said these capacitors don't deliver an instantaneous 21 kilojoules, they drain it over time. With the capacitor rating mentioned above. It can store a total of 7777 amps. 21000 kilojoules / 2.7 volts = 7777 amps. Its just a matter of how fast it releases this energy.

[Tyler Christensen]

The units in this response also have some issues... for instance a capacitor doesn't store a quantity of amps, it stores a quantity of amp-seconds, or coulombs. In this case, Q=CV=16200Coulomb. So draining a 6000 farad capacitor from 2.7V fully down to 0V could supply 16200 amps for one second (although the supercapacitors as you said are not likely to be able to actually do this).

Also you said 21000 kilojoules, but I'm guessing that was just a unit mistake.

[Spencer DePue]

My bad 21000 kilojoules was an error, i meant 21 kilojoules. On another note, I used a solenoid calculator to find out how many teslas a ultracap powered coil can produce an the number was staggering. Of course I know that these calculators are the most ideal in terms of what it calculates to and ignores many variables. Using a 10 inch length, .5 inch radius, 10 turn coil of wire with a magnetic permeability of .00875 H/M ( permeability of steel) at 100 amps/second this electromagnet can produce 34 teslas. MRI magnets usually produce 7 teslas. Something is wrong here..

[Chris Bradley]

Spencer DePue wrote:
> Using a 10 inch length, .5 inch radius, 10 turn coil of wire with a magnetic permeability of .00875 H/M ( permeability of steel) at 100 amps/second this electromagnet can produce 34 teslas. MRI magnets usually produce 7 teslas. Something is wrong here..

100 amps/second at the start of a calculation usually leads to something wrong.

[Spencer DePue]

That 100 amps/second is probably the average amount of current given off by an ultracapacitor every second. The spec sheets do not give a discharge rate per second. But they do say that a continuous current output of these things can provide way more than 100 amps/second. One of the problems is that couple hundred amps/second provides a shorter cycle than 100 amps/second. And ultracaps have an estimated 1 million cycles lifetime. I don't get why the magnet can have far more powerful than a superconducting MRI magnet, I guess the calculator is not factoring the other variables..

[Chris Bradley]

Spencer DePue wrote:
> I don't get why the magnet can have far more powerful than a superconducting MRI magnet, I guess the calculator is not factoring the other variables..

..but variables in what units, that must be the starting question here?! With 'powerful' Teslas, capacitor ratings in amps/sec, and plasma temperature in Watts per microJoule Farads per Volts Hertz already discussed, I think the problem is that you've missed out the number of turns per remanence megaMaxwell as the key variable for your solenoid.

You might find that the calculation, with respect to MRI machines, might look more correct if you started out on the assumption that patients in MRI machines need to be embedded in a solid block of steel when undergoing their scan.

[Carl Willis]

Magnetic field in a solenoid is an example problem in most intro physics texts and also on Hyperphysics, http://hyperphysics.phy-astr.gsu.edu/hbase/hframe.html . A few basic assumptions are transparently introduced, through the use of which all math reduces to simple algebra. So there is no need to blindly entrust this kind of calculation to online calculators. You can take your numbers and do the calculation with a pen on paper, being mindful of the units.

There remains an evident dimensional misunderstanding in your 100A/s number. A likely reason for a bad result from a calculator is "garbage in, garbage out." Check and double-check what numbers you are putting into the calculator and make sure the units are consistent with its scheme (inches or meters, teslas or gauss).

One practical reason why a steel core will not support ~34 tesla fields is saturation. A plug-'n'-chug Ampere's Law calculator typically has no clue about this. Anyway, I thought your solenoid was intended to confine a plasma, not have a steel core placed in it.

-Carl

[Spencer DePue]

here is the URL for Maxwell Technologies datasheet on one of their brands of ultracapacitors.

http://www.tecategroup.com/capacitors/d ... 015370.pdf

I was confused on a few matters, under power and energy , they say the BCAP3000 has 3020 W of usable power. They don't specify if that is the usable power in a discharge cycle. I assume it is, but they don't say. If this is true, 3020 watts / 2.7 = 1118.5 amps per discharge cycle. They also don't specify how much it can discharge in a second. Please advise.

[Tyler Christensen]

Both Watts and Amps are not a quantity of stored energy... it's not how many total watts or amps it has per discharge, it's how many watts and amps it can instantaneously provide during any given discharge cycle. In this case it looks like the safe constant draw is 147A, 3020 Watts/1118 Amps is the rated energy (not necessarily constant), and 2170A is the peak suggested current for momentary draws.

So regarding how much it can discharge in a second, use those numbers... an amp is a coulomb/second and a watt is a joule/second, so you can use these numbers to figure out what I'm guessing you're after.

[Spencer DePue]

I been thinking about a lot of things all day today so I may look like an idiot when I say this, I think the amps delivered in 1 second is 147 amps..... If I'm right, than this is higher than my original 100 amps/seconds which was an estimation. If this is true with the safe constant draw being 147 amps, which is 147 coulombs is 1 second, and the rated energy is 3020 watts than that ultracapacitors discharge cycle would only last 7 seconds.. My electronics skills are terrible, I just know a lot about nuclear power.

[Chris Bradley]

Spencer DePue wrote:
> I been thinking about a lot of things all day today so I may look like an idiot when I say this, I think the amps delivered in 1 second is 147 amps..... If I'm right, than this is higher than my original 100 amps/seconds which was an estimation. If this is true with the safe constant draw being 147 amps, which is 147 coulombs is 1 second, and the rated energy is 3020 watts than that ultracapacitors discharge cycle would only last 7 seconds.. My electronics skills are terrible, I just know a lot about nuclear power.

[Chris Bradley]

Spencer, you really are not taking some sledge-hammer sized hints that you need to go read a book on electronics basics (and some elementary text on physical units).

Once you have figured out/found out that capacitance is a function of charge over voltage, and that energy is a function of capacitance by the square of voltage [a half of it, to be precise] then maybe all this will become clear and you can move on from your day job of nuclear power and move into electronics.

[Richard Hull]

Being a person who is used to pulse discharge work, these capacitors are not meant to be used for high peak currents over short periods. They are more battery like, yet suffer far more horribly at being mishandled than a battery.

The construction of these are inherently weak due to a number of factors, thus, the 2 year shelf life while doing nothing. Lead acid batteries do bettter. The value here is found in their data sheet's suggested uses and the energy stored per unit mass over the lead acid batttery. These ultra caps shine for portable apps where weight is the defining factor and is perhaps their single saving grace.

Intelligent useage of the wonderful features in the super cap is based on a complete knowledge of what it can and can't do. Data sheets are deliberately slanted to show the best features of the component at its extreme limits. These limits often demand a derating of all the other related specs found in the sheet. (tradeoffs)

You would never see a tiny fraction of the rated cycle life if you brutalized the capacitor with monstrously heavy discharges over very short periods.

In short, these are not a solution for giant pulsed coil operation or they would have noted it as a feature in there suggested applications. Reverse induced voltages might just give a life span of one cycle if dumped into the right coil.

Lots of technology and ground level experience needed here to apply any capacitor in every situation.

Have you priced these?

Richard Hull

[Spencer DePue]

Thanks for the advice Richard, I appreciate it. On Electronics Goldmine, I could buy a 9 ultracap array rated for 73 kilojoules of storage for $136.00. A car battery is between $50-80 bucks depending on the quality. I guess the attractive portion of the ultracapacitor is its compactibility. In order to power a plasma core solenoid, what would you recommend for the power source, car battery or slightly faster discharging ultracaps? I'm looking for more power draw for every second.

[Rich Feldman]

Spencer,
The car battery you mentioned probably stores over 3,000 kilojoules. And you really do need to learn and understand physical units of current, power, time, and energy.

As Richard said, ultracapacitors occupy a middle ground between conventional capacitors and chemical batteries.

Batteries have the highest energy density but relatively low power density. Discharge times range from minutes (e.g. for RC racers, or hybrid automobiles) to many years.

Most regular capacitors can be fully discharged in less than a millisecond, with low energy density but high power density.

Ultracaps have an intermediate energy density and power density. Discharge time frame ranges from a fraction of a second (e.g. for a powerful car audio system) to seconds (e.g. to accelerate a hybrid city bus after one regenerative braking cycle).

[edit] I looked up datasheets for the 73 kJ bank you cited.
It could deliver 1000 watts for 70 seconds (as could the car battery, time after time without recharging);
the capacitor voltage would drop from 22.5 to 5V as current rises from 44 to 200 amps.

What the cap bank can do that the car battery probably can't do is give you 18kW for about 2 seconds without exceeding datasheet limits. (models below includes ESR)

[Dan Tibbets]

Back to IEC vs Tokamaks-
You mention increasing the density or pressure of a plasma, simply by pumping more energy into it. While this might work in the very short term where inertia confines the plasma (like a bomb), but when more than a few nanoseconds are involved, you need some other type of confinement system. Usually this is a magnetic field (Tokamak). In an IEC machine like a fusor, the ion confinement (or electron confinement- you cannot do both at the same time) is driven by electrostatic fields. Various attempts have been tried to contain both charged particles species with combination of electrostatic fields and magnetic fields- various Mirror machines. Perhaps the most successful derivative of both the magnetic and electrostatic confinement combined in one machine is the Polywell. Research continues with this machine.

The problem with purely magnetic confinement, is that the heavier charged elements are not contained well in small machines due to the ion's (heavy particles) gyroradius, and associated ExB drift through the magnetic field. To contain ions in such a fashion for adequate amounts of time requires very large magnetic fields- thus the ever growing sizes of Tokamaks. There are other problems, but this is the primary factor that determines the size of Tokamaks. The Polywell also uses magnetic fields to contain one of charged particles, but in this case they are only trying to contain the electrons. Because of the electrons much smaller mass and associated gyroradius, the ExB drift is much smaller, and thus a much smaller machine can contain them for equivalent times. In the Polywell, the ions are primarily contained by the electrostatic field developed from the excess electrons that are contained magnetically.
There are all sorts of confusing interactions involved. Whether the system actually works is yet to be determined.

As far as pumping energy into a system, in Tokamaks, this is a difficult process (I'm not sure why), while in a fusor, it is easy. Remember temperature is equivalent to speed. Electrostatic fields easily accelerate charged particles.. In Tokamaks (perhaps partially due to their large sizes and resultant surface areas that can lose energy) it is a fight to get the plasma hot enough, and to economically maintain the energy against losses you need to reach ignition conditions, where the resultant fusion, adds enough energy to keep the plasma hot , so that the input energy can be essentially turned off. The input energy needs to be so great to heat and maintain the plasma up until the fusion energy surpasses it, that positive energy gain requires this recycled energy. Thus the fusion products need to be contained also, so that they have a chance to transfer energy to the fuel plasma. This is apparently not needed in electrostatic machines. Here it is mostly a matter of decreasing losses and or magnifiying the fusion rate (mostly through increasing the density). I suppose an electrostatic machine might reach ignition conditions, but is not required. Apparently the Polywell cannot contain these higher energy ions (the electrostatic field/ potential well is not strong enough) so it is good that ignition is not required. This also introduces alternate methods of extracting the fusion energy for useful purposes (direct conversion) which is not possible with ignition machines (or at least more difficult).


Actually, using electrostatic means may play a role in magnetic machines like a Tokamak, not for general containment, but to suppress macro-instabilities.
There are some hints that this may be part of the TriAlpha Reverse Field Confinement scheme.


Concerning the strength of magnetic fields in coils (or transformers). Apparently the strength in iron cores can be surprisingly high. But the strength in air cores is substantially less with the same amp turns. And it is this air (or plasma) core that is applicable to containment in fusion machines. The link illustrates the efforts needed to achieve a 20 Tesla field in an air core selenoid. What would be the strength if it had an iron core?

http://www.ru.nl/hfml/research/levitati ... _solenoid/


Dan Tibbets

[Richard Hull]

Golf cart batteries are massive. They are normally 6 volt and are especially made for ultra deep discharge and long periods at high currents. Before that though, I would check into surplus filament transformers especially for klystrons. Some of these, while rare, are easily rated at 5 to10 volts and 300 amp continuous. An appropriate full wave bridge rectifier on one of these with super filter caps would be a forever supply of current.

You will first have to figure just what you are going to need in the way of current and balance the amp turns in your air coil. Obviously, you seek the largest number of amp turns with a minimum of overall inductance and limited current to achieve the air core field desired in the solenoid. Lots to know before you go.

Richard Hull

[Richard Hull]

Iron cores are easily saturated and allow a low reluctance path to concentrate the magnetic field in a solenoid to intense levels based on energy spent per tesla in the iron, but there is a limit, a stopping point, if you will beyond which no more magnetic flux can be achieved.

Air has virtually no imaginable limit in the number of tesla it can support, but is the highest reluctance path known next to a vacuum. Thus the energy spent per unit tesla achieved in air within a solenoid is vastly greater than that in iron.

Richard Hull

[DaveC]

I've used a pair of 12 volt lead acid truck batteries in parallel, to put out upwards of 2000 Amps for several second bursts - all day long.. A standard battery charger, easily kept up with the low average current draw.

Fully charged, with the acid specific gravity on the high side of 1.260 the internal resistance of the typical lead acid battery is in the vicinity of a couple milliohms. This governs the maximum "zero resistance" current available, and obviously varies with the physical size and spacing of the plates. Usually, the Gel -Cell designs give the highest currents.

Paralleled Ni-Cads or Alkalines, can probably give over 1000 Amps if done properly.

The surge impedance of the capacitor similarly limits the available current upon short circuit. It can be approximated by assuming a strip line with the dimensions of the dielectric thickness and electrode geometry. This ignores magnetic coupling between electrodes during the discharge... aka inductance which may significantly raise the overal impedance.

Plates in parallel lower the overall impedance, plates in series add to it... all kind of obvious....

To get really high currents, a little of this noodling will quickly indicate you need to have very very low impedance connections, good impedance matching along the way, and no reflections at the final destination. None of this is the slightest bit trivial, but definitely good for a refresher course in E-M and transmission line theory.

Dave Cooper