When your objective is to generate a large number of high-voltage plasma-igniting flashover events, degradation of the ignitor is a problem; this design minimizes the problem.
When many engineers think of “power” they associate it with the basics of getting a power supply that can deliver the needed voltage rail at the required current level and do so efficiently. While that may be the objective for many applications, there are some where the supply itself is a critical point of the system design and function, as it must provide an unusual voltage/current/time profile to the load. These situations have unique electrical, mechanical, and materials aspects, and mandate a deep understanding of the load’s characteristics. Published papers on their design and operation provide interesting reading and a sense that maybe your problems are not so difficult.
That’s what happened when I came across the Purdue University project-research paper published in Plasma Research Express, “Low energy surface flashover for initiation of electric propulsion devices,” which is technically situated at the junction of two fascinating topics – CubeSats and plasma ignition. Normally, I don’t associate the tiny CubeSat satellites (Figure 1) with topics such as plasma ignitions, so right there I learned something worthwhile (see References 1 to 3 for more about CubeSat).
Figure 1 Artist rendering of Montana State University’s Explorer 1 CubeSat (left); a CubeSat created by the University of Michigan called the Michigan Multipurpose Mini-satellite (M-Cubed) (right). Note the use of a standard hardware-store tape measure as a foldable pop-out antenna; this is commonly done. (Source: NASA/JPL-Caltech/Montana State University via physics.org)
The paper discussed the challenges of developing a viable propulsion system for CubeSats. Chemical-based propulsion offers large thrust but also brings major concerns due to its need for large propellant mass, high temperature and pressure, and a threat to the main payloads posed by the reactive propellant materials. An alternative is electric propulsion which has very high exhaust velocity and fuel efficiency, in principle, and includes technologies such as pulsed plasma thruster (PPT), miniature Xenon thruster, electrospray, and vacuum arc thruster (VAT). At present, these propulsion systems are still too new to be used and are rated below 7 in the Technology Readiness Level scale used by NASA (Reference 4 and 5).
These electrical propulsion systems need ignitor subsystems to initiate discharge (Figure 2). There are many available techniques to ignite a discharge in a vacuum, including gas injection, high-voltage breakdown, mechanical actuators for drawn arcs, fuse-wire explosion, vaporization of conductive coating between the anode and cathode, and ignitor plugs in a pulsed plasma thruster using point ablation of semiconductor layers. Most of these triggering mechanisms have one thing in common: they operate by providing a “seed plasma” to bridge the electrodes and initiate the discharge; each has relative pros and cons, as you would expect.
Figure 2 This is an artist rendering of a device that electromagnetically accelerates plasma to produce thrust; to create the propellant plasma, the Purdue group developed a low-energy surface flashover (LESF) technique that is used at the beginning of the channel. (Source: Purdue University)
One particular type of vacuum-discharge triggering is a surface flashover, where two electrodes are separated by an insulating layer and the breakdown over the insulating surface is initiated when the applied high voltage exceeds the breakdown threshold. This phenomenon, of course, has been extensively studied with respect to high-voltage vacuum devices, where high-voltage discharge holdoff is critical and surface flashover and subsequent breakdown are undesirable effects. The surface-flashover voltage of insulators in vacuum depends upon many parameters such as materials, geometry, processing history of the insulator, the applied voltage waveform and its duration, and the number of previous flashovers, among others.
The problem is that the ability to consistently generate a defined, predictable flashover event is also a function of how many previous flashover events have been initiated, among these other factors. As a result, the ignitor degrades with use, and that’s unacceptable for the CubeSat application. The Purdue team, therefore, set out to modify traditional surface flashover via a significant reduction of the energy of the individual flashover event, and so allow for a large number of flashovers using the same electrode assembly, and without significant damage or degradation to the assembly.
They called their proposed arrangement the low-energy surface flashover (LESF) device (Figure 3). Their paper details the experimental set-ups, the tests they ran, and the results. Remember that in this world of flashover events, things happen fast, and each test tells a new story, so you have to have a good idea of what you are doing and how you intend to grab the data before you initiate a flash event. The paper also discusses the changes of their ignitor as they exercised it. They were able to achieve over 1.5 million flashovers with minimal degradation (Figure 4).
Figure 3 This is the flashover assembly used in experiments after conducting over 1.5 × 106 flashover events. (Source: Purdue University)
Figure 4 Evolution of the breakdown voltage of the electrode assembly (Vbr ) over its >1.5 × 106 flashover events (N)
If you are “bored” with supplies providing steady-state DC at single-digit voltages and milliamp/nanoamp currents, there’s a whole other world out there which depends on controlled material breakdowns. Reading about these will give you some appreciation of how pulsed high-voltage engineers live and work, that’s for sure.
Have you even been involved with high voltages (>1000 V) and events? Was this with steady state or pulsed events? Was it an intentional occurrence or not?
Bill Schweber is an EE who has written three textbooks, hundreds of technical articles, opinion columns, and product features.