PSA: The Steam Controller’s magnetic charger can be a fire hazard
- Thread starter JournalBot
- Start date
Well, as they say, "There's no Steam without fire."
Okay, that was awful, I admit. Sorry.
Okay, that was awful, I admit. Sorry.
Upvote
63
(63
/
0)
Why not a trickle voltage, and negotiation? We have microscopic chips at this point, why is this still a thing on such an expensive product? They exposed the pins, therefore one would have assumed that they weren't just pumping pure amperage the moment of contact. This feels particularly naive for such a product. Almost bizarre.
Edit: Thank you @coonwhiz
It appears they do. Another comment that the watch might have been charging at the same time breathes some idea of a ground fault or common fault into the idea. Or the resistance of the link band matched the resistance of whatever it negotiated. Either way the PSA applies that these systems are not foolproof.
Edit: Thank you @coonwhiz
It appears they do. Another comment that the watch might have been charging at the same time breathes some idea of a ground fault or common fault into the idea. Or the resistance of the link band matched the resistance of whatever it negotiated. Either way the PSA applies that these systems are not foolproof.
Last edited:
Upvote
75
(76
/
-1)
Wow, that's pretty bad! Especially so if the puck goes and "flips", attracted to, say, the metal case of the PC tower you left it on for convenience.
USB, modern high voltage versions at least, as well as POE devices, implement a safe low voltage "handshake" before delivering the desired higher end energy. I think perhaps the puck and controller could implement something similar. I'm thinking this would require a revision to the design, but maybe both have upgradeable firmware and it just MIGHT be possible to implement that sort of preliminary "handshake" through an update alone. It'd take everything lining up JUST right, but there's a CHANCE the early adopters aren't left holding a fire hazard.
Edit: It appears Zacharot managed to make the same point a minute ahead of me. Zacharoooooot!
USB, modern high voltage versions at least, as well as POE devices, implement a safe low voltage "handshake" before delivering the desired higher end energy. I think perhaps the puck and controller could implement something similar. I'm thinking this would require a revision to the design, but maybe both have upgradeable firmware and it just MIGHT be possible to implement that sort of preliminary "handshake" through an update alone. It'd take everything lining up JUST right, but there's a CHANCE the early adopters aren't left holding a fire hazard.
Edit: It appears Zacharot managed to make the same point a minute ahead of me. Zacharoooooot!
Last edited:
Upvote
13
(15
/
-2)
The metal pins in the MagSafe charger are recessed, so they present less of a short circuit hazard. I'd also speculate that the Apple device has short protection circuitry that the Steam device clearly lacks.
Upvote
27
(28
/
-1)
A user on reddit tried to replicate it with a variety of metal objects and was unsuccessful. Its possible that OP's scenario was a fluke where his pixel watch was charging and somehow shorted through the puck.
View: https://www.reddit.com/r/SteamController/comments/1tkelu6/psa_puck_is_not_an_immediate_fire_hazard_but/
View: https://www.reddit.com/r/SteamController/comments/1tkelu6/psa_puck_is_not_an_immediate_fire_hazard_but/
Upvote
81
(81
/
0)
I don't know if the device is any smarter, but design wise, the pins are indented slightly below the magnet. The Steam controller's raised pins seems like a bad idea
Upvote
16
(16
/
0)
Sorry valve but this seems like recall level material, exposed pins that can start a fire is no beuno no matter how many warnings you slap on it
Upvote
2
(16
/
-14)
For everyone else, let this be a reminder to keep your Steam Controller puck away from any metallic or magnetically sensitive objects.
I dont understand how this is the advice and not a mass recall and a redesign to avoid this. How many houses will have to burn down or people die before this is addressed?
I dont understand how this is the advice and not a mass recall and a redesign to avoid this. How many houses will have to burn down or people die before this is addressed?
Upvote
-1
(15
/
-16)
Fail.
EDIT: Please see @Quasarsaurus 's link, it seems it could be more complicated then just at a glance.
Last edited:
Upvote
12
(12
/
0)
Looking at the physical pins in the photo, it does look different in one key way - Apple's laptop magsafe keeps the pins recessed in a lowered area inside the magnet, which prevents easy contact with the pins by a stray watch band.
Upvote
30
(30
/
0)
Does Ars reach out to redditors to confirm their story? This all sounds plausible but it would be nice to read "we reached out to so and so..." rather than "they wrote on reddit...". It builds more confidence in the reporting.
Upvote
36
(38
/
-2)
According to a quick Google the issue with hoverboard batteries hit 50+ complaints to the CPSC before any noted recalls were made.
And $2mil in property damage.
Upvote
9
(10
/
-1)
Upvote
10
(10
/
0)
I always assumed these types of things had some sort of smart logic that only sent any measurable power through after confirming it was properly seated. What happens if I put a lightning cable in my mouth? Does it send 15w to me or confirm Im an iphone?
Upvote
6
(7
/
-1)
No idea how the MagSafe charger works, but I am an electrical engineer, and if I were designing a connector like this, I'd have it have some sort of digital "handshake" before energizing the contacts as well as current sensing to ensure that any time the current abruptly drops (like when disconnected) it has to repeat the handshake before re-energizing the pins. Probably $0.20 in parts, tops.
Bluntly, this is just bad engineering.
Upvote
18
(22
/
-4)
"Yes, we knew this was a problem. Yes, it happened in testing, or the engineers told us this could happen.
Why didn't we include a circuit to detect shorts on the pins? Well that would have cost another $1 per controller, and meant a production tooling change, which is expensive. "
/S
Is this something that every manufacturer who produces objects with exposed wiring pins deals with? Well, yes. Should they have included a protection circuit? Well, yes. Are they going to be sued by people doing this inadvertently, or advertantly? Oh, yes.
Why didn't we include a circuit to detect shorts on the pins? Well that would have cost another $1 per controller, and meant a production tooling change, which is expensive. "
/S
Is this something that every manufacturer who produces objects with exposed wiring pins deals with? Well, yes. Should they have included a protection circuit? Well, yes. Are they going to be sued by people doing this inadvertently, or advertantly? Oh, yes.
Last edited:
Upvote
-6
(3
/
-9)
Stop! Do not try this at home!
Do it in a public space where you can be recorded by others.
Upvote
36
(36
/
0)
I would like to know more about how it came close to starting a fire. I don’t know enough about the material that the watch band indented to know if this is an obvious problem or a misinterpretation of an admittedly disturbing event.
Upvote
0
(0
/
0)
Also an electrical engineer here. I don't think it's that simple if the link posted above by Quasarsaurus is also true.
Upvote
13
(13
/
0)
Scrolled back and read that. Agree that it looks like Valve did the right thing and something else is going on here.
Upvote
10
(10
/
0)
Agree. Which also extends to the physical design. It's pretty fundamental that the current carrying devices should not be the first thing to make contact.
For example, MagSafe recessed connectors as has already been mentioned, or (well designed) electrical plugs where the earth connection is the first connection before the live one.
Upvote
2
(3
/
-1)
jonbob_newcastle
Ars Scholae Palatinae
Lightning cables have a chip embedded in the connector that does authentication to prove it’s an officially approved cable. It seems a safe bet that there’s some current negotiation involved too. (It got cracked eventually.)
In days past I think they used to put a resistor across the USB data lines, the value of which told the charger how much current to supply. Some of the cheaper chargers and cables didn’t implement this correctly so Apple devices would either fall back to the standard 500mA or refuse to charge at all.
Upvote
2
(2
/
0)
I hope that's the case. If it's a matter of the "handshake" being negotiated by another device, I am curious just how that was achieved so as to engineer devices that may be next to other such devices so that they don't do this. I'd have expected the negotiation to have been done through the pins themselves rather than through NFC of some sort anyway.
Upvote
-6
(1
/
-7)
From the testing that sgasgy did, it has those circuits since he couldn't reproduce it. It is very likely it is an failure mode they were not aware of. It could be a software bug that left the charger in charging mode after the removing the steam controller. It could be a marginal reset circuit that is only problematic with a short cable. Or maybe a long cable causes issues. We are in the middle of part shortages(again) and I am sure Steam has to use contract manufactures. It is possible a key part was replaced with a counterfeit.
Upvote
0
(3
/
-3)
Well, not everyone. I go by the "extraordinary claims require extraordinary evidence" doctrine. If someone says they have a cat, I believe them wholeheartedly unless I'm given reason to doubt it. If someone says they have a goat, I might have some questions about how difficult that must be or if they live on farmland, but that's just curiosity. I don't really doubt that. If someone says they own a dinosaur, now I'm going to have to ask for some evidence. If they provide a picture of a crocodillian, I'll say "touche salesman" and move on. If they say they own a unicorn, not only do I doubt them, I simply leave the conversation.
Nothing about someone having a faulty electrical situation is unbelievable. While it MAY be a scam, it's likely enough that it warrants further investigation and a warning to others in advance that this may be a risk. So what if it hurts "potential sales"? I'm a lot less concerned with Valve's profits than I am with fire hazards.
Upvote
-1
(3
/
-4)
There's a reason why Apple's magsafe cables have recessed pogo pins plus a lot of smarts in the charger to avoid sending too much voltage or current unless they did a handshake with the notebook.
Upvote
6
(6
/
0)
jonbob_newcastle
Ars Scholae Palatinae
I must have seen a dozen YouTube videos of Americans heaping praise on the Type G plugs we have in the UK. They have at least five safety features, except one to stop you standing on them in bare feet, which hurts. Imagine standing on an errant Lego brick then multiply that by three.
- Earth pin is longer so is always the first connection.
- L and N pins are covered in plastic for most of their length so by the time they are connected there is no exposed metal.
- Plugs have a fuse rated for the appliance; there’s no need for a table lamp to draw more than 3A, so that’s what the fuse is.
- If you somehow manage to pull hard enough to rip the cord out of the plug the wires inside are cut to different lengths, such that the live wire is the first to be disconnected.
- The L and N connections in the socket are behind plastic doors that are opened by the earth pin, so it’s pretty much impossible for a child to stick anything into a socket that doesn’t belong there.
By the by, I’ve often wondered why American sockets aren’t installed upside down. Having the earth pin at the top would stop the rare instances of something conductive falling onto a half-inserted plug and shorting live and neutral.
Upvote
20
(20
/
0)
Oh, now I'm triggered. My house has upside-down outlets to indicate half the outlet is controlled by a wall switch. I freaking hate these outlets because I use flat-plug power strips (where the wire is supposed to go straight down flat against the wall), but on those outlets the wire goes straight up putting tons of strain on the plug.
Upvote
5
(5
/
0)
Because that failure mode basically never happens. If the pins are exposed enough for that to matter then the plug has basically fallen out of the outlet at that point. A loose outlet in practice exposes maybe a millimeter of conductor, and that's with a heavy cord and a very worn receptacle
And on a completely not serious reason, because then our outlets wouldn't look like surprised faces
Upvote
11
(11
/
0)
Another EE here, based on the reddit link above it looks like it might have been caused by a ground loop. The power itself maybe not have been directly generated by the controller puck or the pixel charger, but a circuit forming between the two. If that was the case, it's possible that the voltage and power going through the watch band was much higher than either charger could produce. Like others have mentioned US home power layout can have issues that places a voltage potential on what should be a neutral or grounded line. The cavate here is that good chargers should isolate everything on the low voltage side, but like most things that is a should and not a 100% do. Isolation of grounds is not always a given simply due to the nature of what a ground is expected to do.
Upvote
6
(6
/
0)
The watch was a Pixel Watch 3. These use induction chargers. Could the Pixel charger induce enough current in a metallic strap to cause damage if the watch was still charging when it contacted the pogo pins? I think that it's highly unlikely.
US plug designs also make it almost impossible to have partially "sleeved" conductors (where only the ends are conductive) or to have recessed plugs. Recessed plugs are used for some AV installations, but that makes them incompatible with many power cords that have the power leads coming out of the side because recessed outlets are not a standard in the US.
It's a very poorly designed plug from a safety perspective. The standard was locked in very early in history and it's almost impossible to update at this point without obsoleting enormous numbers of appliances.
That plug is one of the most over-engineered plugs in existence. The UK is unique in that it allowed "ring circuits" for AC power, when every other country used branch wiring, so the safety fusing had to be integrated into every plug as well as the fusebox, and this made the UK plugs so huge to accommodation room for a fuse.
That's common in commercial wiring now, to the extent that outlets sold/designed for commercial use have the markings "upside-down" relative to outlets sold for residential use.
US plug designs also make it almost impossible to have partially "sleeved" conductors (where only the ends are conductive) or to have recessed plugs. Recessed plugs are used for some AV installations, but that makes them incompatible with many power cords that have the power leads coming out of the side because recessed outlets are not a standard in the US.
It's a very poorly designed plug from a safety perspective. The standard was locked in very early in history and it's almost impossible to update at this point without obsoleting enormous numbers of appliances.
Upvote
9
(9
/
0)
What ground loop? The Pixel charger uses inductive charging.
EDIT - I stand corrected. Comments below show that the Pixel 3 charger uses actual pogo pins in their charger as well. That definitely opens up the possibility for a short circuit.
Last edited:
Upvote
-1
(1
/
-2)
I've looked through this thread several times and ctrl-F'd for coonwhiz, I'm not seeing this comment/link that is being referred to.
Edit: Is this referring to @Quasarsaurus comment? Was there a name change?
Upvote
6
(6
/
0)
Their story doesn't add up, we know the controller can communicate to the puck to turn on charging or not, therefore it wouldn't just send power by default.
Not saying that person is lying, but there's something missing in the equation here.
Not saying that person is lying, but there's something missing in the equation here.
Upvote
3
(5
/
-2)