Plasma-Powered Thrusters For Your Homebrew Satellite Needs

It seems as though every week we see something that clearly shows we’re living in the future. The components we routinely incorporate into our projects would have seemed like science fiction only a few short years ago, but now we buy them online and have them shipped to us for pennies. And what can say we’ve arrived in the future more than off-the-shelf plasma thrusters for the DIY microsatellite market?

Although [Michael Bretti] does tell us that he plans to sell these thrusters eventually, they’re not quite ready for the market yet. The AIS-gPPT3-1C series that’s currently under testing is designed for the micro-est of satellites, the PocketQube, a format with a unit size only 5 cm on a side – an eighth the size of a 1U CubeSat. The thrusters are solid-fueled, with blocks of Teflon, PEEK, or Ultem that are ablated by a stream of plasma. The gaseous exhaust is accelerated and shaped by a magnetic nozzle that’s integrated right into the thruster. The thruster is mounted directly to a PCB containing the high-voltage supplies and control electronics to interface with the PocketQube’s systems. The 34-gram thrusters have enough fuel for perhaps 500 firings, although that and the specifics of performance are yet to be tested.

If you have any interest at all in space engineering or propulsion systems, [Michael]’s site is worth a look. There’s a wealth of data there, and reading it will give you a great appreciation for plasma physics. We’ve been down that road a lot lately, with cold plasma, thin-film plasma deposition, and even explaining the mystery of plasmatic grapes.

Thanks to [miguekf] for the tip.

22 thoughts on “Plasma-Powered Thrusters For Your Homebrew Satellite Needs

  1. any idea what kind of thrust or specific impulse these things is? officially its “to be determined” but what are the typical performance characteristics for this kind of thruster?

    1. Thanks for your question! Unfortunately, the short Hackaday article can’t go into the immense detail for all of the work put into these thrusters. Currently, my newest gen-3 thruster, with fully integrated electronics, has not been qualified for impulse bit and thrust yet, having just passed high vacuum ignition testing for the first time literally this past Friday. However, my prior gen-2 prototype, the AIS-gPPT2-1C thruster was in fact qualified with impulse bit, thrust, and lifetime measurements. I built a simple micro-pendulum test stand, and knowing the length, mass, and displacement, was able to calculate thrust. Impulse bit for that thruster ranged from 0.78uN-s to 3.52uN-s, depending on input energy from the main cap bank, which varied from 0.40J to 0.84J respectively. Total lifetime was 500 shots, however the newest AIS-gPPT3-1C thruster is built to greatly increase lifetime. I have published full testing reports, technical details, and data on my website if you are interested in looking at the current parameters. In general, these types of thrusters can widely vary in performance. This particular thruster is unique in that it is possibly the lowest energy PPT ever designed and fired, with a total bank energy of 0.09J for a 1kV main bank voltage. Typical performance however is in the uN-s range of impulse bit, where thrust can be calculated by multiplying the rep rate by impulse bit. This current thruster is designed to operate at a nominal rep rate of 1/3-1/4 Hz. I regularly post updates to the Applied Ion Systems Twitter and Instagram in addition to the website if you are interested in following along the development. This effort is currently the first and only fully open source at-home advanced electric propulsion testing and development program of its kind out there, and there are some very unique features about this thruster that make it the first of its kind in the pulsed plasma thruster category of electric propulsion.

      1. That’s very cool stuff and looks like a lot of fun.

        Inferring a bit here pulls out some interesting numbers:
        3.5 uN-s will give a 100g satellite a delta-V of 35 micrometers per second.
        A 500 shot lifetime, or total impulse of 0.0018 N-s, will give it a total deltaV of 1.8 cm/s.

        It would be really nice to know the real specific impulse. A total deltaV of 1.75 cm/s can be had with a lot less trouble than a plasma generator.

        The newer one, running at 4 Hz, 0.09 J/shot requires 0.36W of (output) electrical power. Roughly speaking, that needs about 100 square cm of solar cells. So some kind of unfolding panel will be needed for a 5x5x5 cm satellite, though a 1U might manage fine.

        I can’t help but wonder how this compares in mass/size/efficiency/complexity with the Accion Systems’ TILE 50 thruster, which claims a specific impulse of 1250s (or, in rational units, about 12500 N-s/kg), a minimum impulse bit of 0.5 uN-s, and a total impulse of 20-60N-s, in a 50g package 25cm^3 in size.

        1. Paul,

          Thanks for your input! Realistically, the new version will probably operate at around 0.5uN-s. The 3.5uN-s for Gen-2 was only for highest energy shots, and due to power budget restrictions for a 1P PocketQube, not sustainable. However, I do have a design for a higher-energy version for 2P and 3P PocketQubes I will be releasing sometime in the near future.

          For this module, rep rate is limited to the charging supply. I am using an EMCO Q-series 2kV supply, underdriving it at 3.3V instead of 5V to reduce power consumption and make integration easier, looking at tradeoffs between charging time, energy, etc. Due to this, nominal operation will be limited to 1/3Hz at most, maybe 1/2Hz pushing it. Average power draw at this rep rate however was simulated to be 163mW. In this case, a 1P PocketQube with standard side-mounted non-deployable high efficiency solar cells should be able to manage it fine. The goal is on the order of thousands of shots, with potentially up to 10k shots or more ideally. The limiting factor to the prior Gen-2 lifetime was purely a result of Teflon fuel charring. I will be releasing a full end-of-life analysis report on this. However, I believe that the shot energy was far too high for the small fuel bore size, which caused excessive charring and eventual shorting of the thruster.

          It is interesting you mention Accion Systems’ TILE thruster. I have also looked into their tech extensively as well. The first key difference here is cost – a module like this will be on the order of ~$1k, where a TILE thruster would be most likely orders of magnitude more. The second is that I do not believe there is a current TILE module actually designed specifically for PocketQubes – this module is in fact meant for PocketQubes. In general, based on literature, low-energy electrothermal PPTs will have a specific impulse around 300s or lower. Lifetime of the TILE thruster is also significantly higher, and PPT technology cannot compete in terms of total impulse with electrospray technology such as ionic fluid electrospray or FEEP. However, such thrusters are made for high lifetime, high critical applications. This thruster is meant as an ultra-low-cost solution to start breaking the barriers of propulsion at this level. Complexity is another key difference – electrospray like TILE requires advanced micro-manufacturing technology, and is significantly more complex not something that could be built at home. The AIS-gPPT series thrusters however are specifically designed in mind for very easy and simple manufacturing. In fact, the gPPT1 and gPPT2 were both built on the table with nothing more than a dremel and hand tools!

          However, I have started an initiative to look at ultra-low cost simplified FEEP thrusters for PocketQubes as well. I haven’t released the details yet, but have preliminary designs for low-power, easy to manufacture liquid-metal ion spray thrusters.

          Really, the key comes down to end use. You wouldn’t use this thruster on a government satellite, whereas TILE would be a viable candidate. However, current propulsion technology like TILE is still far to expensive and still not quite scaled down in terms of available power and size specifically for the PocketQube community. My hopes are this thruster will change this, and allow for propulsion to emerge and become standard on these ultra-small satellites, especially as the community continues to expand, and overall launch and development costs further drop. As is, PocketQubes have opened up an incredible door to space for low funded research groups, start-ups, and enthusiasts. Adding propulsion to the toolkit only improves their utility even more.

      1. 500 firings is relatively small for the thruster, however the article is a bit behind on my current work. The previous gen-2 version was tested to 500 shots. However the newest gen-3 thruster is designed for significant more. The primary purpose of this thruster is to provide propulsion for PocketQubes, which currently does not exist. This is the first fully integrated propulsion solution that is scaled small enough in terms of size and power available for such tiny sats. The average orbital lifetime of a PocketQube is around 6 months. By using propulsion to increase orbital lifetimes, value per launch is significantly increased, and opens up new possibilities and mission capabilities that has not been possible prior to propulsion. By taking an open-source approach, accessibility to propulsion for the community radically increases as well. In terms of propulsion, open-source is unheard of in the field – everything is currently behind tightly guarded walls.

    1. Pulsed plasma thrusters, or PPTs, are already scaled up much larger and well demonstrated as one of the oldest electric propulsion technologies out there. The have been used extensively in maneuvers and propulsion in space on larger satellites and Cubesats. However, the current major challenge is actually scaling them down for nano-satellites. The new AIS-gPPT3-1C thruster module represents the first available, truly scaled complete propulsion package specifically designed for PocketQubes, which are even smaller than Cubesats. Currently, no propulsion exists for PocketQubes, and with the advent of propulsion for the PocketQube community, new possibilities and mission capabilities can be realized, increasing both performance and per-launch value. The primary goal of these thrusters will be to increase average orbital times for PocketQubes, as well as controlled de-orbit.

  2. This looks exactly like the plasma MCAT thruster technology pioneered by Micheal Keidar at GWU states “2000-3000 sec with a Delta V=300m/s for 3 Kg satellite and 30 gr of propellant”. These only work in vacuum and the force is very small but for a larger thruster configuration, there is enough Delta V once you’re in orbit to send a 6-12u cubesat to the moon. The thruster is fundamentally a spark gap transmitter from anode to cathode that ablates (i.e. erodes) solid material from the cathode. The nice thing is the ablation is very uniform so the metal propellant can be almost all consumed. The trick for higher impulse that the folks at GWU figured out is accelerating the ejected material using a magnetic field that is induced during the discharge. The one downside is high power consumption.

    1. I have actually seen a lot of work by Micheal Keidar. However, this thruster is a bit different. What he is working on is triggerless Vacuum Arc Thrusters, or VATs. These are also small ablative electrothermal-plasma based propulsion, but slightly different principles than pulsed plasma thrusters (PPTs). While this current iteration uses Teflon fuel, I will be exploring other plastics as well as low-melting point eutectics such as Bismuth-Tin. This PPT uses a trigger electrode still, however it is in a very unconventional configuration that standard to help streamline physical design. The whole PPT is actually built from flat-stacked plates, which is highly unconventional and I believe the first of its kind in this configuration as well, in order to minimize required space and make it easier for mounting. The key difference here is that this thruster is the first of its kind to be scaled small enough and low enough power for PocketQubes as a complete module. Only 40mm x 38mm x 24mm, and consumes less than 1W peak power, about 163mW average power during the charging cycle. At only 0.09J stored energy for the gen-3 version, it is possibly the lowest energy PPT ever designed and fired. This thruster and full vacuum infrastructure is also completely home-built and qualified, with no external ties to academic research labs or propulsion companies, and designed at significantly reduced cost. The goal is to bring electric propulsion to the maker community, and lead advances in propulsion using an unconventional approach.

    1. Thanks! The Libre Space Foundation has greatly helped introducing me to the open space community, and is in part why I have pushed development of these thrusters this far to begin with! I am very eager and excited to meet members of Libre Space at the upcoming OSCW this October in Athens, who have also helped make this opportunity a reality for me by sponsoring me to fly out to give several talks on my work! I definitely hope to future collaboration and making open source thrusters a reality for the community, and getting them flown in space one day!

  3. A resonant cavity version is tempting but since the space anchor effect is relative to the root of the volume, they would lend themselves to something larger.
    I’ve been debating an inflatable resonate cavity so it would fit in one of these micro satellites and expand massively once deployed.
    But at least for the time being there’s too much space junk to avoid.

  4. I think one major thing to point out here, possibly the most important aspect to this entire effort that the article does not actually touch upon, is the fact that these thrusters, and all of my R&D efforts, are 100% fully open source. Design details, CAD, PCB, schematics, test reports (regardless of success of failure), testing infrastructure, instrumentation, full step-by-step picture galleries of builds, etc are excruciatingly documented as I go along, which represents the first true, intensive open source independent at-home based propulsion program out there. At this point, advanced electric propulsion has only really been tackled in an academic lab/university environment, or through high-tech R&D companies – millions of dollars of invested testing infrastructure and state-of-the-art facilities supporting work that can go on for over a decade before even remotely entering the market. The goal of these efforts is not only to provide very low-cost propulsion solutions for PocketQubes, which currently do not exist, but to engage with and provide the community with full details and resources on electric propulsion in general, and help lower the barrier of entry into the field and technology by showing others how such advanced electric propulsion research and design can be done with minimal tools and resources. Using a very unconventional open-source approach to the field, which is currently shrouded in very high entry-barrier costs and closed doors, new advances can be rapidly iterated, despite significantly less resources, and can help inspire and encourage other enthusiasts in the open space community as well, and promote open collaboration with other makers.

    Full details and updates are regularly posted to the Applied Ion Systems website, Twitter, Instagram, and most recently YouTube, where you can follow along with these propulsion developments. I live-tweet every propulsion test, and will eventually be moving to full live-streaming of propulsion tests, to allow enthusiasts in the community a first-hand chance to experience real propulsion testing and breakthroughs directly.

    Advances in space technology and advanced electric propulsion CAN be accomplished in an open-source manner, even at home!

    1. James,

      Thanks for your feedback! It’s great to see some excellent discussion evolving on the subject and to be able to engage other knowledgeable enthusiasts and makers on this effort! More collaboration in advanced open source research areas such as this can only benefit everyone as more people get involved!

  5. Some quick updates on the propulsion system development. I now have some performance characteristics for the new AIS-gPPT3-1C module. Recently completed impulse bit measurements on the thruster, which is so far performing at around 0.65uNs impulse-bit, which at 0.33Hz, correlates to about 0.22uN of thrust. I am already on V3 of the board, which should hopefully be the first flight module. The new V3 board includes upgraded high performance pulse rated capacitors to replace the older caps to improve lifetime issues and performance. I will probably be reducing total bank energy from 0.09J to about 0.06J, and increasing rep rate to about 0.5Hz. This will reduced I-bit by a little, however I should see a gain in total thrust to about 0.32uN. Once the new V3 board is built next week, I will start lifetime testing of the thruster. In theory, assuming an average of around 1 microgram of fuel ablated per shot, the thruster could have a lifetime of around 100k shots assuming no other issues or failures, but this will have to be fully qualified with extensive testing.

    1. This is such cool stuff. Conceptually, it’s so simple: an arc discharge to ablate a small amount of propellant to produce a momentum impulse. I love the idea.

      I couldn’t help inferring from the numbers provided the usual parameters of interest:

      Total lifetime impulse = 100 k shots * 0.65 uNs/shot = 0.065 Ns. Over the approximately 3-day thruster lifetime, this would be sufficient to change a 100-gram satellite velocity by 6.5 cm/s.

      At 1 microgram per shot, 0.65 uNs impulse implies an ejection velocity of 0.65 uNs / 1 ug = 650 m/s, or a specific impulse of 65 seconds.

      That 1 ug at 650 m/s is an output energy of 0.2 mJ, from an electrical pulse energy of 90 mJ, for an efficiency of 0.2 percent.

      It’s clearly a fertile area for R&D, with plenty of scope for dramatic increases in performance. I’d love to see what the factors are that are currently limiting performance.

      1. Paul,

        Thanks for your input! I will definitely need to get a measure of actual ISP at some point, although that parameter is a bit out of reach for me to measure currently due to my limited resources. Hopefully I can at least get some ion velocity measurements eventually with a simple time-of-flight setup. 1 microgram is a decent starting estimate on ablated mass per shot, though realistically I suspect it will be much less than this. If I can get access to a microgram-precision scale, I should be able to confirm the actual ablation mass and ISP. Once I have more accurate ISP measurements, it will definitely give a powerful indicator of performance parameters comparing it with improved versions as I progress. It will be very interesting to see how the other two fuels that I will be starting with after Teflon qualifications (Ultem and PEEK) will compare, and if this improves anything at all.

        Something else of interest to also look at eventually would be energy density on mass ablation vs. performance. For the AIS-gPPT3-1C, the starting bore area is 0.24 cm^2. With an energy of 0.09J, this correlates to an energy density of 0.38 J/cm^2. My previous Gen 2 version, operating at a max of 0.84J, with a starting bore area of 0.38 cm^2, had a max energy density of 2.22 J/cm^2. It only lasted 500 shots due to fuel charring, which I suspect is in part due to too high energy density. The Gen 3 so far has fired 130 shots with no charring or visible wear. At the new reduced energy of 0.06J, energy density should be even lower. It will be interesting to see how this affects ablation and performance based on material properties.

        On thing that I strongly suspect is a limiting factor of its performance is the unusual aspect ratio. Having such a short discharge length and very unconventional aspect ratio compared to conventional PPT designs will make performance suffer and reduce efficiency, which is unfortunately the price to pay for miniaturizing the system to be use-able with the incredibly small PocketQube space constraints. It is definitely exciting that there is potential for improvement however, and I am sure as more data and test results come in as I progress the design will improve over time!

        Thanks again for your input! I think your extrapolations make sense and are reasonable indications of performance. Hopefully I will have more numbers for you and others to analyze!

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.