In 2013, on a visit to a Ugandan clinic, Manu Prakash noticed a small metal machine propping open a door. It was a centrifuge—a device that spins tubes of liquid at extremely high speeds, so that any particles within them sink to the bottom. This process is used to separate out everything from cells to viruses to DNA, which means that centrifuges underpin a lot of modern science. They’re as essential to hospitals and laboratories as saucepans are to kitchens.

But centrifuges run on electricity, which is why many clinics in the developing world can’t use them, even if they can afford them. That’s why the one that Prakash visited had repurposed their centrifuge as a doorstop. It’s also why the clinic couldn’t carry out a lot of simple diagnostic tests. “On a chart, they listed all the tests that they do, but they were only really doing two out of 10,” recalls Prakash. “The other eight required centrifugation. I’ve seen this play out over and over again. It was clear that we needed a centrifuge that could operate without electric power.”

Prakash has now created one. Modeled on an ancient children’s toy, and made with little more than paper, string, and tape, it can spin at speeds of up to 125,000 revolutions per minute (rpm). That’s more than enough to, say, separate cells or malaria parasites from blood samples. And it’s actually much faster than a lot of desktop centrifuges, even though Prakash’s device is entirely hand-powered, weighs less than 2 grams, and can be made for just 20 cents. He calls it the paperfuge.

“This type of innovation has the potential to support the creation of cost-appropriate, rapid, and robust diagnostics that can be administered by local health care workers in the lowest-resource communities in the world,” says Carol Dahl, Executive Director of the Lemelson Foundation.

Prakash, a biophysicist who grew up in India and now works at Stanford University, has a habit of creating devices like this. Last year, I wrote about his Foldscope—a $1 pocket microscope that can be folded from a sheet of paper. He has also developed a skin patch that detects parasitic worms in the manner of an ultrasound, but for less than $10. He has been working on tools for extracting malarial parasites from mosquito bites, and computers that run on water droplets. For these feats, he was awarded a MacArthur “genius grant” last year—a coveted $625,000 award given to exceptional and innovative people.

Having decided to make a cheap centrifuge, Saad Bhamla and other members of Prakash’s team started thinking about simple spinning devices. Yo-yos were an obvious pick, and one of the lab members was adept with them, having once been a circus performer. But the group soon learned that yo-yos can only achieve about 4,000 rpm. At such low speeds, it would take far too long to spin down a typical laboratory sample. The same problem has plagued other attempts to build hand-powered centrifuges out of egg-beaters or salad-spinners—they just don’t spin quickly enough.

“So, I asked the [team] to bring more toys into the lab,” Prakash says. And someone brought in a button spinner.

The simple toy, also known as a buzzer, is part of the whirligig family, and has been around for more than 5,000 years. They’re made by passing strings through the center of small discs—everything from buttons to bones and flattened musket balls. By twirling the ends of the strings, you spin the disc and wind the strings together. If you now pull the strings outwards, they unwind and rapidly spin the disc in the opposite direction.

By filming a basic whirligig, Prakash’s team were astonished to find that it could spin at 10,000 rpm. “We realized that this is a toy that no one had thought about,” he says. “The physics of how it works weren’t understood and its fundamental limits were completely unknown. So we spent six months thinking about the math, all with the goal of asking how fast it could really go.”

By playing around with the disc size, the position of the holes, the strings, and the speed at which the users pull them apart, the team managed to hit 125,000 rpm. That’s “the fastest rotational speed every recorded for a human-powered device,” they wrote. “We have submitted an application to Guinness World Records. But anyone can make a whirligig to our specifications and get those numbers. It’s just physics; it’s not special materials.”

In the actual paperfuges, the discs are just stiff, waterproof paper, like the stuff that bank notes are made from. The strings are just fishing lines, tied to two wooden handles for easy winding. Each paperfuge has two discs, which are stuck together by bits of Velcro and sandwich two sealed drinking straws. If you want to spin some liquid, you inject it into thin plastic tubes and load these into the straws. Stick the discs together, wind up the device—and off you go.

In less than 2 minutes of spinning, the paperfuge can separate the liquid part of blood (plasma) from the cells within it—an essential step in many diagnostic tests. In 15 minutes, it can isolate malaria parasites from a finger prick’s worth of blood.

And despite its simple nature, it has a number of seals, plugs, and other safety features, designed to stop users from spraying themselves with blood and other fluids. “We’ve built a thousand of these, taken a hundred into the field, and aggressively tested them,” says Prakash. “We’ve even tried throwing them into the street to see if any samples leaked out.”

“It’s so simple and yet so clever,” says Andres Martinez from California Polytechnic State University, who also works on cheap diagnostic tools. “Beyond the potential applications of the [paperfuge], I think it will inspire others to think about how other simple technologies could be applied toward improving life in developing areas of the world.”

“The importance of sample preparation is often overlooked and under-estimated. Yet it is the wheels of a car, without which a diagnostic test will not go far,” adds Helen Lee from the University of Cambridge. If the paperfuge can be integrated directly with tests that only require a few drops of blood, “it would be a great addition to the diagnostic tool box in resource-limited settings.”

The team is now testing other materials. By switching from paper to certain plastics, they could use 3-D printers to mass-manufacture millions of paperfuges; they could also print discs with more complicated liquid-holding channels, for carrying out more sophisticated chemical reactions.  Or, by using transparent plastic discs, they could create paperfuges that could double as microscope slides, and be examined immediately after being spun.

They have also taken their paperfuges to rural Madagascar to see if health workers can actually use them in the field. “The first time you explain it, people look at you speculatively,” he says. “But when you take it out and demonstrate it, it’s almost like a lightbulb goes on—ironic, since these are places that have no electricity.”