Too Lit: The Exploding Fidget Spinner Teardown

Like so many things in the Mindtribe office, it all began with a Slack post;

Shortly followed by this Gizmodo article. Fidget spinners, like the hoverboards and Note 7s that preceded them, were exploding, their lithium polymer (LiPo) batteries causing fires in people’s living rooms.

Photo credit: WEYI via Gizmodo

At this point, you’re probably wondering what a battery is doing inside a fidget spinner. It turns out that some enterprising manufacturer decided that fidget spinners are good, but would be even better with a Bluetooth speaker and cool flashing LED lights. Many people agreed, and were probably quite happy with their purchase right up until it caught fire. At that point, the story got picked up by local news and then Gizmodo, the Verge, CNET, and a bunch of other outlets.

We soon finished our giggling and moved on with our lives. We were secure in the knowledge that as engineers and older millennials, we would never buy something as silly as a cheap fidget spinner with blinky LEDs and a Bluetooth speaker.

Until, of course, we did.

One of our engineers returned from a recent trip to China with not one, but three of the same exploding fidget spinners that we saw on Gizmodo. She found them in the Shenzhen electronics market (which is a fascinating story all on its own) and bought them for about $2 USD apiece. It was a perfect hilarious gift for an office full of engineers.

We tore off the packaging. We played with them. We affixed “CHARGING = FIRE!!” stickers to the USB ports so we wouldn’t burn the office down. But then we got curious. Now that we had our very own exploding fidget spinners, what makes them explode?

When we tore them open, we found a fascinating snapshot into the world of extreme cost reduction. Most of the teardowns we do on this blog are gadgets that cost several hundred to a few thousand dollars, and it was really interesting to look at something on the other end of the cost spectrum.

In fact, it’s pretty remarkable that these exist at all, for that price point. Buying any one of the major components off-the-shelf could easily cost you more than $2, but they somehow covered all of the components, plus a plastic housing, plus assembly and packaging for less than I spend on a bus fare. Sure, the speaker quality isn’t great, they feel cheap, and the rusty bearings don’t spin particularly well, but what did you expect, for that price?

That being said, there are some things that got left out, notably the charging protection circuit. Let’s take a closer look!

Outside of the enclosure, we can see the top side of the PCB. (Yes, they even managed a double-sided load PCB!) We can see LEDs sprinkled throughout, the speaker (yellow box), the microcontroller (red box), USB connector (green box), Bluetooth antenna (blue box), and the battery (purple box). The thing that jumps out is the little group of solder pads right above the battery, which are missing components. They even have callouts and outlines on the soldermask for R21, U4, U5, and C18. It turns out that unlike the Note 7, this probably isn’t an issue with the batteries themselves (as some have speculated). The batteries seem to be operating exactly as designed. It appears that the manufacturer simply decided not to include the charging and protection circuit. Whatever you plug into the USB gets routed almost directly into the battery.

Lithium Batteries are notoriously fussy about how they like to be charged. Normal cell voltage should never exceed 4.2V and the charge current should be the same as their rated capacity in milliamp hours. If you have a 100mAh battery, you should charge it at 100mA to a cut-off voltage of 4.2V. Most devices include a little chip or two to protect the cell from under-voltage, over-voltage, short circuit and high current. There seem to be spaces on the board for this (the empty soldier pads in the image above), but none of the actual chips are included. Instead, they use the very simplistic circuit below:

If we assume the LiPo cell starts off at 2.5V (normal fully discharged voltage) then it would start charging at 230mA! As the cell is probably only 10-20mAh, that is over 10x the charge current it should be using! What’s worse is the cell will try to charge up all the way to 4.67V! Way above the safe limit of 4.2V.

If it’s happening in your fidget spinner, this is very bad news. Charging the battery with too much current will cause it to overheat, boiling the flammable liquid electrolyte inside and eventually causing a fire. Similarly, charging the battery with voltages above the safe limit of 4.2V causes the graphite electrode to decompose into carbon dioxide gas, which inflates the battery and can make it explode. When you add both of those together, it’s no wonder that we see fidget spinner battery fires.

It turns out that charging protection circuits are important.

In low cost products like this, it is assumed that short cuts will be taken to reduce costs. Whether that is using domestic Chinese integrated circuits or excluding non-critical components, that usually has no huge effect of the final product. Omitting crucial things like battery protection however is very bad. Often, manufacturers produce lithium cells with protection circuits built in, but even that seems to have fallen to the relentless sword of cost reduction. If there had been a protection circuit either built into the battery or populated on the PCBA, this wouldn’t have happened. But since there’s no protection circuit anywhere, we’re seeing the inevitable result in smoke and flames.

For consumers, the biggest takeaway here is to look for a UL or CE mark on devices that you buy. UL in the US and CE in Europe are the governing bodies that certify electronics as safe to sell in their respective regions. Each product bearing their mark has undergone a battery of expensive and time-consuming safety tests and been approved as safe for the public. As responsible engineers, we spend a lot of time making sure the devices we design will pass various regulatory tests, but not everybody is so careful. The UL and CE marks appear below, and should appear on both the packaging and the device itself.

For engineers and designers, this is a good reminder that regulatory testing is important. As much as we may grumble at the thousands of dollars that testing costs and thousands of pages of regulations that we have to contend with, testing can save your product from becoming the next exploding fidget spinner.

It’s also a good parable about the danger of handing over your work to a CM without any oversight. People throughout the supply chain will sometimes look at your device and think they’ve found great new ways to optimize it. Indeed, coming up with helpful suggestions is a mark that you’ve found a good CM. However, you need an engineering voice who understands the product and can approve or disapprove those suggestions. If you don’t, bad things can start happening very quickly.

With our biggest question answered, we started poking through the rest of the device.

When we look at the bottom of the board, we find about what we’d expect. Some battery information, a few more chips, and more LEDs. Unfortunately, all of the chips seem to be generic Chinese manufactured, and a quick search didn’t reveal any information about them. If you have any more info, we’d love to hear it!

The mechanical parts are about what you’d expect, as well. Two injection molded plastic parts that hold the board in place with press-fit pins, and cheap bearings of highly variable quality. One of the spinners we got only rotates a few times before it grinds to a halt, while one spins smoothly the way you would expect.

Despite the low cost, the press fits are well tuned. The entire assembly, including the bearing and bearing caps, are held in place with press fits. They feel solid when assembled, but come apart with a little prying. Assembling the spinner is as simple as installing the board and pressing the whole assembly together. The parts aren’t cosmetically perfect – there are some pretty obvious knit lines on the bearing caps and flash throughout, but this is honestly better than I had expected for a $2 device.

It’s clear that like any hardware device, a lot of thought went into this fidget spinner. For all of its flaws, the fact that it exists at all, as a $2 device with Bluetooth, a battery, and a speaker, is a cost reduction triumph. However, you need to know when to stop cutting costs, before you push it so far that something fails. Testing is important, both for designers to know when they have gone too far and for consumers to check, for their own safety.