Why can’t a tokamak just run forever? The short answer is that most of them are built to work in pulses. Researchers at the DIII-D National Fusion Facility in San Diego now report a way to push on that limit. They say they have measured current driven by radio waves called helicons, a method tried and dropped in the 1980s. The result has been published in Physical Review Letters. It is a small, careful lab measurement, not a power plant milestone.
A tokamak is a doughnut-shaped machine that uses magnets to hold plasma, a gas so hot its atoms break apart. Part of that magnetic cage comes from an electric current flowing inside the plasma itself. Keeping that current going is the job of “current drive.” DIII-D is a U.S. Department of Energy user facility, and General Atomics scientists Bob Pinsker, Jeff Lestz and Bart Van Compernolle led the work, according to ANS Nuclear Newswire.
Why Current Drive Decides How Long a Tokamak Runs
Many tokamaks start the plasma current with a pulse from a large electromagnet. That works, but only for as long as the magnet can keep changing its field. Pinsker told Gizmodo that pulse length is “intrinsically limited.” DIII-D pulses last about 10 seconds, Gizmodo reports. Larger machines can reach several hundred seconds or more.
A power plant would want to run far longer. So engineers look for other ways to drive current: beams of fast particles, microwaves, and now possibly helicons. The UK’s planned STEP prototype, which we covered in our look at the STEP fusion plant tender, will rely on other methods. ANS reports STEP will use nearly 200 gyrotrons, which are microwave sources.
What the DIII-D Team Measured
The lab’s own press release, issued ahead of a November 2025 American Physical Society plasma meeting, calls this the first successful detection of helicon current drive in a tokamak. It reports about 20 kilo-amperes (kA) of driven current near the hottest part of the plasma at its center, “consistent with theoretical modeling.” For scale, the release says tokamak plasma currents run to millions of amperes.
Interesting Engineering, summarizing the paper, adds details. It reports 370 kilowatts of radio power at 476 megahertz, sent through a 30-module antenna about 1.5 meters long. It gives the driven current as 20 kA, give or take 10 kA, against a simulated 21 kA. It also reports the plasma center heated by about 1 keV. I could not open the paper itself, so these figures come from that outlet, not from the journal page.
| Item | Reported value | Reported by |
|---|---|---|
| Driven current near the core | About 20 kA | DIII-D press release |
| Uncertainty and simulation | Plus or minus 10 kA; 21 kA simulated | Interesting Engineering |
| Radio power injected | 370 kW at 476 MHz | Interesting Engineering |
| DIII-D pulse length | About 10 seconds | Gizmodo, quoting Pinsker |
Source: the outlets named in the table, November 2025 to November 2026. Figures are as reported, not checked against the paper.
Why Old Experiments Failed and This One Might Not
The press release says early helicon tests “were not successful” because the plasmas did not resemble a power plant’s and the equipment was inadequate. The DIII-D team used a new comb-line antenna and preheated the plasma center to make it more reactor-like. Pinsker told the American Nuclear Society, “We really went to extremes to make sure that we couldn’t be accused of cherry-picking.”
What This Does Not Show
Start with scale. Twenty kA is a small slice of a current measured in millions of amperes. The test plasmas also had “far lower electron pressure than reactors are expected to have,” Interesting Engineering reports. ANS adds that the current was driven only at the core and only in low confinement modes. Lestz told Gizmodo that DIII-D cannot replicate reactor-like conditions.
This was a single experiment series on one machine. It did not measure fusion energy gain of any kind, scientific or engineering. It did not show long operation, and it says nothing about cost. Asked whether helicons could be the preferred method, Pinsker told ANS: “I would say it’s too early to tell that definitively.” He also said extending the work needs “to increase the power quite a bit.”
The next step is already under way. Interesting Engineering reports that the KSTAR tokamak in South Korea has installed a helicon antenna and begun commissioning.
Why It Matters to You
Fusion power is not on any grid today, and this result does not change that. If helicons scale up, they could give plant designers one more tool for steady operation. Steady operation is what turns a lab device into something that could supply power around the clock. Sources describe that as a path to be tested, not a date. For steady clean power you can use now, see our report on Fervo’s Utah geothermal plant.
Sources
- DIII-D National Fusion Facility, “DIII-D in the News” page and linked press release, “Scientists Pioneer New ‘Helicon’ Method of Driving Electrical Currents in Tokamak Plasmas for Fusion Energy Production,” embargoed to 17 November 2025.
- ANS Nuclear Newswire, “A new current drive option for tokamak-based power plants,” 30 October 2026.
- Gizmodo, “Physicists Gave Up on This Fusion Technique in the 1980s. Now It’s Working,” 9 November 2026.
- Interesting Engineering (via Yahoo News), “US nuclear fusion test drives tokamak current with a 1980’s radio-wave idea,” 13 November 2026.
What to watch next: whether KSTAR sees the same effect, and whether anyone repeats it at higher power. One machine’s result becomes a method only when others match it.
