Every so often we see on line storeys of fake chargers burning down houses, and electric shocks. Normally these chargers are refereed to as ‘Apple’ or ‘iPhone’ chargers, however that is just because the small triangle chargers (in a variety of colours) are so prevalent. An official Apple 5W USB Power Adapter (link).
A while ago (pre Britex referendum) I bought a 89p phone charger (with free shipping) from Ali express (link to a 96p one), just to see how ‘good’ it is, lets just say that you wont be surprised by any of this:
The Phone charger is a BS1363 ish plug, so testing it is easer than testing either the HLK-PM01 or the cheap Chinese crap one (And yes I know I promised to revisit the HLK-PM01 to try and get a fix).
Having a quick go at checking to see how well it conforms to BS EN 55032:2015 – “Electromagnetic compatibility of multimedia equipment. Emission Requirements” (until the beginning of this month I would have considered BS EN 55022 [pdf]), It will have to comply to the Class B (Residential / Domestic) limit lines for conducted and radiated emissions
Conducted Emissions testing give us these results:
Radiated Emissions testing give us these results:
Looking at these two graphs, you can see that while the radiated emissions are fine, it fails to meet the requirements of the conducted emissions (Blue squares below the red line, green circles below the blue line). now we know its not permissible to place the charger on the market in the EU (including the UK) from an emissions point of view. We can now assess its electrical safety.
So lets look at what happens when we perform a quick LVD screen test
The results for the LVD screen test:
- Offload DC out 5.3 V
- DC maximum output before supply folded ~1 A
- Power used – 3 W off load, 8 W @ 1 A
- Earth Pin is ok, however spacing issues with Live and Neutral
- Fails a breakdown test 600 Vac @ 1 mA (does not recover).
- No secondary securing of internal primary wire to pins, or PCB.
- Single insulation mains in contact with LV / Secondary side.
- No physical barrier between Pins and PCB
- Capacitor between Primary and Secondary circuits is not Y rated.
- Separation between Primary and Secondary does not meet requirements.
- No mains fuse fitted
- Solder on rear of pins in physical contact with capacitor on secondary side.
- No Class II symbol on regulatory markings
I am sure it is safe to say that I was not surprised by the lack of conformity of the charger, I was however surprised at just how terrible it was!
I have had a number of iPhones over the years, and as such I have two real iPhone Chargers laying around, have a look at the markings on these two:
13 thoughts on “What does 89p get you? Autopsy of a Chinese phone charger”
The problem is that 99% of these uber cheap USB chargers coming from China are RF spewing deathtraps like this. This one has a bonus of being a fake Apple charger but you will find the same circuitry with minor variations and more or less egregious safety issues in most of these junkers. Even the ones sold at big name brick & mortar stores, unfortunately!
Saving money on mains power supplies is pretty foolish – that doesn’t mean one has to buy an overpriced Apple one but there are plenty of decent ones that are UL or TUV certified, for example (i.e. someone has actually inspected & tested them, CE mark *doesn’t require any testing whatsoever*!)
No to mention the “China export” mark
https://en.wikipedia.org/wiki/CE_marking#China_Export
Though how true this is, is debatable
So what happens if one of these guys burns your house down? I’ve wondered about this. You can buy power outlets from Amazon that claim they are UL listed but there’s no actual listing number to be found. Does Amazon push off the blame to the seller if your house burns down? Are consumers truly responsible? Most of my friends are college educated, some are engineers and such, don’t even know what a UL listing is and I’m for sure my mother doesn’t know either. Is there anything being done about this or is it really not a problem? (I’m not expecting you [Will] to know, just asking anyone that zips by)
Safety, as a physics and chemistry-based engineering discipline, has long been ignored and shunned by the ‘hacker’ community. Understand that it is not cool, not to mention frustrating, to hear people say that your wondrous creation may not be safe. And members of IEEE Product Safety Engineering and EMC societies talk about this every time more than two compliance engineers gather in the same spot.
But we are all grown up and haired over – so let us not ignore accredited safety and EMC engineers when they offer a suggestion or two for your magnificent design. And yes, I have submitted articles to HAD, and I do understand why they were ignored. It is time for the HAD writers themselves to continue, at frequent and regular intervals, with the work started by Anool Mahidharia and others.
I suspect, they’ll pass the responsibility to the distributor or manufacturer. If they know the issue, they’ll fall back on plausible deniability. Most resellers don’t look past it’s cheap and it sells for a profit. I’m also not aware of, here in the States, anybody that has taken upon themselves or a family member to sue for the experience of fatal or near fatal shocks or fires in spite of how law-suit happy Americans are.
Here’s a couple of links to another site that has also looked at these things in the States:
https://www.tomshardware.com/picturestory/803-apple-5w-adapter-knock-offs-colorful-a1265-tear-down.html
https://www.tomshardware.com/picturestory/821-a1265-usb-charger-tear-down.html
There is a site that tests batteries, chargers and powerbanks ( https://lygte-info.dk/ ) and most of those ultracheap chargers are deathtraps. HV/LV insulation is almost always below safe specs, sometimes insulation is prone to melting, plastic is usually not flame-safe … And filtering is non-exsistant while noise is measured in hundreds of milivolts. You can buy original for 30 bucks, good quality 3rd party for 15 bucks or complete crap for 2-5 bucks.
The BOM for a safe charger is only around $2. It may take a bit of searching to find a genuine Samsung or LG charger for under $5, but they are out there.
For a while, Poundland in the UK did £1 5W USB chargers that were supposedly OK, though they were rather bigger than the Apple ones. Trying to cram everything into such a small form factor whilst cutting costs to the minimum seems to spell sure doom.
I think Clive knows a song about this, don’t you?
https://www.youtube.com/watch?v=ioAq7PI1Uwg#action=share
I believe the worst thing about these pieces of Chinese junk is that the general public knows nothing of the dangers and the regulators seem oblivious to the dangers of the ones that do make it to the mainstream markets. Doesn’t matter if it’s a petrol/gas station, the local supermarket or convenience store… It’s cheap, so it sells, In the States, these things can pop up with counterfeit markings, and I have no doubt that happens in other markets as well.
“applying 600 VAC common mode to its pins and checking for leakage current through the device”
What is the specific test method being cited and used? Touch current testing is typically done at 110% of rated mains voltage. If you are referring to di-electric withstand, then the test level is typically 3kV, and uncontrolled leakage current or breakdown of galvanic isolation determines pass/fail.
references
1. IEC/EN/ANSI 61010-1, annex A,
2. EN/ANSI 60950-1, section 2.5, annex D
3. EN/ANSI 60065, annex D
3. IEC/ANSI/EN 62368-1, clauses 2 and 5
4. IEC 60990
Note: standards based on IEC609950-x and IEC60065 should be considered deprecated.
Ok, re-read the Skippy article – some stuff was lost in translation to British to American. Perhaps it was intended to say that that, during the di-electric withstand test (‘hi-pot’), that the unit failed at 600V. But 1mA would seem to be an unusually low limit setting for this test.
For those interested, for hi-pot, the test voltage is linearly increased (montonic) to the test level (3kV) and then held at the test level for a certain period. during the voltage ramp or the test dwell periods, there can be no uncontrolled flow of current and insulators cannot arc or otherwise fail. Depending on the equipment and the scoped standard(s), I have typically set the current limit for 125% of the calculated leakage at the peak test voltage. For stand-alone transformers, I have typically set the current limit to 1mA or less. But for power supplies, limits are seldom less than 5mA.
Here is a video of a different EUT undergoing hi-pot test: https://skippy.org.uk/5v-acdc-converter-switch-power-supply-module-3w-700ma-industrial-voltage-regulators/
I have a few different things on my site that I have subjected to EMC and LVD assessment over the last few years,