Current boost

[Paul W Fontana]

If I have a HV supply that doesn't provide as much current as I would like, could I add a second, lower voltage supply as a current source to provide a current boost? I'm thinking of something like in the schematic below, where R2 is chosen in such a way that V2 - V_plasma - I_boost R2 = 0, where V_plasma is the expected voltage across the discharge and I_boost is the extra current available to the plasma. R2 and/or V2 could even be made variable to explore the I-V characteristic of the discharge. (R3 may be unnecessary and may be undesirable if I want to keep the cathode grounded, but would prevent the entire HV potential difference from appearing across the low voltage leg. Ignore that "battery" label - I realized after I sketched that that the low voltage supply would still have to be at least several hundred volts to overcome V_plasma and drive current in the right direction.)

Since low-current many-kV HV supplies are a lot easier to come by (and less expensive) than ones that provide the 10's of mA needed for a fusor, it would be nice if this would work. Apart from fusors, it would also be nice for our glow discharge experiments, which typically operate at about 1 kV or so, but are limited by our power supplies to a few mA.

[Bob Reite]

IMHO, this is not going to work unless the two supplies have equal voltage and resistors R1 and R2 are equal resistance to encourage load sharing.
For safety reasons, you do NOT want R3 where you have it. You can insert a low resistance value resistor (say 10 ohm 10 watt) in the plus connection to ground on each power supply to monitor the individual current if desired.

[Paul W Fontana]

@Bob, thanks for your reply. You're right, on further thought R3 is a bad idea. Not to be argumentative, though, but I think an unevenly shared load is exactly what I want. The point is to let the HV supply provide the voltage and the smaller (in voltage, not necessarily power) supply provide the current. Can you explain further why you don't think it will work with asymmetrical supplies? Maybe I'm missing something.

The two legs together are basically a current adder; if I were dealing with small voltages I would use an op amp at the junction, but is there some equivalent to an op amp that works at tens of kV?

I think I'll try a scaled down version with one of our 1 kV plasma glow discharges to see if it works, unless someone sees any reason something might get fried that way.

[Dennis P Brown]

No, an uneven load isn't what you really want - if you have a high current, low voltage system combined with a high voltage but low current system, even if the low voltage system does provides a high current, the electrons so produced will be at the low voltage! Not worth the cost, effort and more likely to short the high voltage through the lower voltage system damaging it unless the system has a near similar voltage. Sorry but there is no easy solution - a cap system will provide a "pulse" current that could yield some neutrons for that duration but power is power and for a constant current system, there is no way around that limit.

[Richard Hull]

There are no workarounds on getting the energy needed for fusion, beit getting the needed HV or the needed current at HV. "It is the only game in town and it is rigged". (to quote Don Lancaster).

Richard Hull

[Paul W Fontana]

@Dennis Thanks, I knew there was something obvious I was missing, and your comment helped me see it: if V2 is not at least V_plasma, that leg will be reverse-biased and sink current instead of sourcing it (or at best, with diode protection, not contribute anything). So both V1 and V2 need to be HV. However, in that case, it does seem to me that the currents add, so at least it might be possible to string multiple lower-current HV p.s.'s in parallel to get more current capability overall.

@Richard Yes, of course there is no free lunch, but this was not a free lunch scheme. I was not looking to bypass the first law of thermodynamics; the second p.s. was intended to provide the extra energy. With enough power available from that supply and HV available from the other, it still seems like there should be a way to leverage its current up to HV, but maybe that just amounts to building a DC-DC converter, at which point one might as well just build a proper supply.

Still, when I get a chance, I think I'll try putting two of our 2 kV HV p.s.'s in parallel on one of our glow discharges (with some appropriate HV diode protection) just to see if it works. Thanks for the feedback!

[prestonbarrows]

Still, when I get a chance, I think I'll try putting two of our 2 kV HV p.s.'s in parallel on one of our glow discharges (with some appropriate HV diode protection) just to see if it works. Thanks for the feedback!

The golden rule of power electronics: things in parallel must have the same voltage and things in series must have the same current.

Putting supplies in parallel is possible, but it is a risky game for high voltage supplies. You have to remember Kirchhoff's voltage law, the slightest difference in the output voltage of the parallel supplies will drive infinite currents between them in the ideal world. In the real world, this current is limited by the power rating of the supplies and the wire resistances, but it is still not a situation you want to be in.

This becomes extra dicey with high voltage supplies where the topology is usually some high frequency transformer into a rectifier or voltage multiplier and smoothing capacitor. The outputs will always have some finite ripple (especially when drawing current) and the supplies will not be in phase with each other.

This means that in general, there will always be voltage differences between the parallel supplies causing unwanted currents to slosh around. This is on top of any differences in calibrations between the supplies etc. As others have mentioned, this is where you have to start throwing in protection diodes and other complications. It is doable, but expect to see some strange behavior and equipment failure especially if you gang up more then 2 or 3 supplies.

The proper way to parallel supplies is to have one 'master' supply that acts in current control mode and n number of 'slave' supplies acting in voltage control mode to match the voltage output of the master. High end supplies will have this capability built in. But again, high voltage supplies generally tend to not work well for this.


It is actually often easier to stack a number of lower voltage higher current supplies in series if you do it before the output is rectified. This way, you can couple the AC outputs into a stack of series capacitors; this decouples all the individual supplies so that they don't need to be closely matched in voltage or current.

An example can be seen here:
viewtopic.php?f=11&t=4887

[Dennis P Brown]

I agree with Prestonbarrows that Chris Bradley's method for using small, low current supplies to produce large current supplies is one of the best, and easiest methods. He has posted a lot on this issue and he is both very insightful and one of the most knowledgeable people to post here on this subject (and others areas as well.) His approach is just brilliant in its simplicity and elegance. You should see a number of his posts on this subject.

[Paul W Fontana]

@Preston Thanks for your input. I will of course be careful to protect the power supplies with HV diodes.

I had seen that thread by Chris Bradley while browsing a while back, but had forgotten about it. It's quite clever, and does accomplish what I was trying to get at.

[Richard Hull]

The real issue in all of this is....Has anyone actually done fusion and become a member of the Neutron club using any of these methods of obtain HV at the required energy? We will see if it is viable only once someone is properly inducted into that august body using these methods. It all sounds great and is theoretically very interesting, but those who have done fusion know the loading and operational issues in the bizarre ionized gas diode that is the fusor.

The described method of stringing weaker supplies in series-parallel combos would most certainly do a good demo fusor, but so will a simple neon sign transformer.

Richard Hull