Controlling Current and Power to the Fusor

[Jeff Robertson]

I've been tinkering with atmospheric plasma, testing threshold voltages at various pressures, current levels, etc. Reading up on various fusor tests on these forums have led me to believe that, for fusion, ideally you want at least 20 kV and 20-50 mA going to your fusor.

What I've found is that as I crank up my variac, the current will hover around 20-40 mA, depending on the pressure. As soon as a plasma ignites, however, the current spikes up to 150+ mA. I can continue turning up the variac a little bit past this point, causing the plasma to get brighter and the current to steadily increase. Somewhere between 150-200 mA a safety switch in the transformer is triggered (the transformer's only rated for 42 mA), causing the power supply to shut off.

My electrical engineering knowledge is limited, but here's I think I've figured out (please correct me if I'm wrong). This is for a simple system, consisting of a transformer straight from the wall to the fusor.

- The input voltage, from the wall, is fixed (~110-120 V in the US).
- If the coil ratio of the transformer doesn't change, then the output voltage is also fixed.
- The outputted current depends on the output voltage of the transformer and the internal resistance of the fusor (I = V/R)
- The input current drawn from the wall depends on energy conservation (IV = IV).

I want less current to the fusor, and the only way I see this happening is if I can increase the resistance of the plasma inside the fusor. But from what I've seen the resistance of a plasma is a tricky beast, and depends on various parameters such as chamber pressure. Most of my plasma runs are done at 10-20 mTorr, which isn't that much higher than typical neutron-yielding runs.

So the question I'm trying to ask is, how do you guys control the current to your fusor? Do you guys hook up your high voltage and just see what Ohm's Law says? Or do you make conscious efforts to tweak the internal resistance of the fusor so that you can reach a desirable current?

I suspect the question I'm asking is fairly elementary, but I couldn't find much addressing it with the search function. Note that any above references to current or voltage are referring to RMS, as I realize the signal alternates.

Cheers,
Jeff

[KJNW]

Hello Jeff,

You sound at arm’s length like you are on the right track insofar as controlling the current by controlling the resistance of the plasma. It’s some of the numbers you are quoting that leave me a little puzzled. I will begin with the finest document on tuning the fusor that I have found on this board so far, and I suggest you chew it slowly as it has a great deal of salient truth to impart. viewtopic.php?f=11&t=4808#p31430 That done, I will tell you that my team runs 30KV at about 3 to 5 milliamps with a tungsten grid, and we adjust the current primarily with the deuterium gas and vacuum controls, not the variac. As the pressure falls, the resistance increases, and then you can raise the voltage to maintain the same current. After a number of these adjustments, you will hit a point where you can go no further without melting your inner grid. Hope this helps.

Carl Greninger

[Chris Bradley]

Jeff, it's pressure that controls the current, not the external electrical parameters. You need to pump the chamber low, then inch up the volts and pressure gradually, manually limiting the current that way.

Richard has a FAQ which covers this process, and finer detailed discussions also appear in many other threads.

viewtopic.php?f=24&t=3111#p12621

[Steven Sesselmann]

Jeff,

If I can add one more point, and it might save your transformer.

A high value ballast resistor between your power supply and the fusor is the best way to gain some control. Vacuum can go from being a perfect insulator to a perfect conductor in no time at all, a sure fast way to blow up your power supply.

Search the forums for ballast resistor, and see what it turns up.

Steven

[Richard Hull]

The plasma is a conductor. Townsend breakdown and ultimately arc conditions are a function of how many current carriers (ions) there are in the discharge. Reduce the pressure and you reduce the number of gas molecules in the volume of the plasma, thus, limiting the possible current carriers. As the pressure is reduced,.... and it gets hyper critical between 20 microns and 5 microns,.... more voltage is needed to keep the plasma going. Current can runway due to any number of factors involving voltage, and pressure.

Fusion, that is easy to measure, just will not happen until about 20kv in a pure D2 gas environment within the fusor and usable experimenter fusion really requires over 25 kv with 30kv being the starting point for neutron research and activation as a must.

So, you begin to see the complexity of the task at hand. All folks who have done fusion and continue to do it, will tell you that operation to optimum levels is pure art guided by past knowledge, understanding and "owning" all the parameters and simple physics.

In short, we are working on the ragged edge of the physics involved; a razor thin line with a balance of current, voltage and pressure. As I have, and will continue to point out, in our fusor world, the operation of a fusor is an artform conforming to scientific principles that function only in a very tiny region of favorable conditions. Like driving a car or programming a microcontroller, it is a learned thing that demands hands-on action and mind focused attention and understanding.

Playing with plasmas, like you are doing, will tell you more about what is demanded than reading a thousand books and even more than some FAQs here, which really tell it like it is for fusor specific operation. Telling it like it is, however, doesn't put skills in you hands. It is just a good guide written up by someone who has "been there, done that". You have to put it into practice in a manner that works for you.

Richard Hull

[Jeff Robertson]

Hey,

Thanks for all the responses and sorry for taking so long to reply. I read through the thread you provided Carl, and took all the above posts into consideration. It sounds like my intuition was more or less on the right track, at least in terms of understanding the relationship between pressure and plasma resistance.

The other day I was testing the plasma at different pressures, trying to see if I could get a higher voltage at lower pressures. What I found, though, was that the current would rise at an even faster rate at lower pressures! This meant that the maximum voltage I could push the power supply to was even lower than at higher vacuum. I did several tests on this to make sure I wasn't crazy before turning off the power. I didn't have much more time to figure out what was going on. The only test I was able to do was apply an ohm-meter to the inner and outer grids to see if there was a short - it read infinite resistance.

So, what's going on? Have any of you had an issue of the current increasing more rapidly at lower pressures? The only thing I can think of is that the current is shorting somewhere where it shouldn't be, but if that's the case I'm not sure why it appears to be more susceptible to doing this at lower pressures.

Also, I saw in the thread Carl linked a lot of people were recommending using a ballast resistor to limit the current. I've been having trouble finding information regarding ballast resistors, at least in the context of amateur HV power supplies. Could someone explain if one would be useful for my power supply, and if so where exactly I would put it into my circuit? I attached a diagram of my current power supply. I have a smoothing capacitor in there as well but I haven't been testing with it. Does a ballast resistor limit the current to just the fusor, or out of the transformers as well?

Thanks for the help guys!

Jeff

[Chris Bradley]

I think Richard covered this quite well already.

The exact nature and behaviour of your set-up is down to several factors.

The ballast resistor serves a key function and should go in series from your cap to the fusor central electrode. It needs to be a high voltage, high power resistor of 10's to 100's of kOhms. As there is a breakdown in the chamber the volume will go from 'non-conducting' to 'conducting' on an essentially instantaneous basis. With a ballast resistor in place, as the current ramps up so it sustains a voltage across it, lowering the voltage across the chamber. If it lowers the voltage across the chamber too much, the lights go out and no current flows, thus no voltage across the ballast resistor and it's all across the chamber again, so, bing, the discharge comes on again.

Ultimately, it doesn't keep coming on and off (well, not usually, though it happens in certain circumstances) and instead the ballast resistor plays a ballet with the voltage across the chamber on a time-scale too small for you to notice. Effectively, the ballast resistor causes the chamber to self-starve itself of volts when it sucks too many amps, so that it does not enter this runaway that you are experiencing.

[Frank Sanns]

It could be that I was the one that first experimented with a ballast resistor and here is why I used one.

While getting my feet wet while running a fusor, I was getting the commonly reported flashing on and off at the edge of plasma extinction (gas too rare to support the discharge at a given voltage). This happens at voltages that are usually above 10 kv and gets worse as the voltage goes up. At the higher voltages naturally comes higher power (and higher heating of the inner grid which I will get to in a minute). The higher power means that when the fusor pressure/voltage/current cause the plasma to extinguish only to reform a moment later, huge swings in power result and give large RF spikes that interfere with the electronics of the neutron detectors. This gives counts where there are no neutrons so metrology goes out the window. Adding a ballast resistor prevents these harsh power swings that are hard on power supplies and give false neutron counts. As you can imagine, power supplies, distribution cable, and fusor geometry have a profound effect on abruptness of the power swings from the inductance, capacitance, and electrode surface area to name a few.

On top of that, as the voltage and current go up in the fusor to maximize fusion, heating of the inner grid gets greater and greater due to ion impacts. At some power level, the inner grid will heat sufficiently to start to boil off electrons (thermionic emission). When this happens, a large current can be conducted through the large stream of electrons streaming off of the inner grid to the outer shell. This causes a huge increase in current from the power supply and can be an effective, near dead short. Melting of the grid can happen in an instant and that means tearing down the fusor and replacing the inner grid.

Ballast resisters are not necessary in systems that have low capacitance and power supplies that can only put out on the order of a few hundreds of watts. With power supplies that are capable of putting out around 600 watts or more, some sort of ballast resistor becomes a valuable aid in minimizing power swings near extinction and melting of inner grids and over stressing power supplies.

Frank Sanns

[John Futter]

Jeff
Plasmas tend to exhibit negative resistance or nearly so.
As the plasma lights it creates an almost zero resistance between the electrodes -it does this faster than a power supply can respond in its feedback loop --as a result you tend to get pulsing from no current to current limit.
The answer is to put in a ballast resistor to put a real impedance in the psu line. If you know the exact parameters you may be able to make up this impedance with a combination of resistance and inductive reactance. Most amateur and commercial systems use resistance only with values between 1 and 0,01 of the availabe current from the supply at maximum voltage -ie 1Meg for 1 and 10K for 0.01 for a 30kV 30mA supply. the nearer to 1 the more stable the system but the more losses in the ballast resistor.
All this worked out by available voltage divided by available current for a factor of 1.
It pays to put the ballast resistor at the ion source end of any cabling to mitigate capacitance especially from coaxial type supply cables. Capacitance in the circuit makes the supply less stable as it introduces "lead" to the current waveform which the control electronics finds difficult to control /predict so it tends to over shoot--in this case the supply shuts down until it thinks it is in control again.

Real resistance is your friend but is a power waster. I suggest starting with a value of near one and reducing until stability becomes a problem. This will protect your supply from possible damage.

Most modern supplies have event counters in them that shut the supply down if instability continually occurs --this is to protect the multiplier and multiplier drive components from dI/dt and dV/dt causing internal heating damagedamage.

[Chris Bradley]

John Futter wrote:
> Plasmas tend to exhibit negative resistance or nearly so.

Is 'negative resistance' the same as 'negative resistance coefficient', in this context?

[Richard Hull]

Ballast resitors have always been used from day one on all power fusors. They are not needed on Neon transformer or oil burner powered demo fusor systems as these are magnetically shunted and act as a magnetic ballast of sorts.

I have used, and many others have used, obligatory ballast resistors in all real fusor systems since the late 90's. I have used a 60,000 ohm, 150watt wirewound resistor in a full wave non-filtered power supply since 1999. It is in the oil of my x-ray transformer where I attached it for cooling and corona suppression purposes. During my normal super fusion runs at around 40kv @10-15 ma, this resistor wastes only about 10 watts or so of the actual ~480 watts sent to the fusor. The extra wattage is for those high current runaway events already discussed here. It gives you time to back off the applied voltage without blowing the supply or melting a grid. (This assumes you are constantly monitoring current and watching the grid, visually.).

The above is data from a real system that runs and has run for about 12 years now. We are not really worried about wasted power in a resistor though in an amateur fusor since it is all wasted power when looked at against the fusion energy return generated.

Richard Hull

[Rich Feldman]

Chris Bradley wrote:
>> John Futter wrote:
>> >> Plasmas tend to exhibit negative resistance or nearly so.
>> Is 'negative resistance' the same as 'negative resistance coefficient', in this context?

Only with a stretch.

Resistance is the ratio of voltage to current, or slope of voltage with respect to current. In ohmic conductors, it is practically constant over many decades of voltage and current.

Negative resistance coefficient, typical of some conductive materials, means the resistance goes down as some other parameter goes up. Most commonly temperature, hence the temperature coefficient of resistance. Certain HV resistors have a specified voltage coefficient of resistance.. Their resistance change at extremely high voltages might matter in instrumentation, but the absolute and incremental resistance remain always positive.

Negative resistance (typical of glow discharges) means that as current increases, voltage decreases. So dv/di, the incremental resistance, is actually negative. You can't maintain a stable operating point with just a "voltage source" and such a load. V, i, and the absolute resistance v/i are still all positive at any operating point. Passive elements can't have negative absolute resistance -- that would make them electric power sources.

One could argue that the absolute resistance of a glow discharge has a voltage coefficient of large magnitude and negative sign. But that's not mainstream usage.