Progress in Switching – all capacitors firing, but more reliability needed
(6) Comments[From Eric Lerner’s LPP report of February 23 - sorry for the posting delay]
We have continued to make progress with the switches.
This month we succeeded for the first time in getting all 12 capacitors to fire during the pulse, firing at 24 kV. However, we cannot yet do this reliably and repeatably.
At the same time, we have also created conditions in which most of the bank, 7-9 switches, will fire without pre-firing, allowing us to perform the experiments described in the first section.
What we still need to do is to get repeatable and full-bank firing together.
The re-profiling done by the switch manufacturer, R.E. Beverly, did not resolve the pre-firing problem. Neither did our attempt to go to higher charging voltages, up to 35 kV, which the literature shows should make such switches fire better. We ran into significant problems at the higher voltages (see below).
We determined that part of the problem was that, in most spark gap designs, the trigger is located between the two electrodes, not within one electrode as with this design. The existing design meant that it was easier for current from the trigger to short out to the adjacent electrode, rather than going across the gap to trigger the capacitor. We modified the switches to allow the trigger head to stick out a small amount into the gap. In addition, we carefully measured and adjusted the trigger heads and top plates to reduce the variation in the spark gap (the distance between the trigger and the opposing electrode) from almost 10% to about 5%.
Together, these two steps greatly reduced pre-firing and allowed us to get one shot with all 12 switches firing. However, the gas pressure in the switches that allowed all capacitors to fire also allowed too much pre-firing. So we had to increase the pressure, which allowed reliable firing of 7-9 switches, giving us the 0.5-0.6 MA used in the angular momentum experiments.
We intend to further adjust and measure the bottom plates of the switches to reduce gap variability to about 2.5%, hopefully getting full bank firing on a reliable basis.
Fusion Confidence Monitor
LPP and PPPL Informal Collaboration
Comments
For a more in depth discussion, start a thread in the forums.Why not going for a 1% or 0.1%. Fusion power always seems a month away :(
Hi, just to know, why not use one or several Thyratrons in parallel, instead of this Spark-gap switch ?
(or a crossatron, but newer technology = more problems)
Thyratrons seem reliable, at least much more reliable than spark-gap switches given your experience. They have been used since many years in the industry with success.
Sorry if the answer is obvious for you, I am not a plasma specialist;-)
Personally I would go with traditional design and put the spark gaps between the electrodes. Also I would want to be able to tune the distance between the spark gaps for higher repetition rate discharges given a constant current charging source. This is due to the RC time constant involved when charging the capacitors, the capacitors furtherest away will have a resistance from charging source equal to (resistance per stage) times (number of stages) which reduces their charging time which goes as RC. So to get everything to discharge simultaneously this requires reducing the gap length in accordance with the voltage across the capacitor electrodes and the dielectric breakdown strength of the gas your filling it with. But the truth is you want to eliminate the spark gaps. Each spark gaps is a plasma. You want all of your energy dissipated in the focus device not your spark gaps. I would want to discharge the spark gaps in my focus device if that is possible by designing a focus device geometry. Anyway… zeroth order calculation says energy will be dissipated equally in all of the spark gaps, that includes the focus device. So if you have five stages, fusion power will have to be 5x input. Theory is only indicating 1.4 at this point. So in the long run a marx generator may not work for you.
I would arrange the capacitors in a circle with electrodes pointing toward the center kinda like old cray memory banks. This would minimize the distance electricity has to propagate and minimize the inductance. Using a zeroth order guess, 1 foot is 1 nano second. Also the UV from the first spark will help ionize the air for the adjacent gaps thus assist in firing all banks simultaneously without the need for laser ionization or UV LED triggering.
I don’t know enough about your geometry and power plan but perhaps designing it so that your not using a marx generator is the way to go. have each electrode coupled with a separate capacitor. And use an ionizing laser or UV source to excite the gas within uniformly to trigger plasma creation. This way all energy is dissipated within the focus device and not your spark gaps! A little extra jolt from a laser or UV discharge source couldn’t hurt either but maybe it does due to EMF or polarization. I would image that the focus device would have to be re-engineered for lower working voltage and maybe higher pressure. Maybe higher pressure would suffice.
@omodes: picture, for geometry, the whole device as: a metal plate ~2 m², in centre of which is welded a pipe (the anode); given a suitable gap, there is another plate, parallel to the first, connecting sixteen cathodes that surround the anode; the pulse arrives from twelve capacitors, that are arranged all around the outside, each the size of a beer keg, and the switches are mounted directly to the capacitors. the atmosphere for the switches (SF6) need not be the same stuff or pressure as in the focus device (borane+helium). check the gallery for pictures… http://focusfusion.org/index.php/gallery/image_med/65/
http://focusfusion.org/index.php/gallery/image_med/109/
From one of the emails on your site, I read about the Sandia progress with their Z-pinch at: “http://www.sandia.gov/media/z290.htm”
In that paper I read about: “Highly synchronized laser-triggered switches allow the stored energy to be discharged simultaneously through the 36 cables, each as big around as a horse and 30 feet long, arranged like spokes of a wheel and insulated by water.”
I went on the internet and found the following paper about laser triggered switches: http://www.sandia.gov/pulsedpower/prog_cap/pub_papers/053382c.pdf
The switches are made by Ktech: http://www.ktech.com/Pulsed_Power/Main/Pulsed_Power.html
They look very expensive and government contract oriented. And, maybe there is something in these references that will trigger a solution for the LPP switch problems.
Post a Comment