18

I'm currently building a PC peripheral that uses an AC to DC converter to power a set of DC fans. I'm not a trained EE or electrician, so I want to make sure I'm not putting myself a risk with this setup.

Essentially, I have the AC to DC converter pictured below with a screw terminal block on the DC end plugged into my breadboard power rails:

AC to DC converter
My precarious-looking DC screw terminal block

Assuming I'm using a standard 120V US outlet, am I putting myself at serious risk? I know it doesn't take much current to kill someone, so I get really concerned with this exposed wiring when I see this can output 3A at 12V.

DonutGaz
  • 285
  • 2
  • 6
  • 22
    12 V isn't anywhere near enough to kill you unless you stab yourself in the chest with the electrodes or something. And even then I'd be more worried about the stab wound than the voltage. Anything lower than 30 V AC/42 V DC is generally considered safe. – Hearth May 28 '22 at 23:49
  • Thank you! I feel much safer being able to ask these questions. – DonutGaz May 28 '22 at 23:49
  • 2
    I think stabbing yourself in the chest is more dangerous, you will feel a strong sting if you put on your tongue – Tony Stewart EE75 May 29 '22 at 02:52
  • @TonyStewartEE75 9 volts DC, passed through the heart, can be enough to kill you, if it induces fibrillation. (Over the outside of your skin, it's pretty safe.) – wizzwizz4 May 29 '22 at 17:47
  • 7
    @wizzwizz4 Don't stick them into your heart. Don't stick them near your heart. For crying out loud, don't stick anything near a heart that doesn't belong there unless you're a trained surgeon. – Mast May 29 '22 at 18:37
  • 1
    @wizzwizz4 what is the threshold for fibrillation on the heart? Besides that's irrelevant. – Tony Stewart EE75 May 29 '22 at 19:29
  • 1
    It wouldn't hurt to buy a better adaptor from some reputable source. The fear of being sued in the US helps to keep everyone honest when it comes to safety markings. But assuming the adaptor is designed and built properly, there is little or no danger to you. It might be a good idea to just verify that there is no continuity between the AC side and the DC side using a simple multimeter. – user57037 May 29 '22 at 20:40
  • 2
    @TonyStewartEE75 There's an urban legend that somebody (an EE student) managed to kill themselves with a 9V DC multimeter, attempting to "measure their internal resistance" by sticking a probe into each thumb, which passed a voltage across their heart. I don't know what the lower bound for "can it kill you" is, but I know that 12V DC is _above_ the "can it kill you" threshold if it gets inside your body. Do not stab yourself with live wires, no matter if it's a "safe" voltage or not. – wizzwizz4 May 29 '22 at 20:59
  • Ever stepped on one of these in the dark? – Jeroen3 May 30 '22 at 07:08
  • 1
    @wizzwizz4 brief googling tells me the resistance across the heart is about 50 ohms (odd coincidence) and less than 10 mA across the heart can kill, so yeah, that is another reason why stabbing yourself in the heart with probes at 12V would not be a great idea – llama May 30 '22 at 19:13
  • Lick it and find out – ScottishTapWater May 30 '22 at 21:44
  • 1
    Breadboard setups certainly do come with a higher risk of puncturing skin or at least making very positive skin contact (eg by touching a wire end that is unexpectedly sharp or hot), compared to just consumer use of such a wallwart. I would certainly avoid anything downright dodgy as a breadboarding supply. Better ones to look for: Anything that is legit graded medical or toy use. Secondhand original adapters that came with brand name equipment (eg original issue(!) phone chargers by Apple, Samsung, Motorola, or laptop supplies that came with a Dell, HP, Lenovo... laptop). – rackandboneman May 30 '22 at 23:09
  • 2
    Also: If you are serious about running experiments, there is a reason for the existence of laboratory power supplies (they are affordable these days). Also: Wallwarts with a two-prong mains plug have a tendency to unexpectedly lightly shock you due to a commonly used EMI reduction circuit.... – rackandboneman May 30 '22 at 23:15
  • @wizzwizz4 I heard a similar urban legend alooooong time ago, except it was with a (student?) doctor, a 9 volt battery and some saline soaked bandages on each wrist. But as pointed out, it's not the voltage, it's the current that will kill you, but dry skin is a very good insulator. – Peter M May 31 '22 at 14:59
  • It is the amps that kills you, ant three amps is more than enough to do it! In fact, it might actually cook you! Why are you using so much current for a little breadboard? However, if you eat only keto vegan cage-free carbon neutral non-GMO gluten-free organic food, you will not be electrocuted. An extra measure of safety, in case of ground fault, would be to power you device through a GFI outlet, just in case there is a ground fault and your bare feet are in a bucket of salt water. – richard1941 Jun 08 '22 at 22:06
  • The CCP is ultimately responsible for the safety of your power supply. Do you trust them? If not, then, with the power supply plugged in and operating, measure the voltage between either of the power leads and a water faucet. If it is enough to shock you, be careful. Another danger sign is continuity between your black wire and one of the power cord pins. And even if there is no ground fault, remember that baby puts out 3 Amps, way more that what is required to kill you permanently dead and unable to pay taxes. – richard1941 Jun 08 '22 at 22:21

3 Answers3

37

These can be good. They're usually good.

You don't need to guess about quality.

You only need to consider the source, and look at the markings to see if it has the mark of an NRTL (Nationally Recognized Testing Lab, a list curated by USA OSHA which many countries rely on). The most common NRTL marks you encounter in North America are

There's really no problem finding listed Wall Wart power supplies, as the things are absolute commodities made in billion quantity, and the vast majority of units are sold to consumer electronics manufacturers like Philips or Linksys who need listed supplies. Listed ones are so common I'm surprised to see one that is not listed. Maybe it's the multi-voltage switch that is the problem. Whatever, get a single-voltage unit.

There are many, many ways to shortcut the design so one of the DC output pins is energized at AC line voltage or AC line neutral (which is line voltage if certain malfunctions occur). However, UL won't allow that on a listed unit.

And since the vast majority of wall warts are approved, it's understandable to assume they all are.

Not this one, though.

All NRTL marks are conspicuous in their absence from this one. What you see is the marks that are universally faked in the North American market, because there are no consequences for doing so when your boots are on a faraway territory of a nation that does not cooperate with mark enforcement.

  • FCC. That is a self-certification that it complies with radio emissions rules, but self-certification is meaningless from overseas junk sellers, because the US agency FCC does not have the public funds to go on overseas adventures into uncooperative nations to defend their mark. *
  • CE. This is a European self-certification of compliance with EU safety rules. But the EU also won't spend public funds chasing miscreants outside the EU, so the mark has no force outside the EU proper.
  • RoHS. Ditto for EU electronic waste rules, e.g. use of lead-free solder.
  • CCC. China's competitor to CE, that China doesn't enforce on goods destined for export.

Reputable suppliers help.

It is rare to see falsification of NRTL marks in the consumer space, because the Federal law that enables NRTLs requires them to legally defend their mark aggressively. However, it is not impossible for someone to sneak something out before UL notices, and that is where it helps to use reputable suppliers who have a chain of custody. I.E. the item was shipped from Philips' warehouse to Home Depot or Mouser's warehouse. Reputable bricks-and-mortar stores are almost always reliable product, as they have good chain of custody and are heavily focused-on by consumer protection agencies.

Direct-mail sites such as eBay or Amazon generally involve 3rd party sellers, and there is no chain of custody whatsoever. I don't see much risk of forged NRTL marks in the consumer space, but if it happens anywhere, it'll happen there.

* If I were the FCC, I would go after wholesalers and drop-shippers of this garbage, but I am not the FCC.

Harper - Reinstate Monica
  • 21,639
  • 27
  • 68
  • 1
    Actually, this thing has two screws, so one can probably open it and take a look at the insulation and general quality. I think the failure modes that lead to mains voltage on the output side are mainly unintended contacts inside the device, whose possibility should be visible. – Peter - Reinstate Monica May 29 '22 at 14:11
  • 1
    I haven't looked in years, but IIRC a long time ago Radio Shack a similar multi-voltage wall wart. No reason I can think of why someone couldn't make an approved one today. – manassehkatz-Moving 2 Codidact May 29 '22 at 17:13
  • 1
    The logos mean nothing about quality and quality controls affect safety. You need more due dilgence than logo definitions and implications. – Tony Stewart EE75 May 29 '22 at 19:27
  • 2
    @TonyStewartEE75 I've edited to discuss that, but in my experience it's not a big problem in the consumer electronics space. UL has *robust* legal defense of their mark, and they're required to in order to be a NRTL. I've heard it's a bigger problem in the component space. – Harper - Reinstate Monica May 29 '22 at 19:46
  • 4
    @Peter-ReinstateMonica You can't practically look at the insulation inside the transformer, though, which can frequently be rather skimpy in the very cheapest ones. – Hearth May 29 '22 at 19:49
  • 1
    @Peter-ReinstateMonica I have to agree with Hearth here; there is no way to look at the insulation, unless you dismantle the transformer to see what kind of isolation it has (it seems to be a switch mode power supply), and even then you would have to be familiar with what is safe and what isn't, and you end up with dismantled power supply you can't use. – Justme May 29 '22 at 19:52
  • 2
    It is perhaps rare to see falsification of UL marks, but it's not *that* rare when it comes to wall warts in particular. [Example](https://www.ul.com/news/ul-warns-counterfeit-ul-mark-usb-power-adaptors-release-12pn-43). [Another example](https://www.ul.com/news/ul-warns-counterfeit-ul-marks-usb-chargers-release-20pn-27). [White paper by UL about counterfeit iPhone adapters](https://code-authorities.ul.com/wp-content/uploads/sites/40/2016/09/10314-CounterfeitiPhone-WP-HighRes_FINAL.pdf). *Especially* with online orders I would not be surprised at all to receive a counterfeit. – Chris May 29 '22 at 20:26
  • Yes @Peter, not least there are lots of "invisible" safety violations such as faulty components. Bet you don't see any UR Recognition. But how do you tell about plastic? Listed products require special expensive plastic that resists combustion and does not emit toxic gases that would incapacitate those trying to escape a fire. – Harper - Reinstate Monica May 29 '22 at 21:51
  • 3
    Is the second "CE" in your list of certifications supposed to be RoHS? – supercat May 30 '22 at 00:10
  • @Hearth Interesting, I didn't know that. (As an example, the video ttps://youtu.be/ehYYoU13i-I is one of many that show dangerous devices, including missing insulation inside transformers (the first one, fake astonishment or not, at 12.00). As somebody commented there, "Damn, this thing makes me feel much more safe using power supplies that I've made myself :D.") – Peter - Reinstate Monica May 30 '22 at 07:26
  • 2
    Note that letters C and E should be half-circles. These looks like half-ovals. It might not even be an European CE, it may be China Export mark (https://www.aydinlatma.org/en/how-to-distinguish-a-real-ce-mark-from-a-fake-ce-mark.html). CE on products from China is meaningless unless it is imported to European Union by a brand that has its reputation at stake, and big enough to make it worth prosecuting by regulators and media. – Mołot May 30 '22 at 12:02
  • 2
    LACK of certification marks is indeed a bad sign, but there is plenty of evidence of products bearing 'fake' markings, so it is pretty pointless to advise someone to rely on them. – MikeB May 30 '22 at 12:02
  • 1
    Indeed, this isn't even a European CE mark. In a European CE mark, the "C" and the "E" should be far enough apart that the "C" could make a complete circle without overlapping the "E". This one isn't even close. See [here](https://www.cemarkingassociation.co.uk/ce-marking-and-the-chinese-export-logo/). – Chris May 30 '22 at 19:15
  • How would you rate Reliability and Safety as separate issues on this product? – Tony Stewart EE75 Jun 04 '22 at 04:54
  • @supercat yes it is, good catch. – Harper - Reinstate Monica Jun 04 '22 at 05:46
  • @MikeBrockington that is where chain of custody is important. Home Depot certainly got the items straight from Eaton. Mouser certainly got it from Cree. With direct mail all bets are off, thanks to Amazon Commingling. – Harper - Reinstate Monica Jun 04 '22 at 05:48
10

No. Those wall warts are quite safe. The high voltage input is isolated form the low voltage side by a transformer that steps down the voltage right off the bat to a safe level, and pretty much all the ways that a transformer can fail won't ever allow the the high voltage to appear on the output.

That's why so many inexpensive consumer electronics use them rather than directly accepting the mains voltage and regulating it down inside the product.

DKNguyen
  • 54,733
  • 4
  • 67
  • 153
  • 13
    I've seen cheap Chinese ones that could kill you with fake certs. – Voltage Spike May 29 '22 at 00:10
  • 2
    @VoltageSpike Ehhh, yeah don't buy shit ones off alibabi or wherever. – DKNguyen May 29 '22 at 01:14
  • 31
    It’s pretty suspicious that the brand named of the adapter is one letter off a major brand and uses the same font. Never heard of “belker” myself. – Bryan May 29 '22 at 01:28
  • 2
    "Belker" huh? Close enough phonetically to "Belken". Seems it could be generic internally but some corners could be cut. I would not rely on it to last long above 240V. A rattle when you shake them is also a worry. – mckenzm May 30 '22 at 04:00
  • I've seen one with switching power regulation and no transformer at all, just some inductors. I'm not sure how you can tell the difference from the photo, but if you can, please share. If you can't, this answer is based on assumption that may be fake. – Mołot May 30 '22 at 12:05
  • @Mołot Those are the newer switchers and yes they do not use a transformer. You can tell by them being a lot lighter. I believe they use capacitive isolation. Not quite as fool proof but safe enough if properly designed. Definitely don't cheap out on those. – DKNguyen May 30 '22 at 15:49
  • How do you rate Reliability and Safety as separate issues and specifically on this example product? (anyone?) – Tony Stewart EE75 Jun 04 '22 at 04:53
6

Yes, definitely.

When properly designed, wall warts are quite safe. When improperly designed, there are realistic failure modes that put mains voltage on the output.

There are many bad wall warts on the market. Even a UL certification stamp is insufficient, since there are wall warts sold online (and even in stores) that put counterfeit stamps on while flagrantly ignoring the actual safety standards. Example. Another example.

UL released a white paper about counterfeit iPhone adapters. A very relevant quote:

In total, we tested 400 adapters and the results were literally shocking. The overall failure rate exceeded 99 percent. All but three failed our basic safety tests and were fire and shock hazards. Twelve were so poorly designed and constructed that they posed a risk of lethal electrocution to the user.

So the question really becomes, do you trust your life with this wall wart? If it were my life, I'd shell out a few more dollars for a more reliable one.

Chris
  • 442
  • 3
  • 9