An electrical experiment for Easter

Chris Pearson said:

A USB charger (Apple Model A1696) was plugged into the adjacent socket-outlet. A USB-C to USB-C cable was plugged into the charger and the other end was immersed in the sodium chloride solution. The charger was energized.

Correct function of the charger was confirmed after the experiment. Confirmation that the earth was effective was obtained by measuring the earth fault loop impedance at the socket outlet.

Results

No current flow was detected. The EFLI was 0.80 ohms.

I would question whether your experimental technique is a valid representation of the real-world risk of a USB device in water, because the more recent versions of the USB standards, including USB-C are not a simple, always-on, DC power supply but rather they have a set of communications protocols and hand-shaking between the host and the connected device, whereby they communicate details of the respective devices, power capacity (volts/amps) and (from a quick skim read of Wiki) it looks like the host (i.e power supply) might be checking for certain conditions at the device end of the cable before powering up the supply. 

So it's possible that with a compliant USB-C power supply and USB-C cable, the plug you immersed into the solution may not have been powered up at 5V at all.

There's also the question of 'Power Delivery' functions which allow the USB power supply to raise the supply voltage above 5V depending on the connected device, in a real-world situation a lot of newer USB-C power supplies have PD functionality, so connected to a phone could be providing upto 20V if turbo-charging the phone, but to test that would require a suitable power supply and device on the end of the cable, or some other way of ensuring the power supply was outputting the higher voltage onto the USB-C plug when immersing it into water.

A (standards conforming) USB port will always (assuming it hasn't been turned off in some way) have 5V on it and be capable of delivering 100mA, before anything is plugged in.  There are various ways in which the port and the connected device can then negotiate different voltages and currents, up to 5A at 48V (notably the USB Battery Charging and Power Delivery Specifications), assuming that those voltages and currents are supported at both ends (and in the higher power cases (more than 20V or more than 3A), by the cable itself which also has to contain an electronic ID which is validated before allowing the higher powers).

High end phones now charge at 45W or more with a USB charger capable of delivering at least 15V, after negotiation.  I suspect that in many cases those phones are only charged at ~2A @ 5V most of the time as that is what most non-OEM chargers will deliver.

A (standards conforming) USB port will always (assuming it hasn't been turned off in some way) have 5V on it and be capable of delivering 100mA, before anything is plugged in.  There are various ways in which the port and the connected device can then negotiate different voltages and currents, up to 5A at 48V (notably the USB Battery Charging and Power Delivery Specifications), assuming that those voltages and currents are supported at both ends (and in the higher power cases (more than 20V or more than 3A), by the cable itself which also has to contain an electronic ID which is validated before allowing the higher powers).

High end phones now charge at 45W or more with a USB charger capable of delivering at least 15V, after negotiation.  I suspect that in many cases those phones are only charged at ~2A @ 5V most of the time as that is what most non-OEM chargers will deliver.

Are you sure that is correct for USB 2.0 on USB Type C ? I wouldn't disagree for the earlier USB versions, but have 2.0 / Type C changed this?

I don't work on USB standards, but I've seen some sources stating that USB-C requires a specific pull-down resistance be applied to one of the pins, CC, to trigger the host device to power up the V-Bus. Suggests that without the required pull down resistor, V-Bus is switched off.

DShutt said:

A (standards conforming) USB port

It's not the USB end that is the [source of] danger, but the power plug, it's adapter, and the wider EMC and product design standards (real or imagined).

Bad component select, lack of galvanic isolation, addition of EMC filters, assumptions about 2 vs 3 pin power all play a part in the safety (or lack of) for the usage of the charger in scenarios like these. 

It (understanding) is made worse by the rareness of the events and consequent 'poor' failure analysis after the event.

I greatly doubt that the charger was operating 'normally' and yet creating a fatality. 

We may be doing that 'searching for the car keys where the light is better' rather than the darker truths...

Hi,

Sorry, you are absolutely correct for type USB type C connections.  Given that type C leads are identical both ends, there has to be a means of preventing someone connecting two power sources together and the CC pin(s) are used to ensure that power is not delivered unless a sink is attached (amongst other things).

Having said that, I wouldn't be relying on that control for safety purposes although it might affect the results of the experiment.  It might be especially relevant if the absence of the pull down on actually turns off the charger as an energy saving feature.

My comment (which is wrong for the USB C case by the way as noted in my last post) was more about presence of 5V on the lead rather than the safety, or otherwise of delivering that 5V.

Unfortunately, as noted by others, it may be quite possible for the charger to be operating as designed and still pass enough current to cause harmful effects to someone in a bath.  I would call such a design defective but others might call it "value engineered"; regulations / standards will have a third view.