900W+ PSUs on 120V with the C14 inlet — whose fault is it?

Hardware Insights specializes in power supplies. In this part of the forum, you will find detailed information on every aspect of power supplies, including how to repair them and how to judge if a power supply is a good quality unit or not.

Who's to blame for leaving the monster-PSU-owning North Americans to check their power cords?

The UL (and CSA), for seeing fit to uprate the C14 inlet from 10A to 15A
The safety regulators there as a whole, for not mandating cables sized to carry the full connector rating without overheating
No votes
The monster-PSU manufacturers, for not putting a big fat warning label near the inlet
No votes
The monster-PSU manufacturers, for proceeding to take UL/CSA words and use the C14 inlet above 10A
No votes
None of them — I still think the responsibility lies on the user
No votes
Those who left us with the two different voltage ranges to begin with
No votes
The market, for demanding such huge PSUs in the first place :D
No votes
Total votes : 2

900W+ PSUs on 120V with the C14 inlet — whose fault is it?

Postby LongRunner » November 24th, 2014, 11:29 pm

So as we know, you can't just use any old power cord for running a 1375W (or anything greater than around 750W output, to be realistic) PSU on the lower mains voltage range — at least, not if you don't want the cord to melt down sooner or later (or at the very least, age prematurely). (For the convenience of those outside the States, 18AWG is nominally 0.823mm² — which is marginally larger than the smallest metric size allowed for such cords, 0.75mm². In practical terms, that probably makes about a half-amp of difference to the maximum continuous current.)

But whose fault is it that the monster-PSU users have to check what size cord they use, anyway? It bothers me when the authorities leave responsibility on the public that they could easily take into their own hands if they cared (I did make the OCP-less multi-way adapters a contender for the most dangerous electrical item available, after all). Frankly, it seems that the North American electrical "safety" authorities have lost the plot — as anyone who compares their plugs to the other types (except for the even stupider Japanese type which resembles a NEMA 1-15P with a separate earth-lug-on-a-wire) will notice:

Comparison of US, Australian, and UK plugs.jpg
From left to right: US NEMA 5-15P, an older (pre-2005) AS/NZS 3112 10A plug with uninsulated pins, a current AS/NZS plug, and a BS 1363 plug — with polarity labelled. All four cords have normal C13 sockets at the other end. And yes, I made the photo (the US cord, a Japanese cord, and a whole bunch of British and Schuko/French cords having made their way into my collection).
Comparison of US, Australian, and UK plugs.jpg (98.44 KiB) Viewed 20641 times

Sorry, but you can't place bare pins 0.15″ (3.8mm) from the edge of the plug body, with non-recessed outlets, and not expect shocked users aplenty. :dodgy: (The last time I checked, they hadn't improved matters. Oh, and the US cord is an 18AWG unit; the older Australian model shown is 1.0mm² and the other two are 0.75mm².) Even before the requirement for pin insulation, the Australian plug pins were at least a reasonable distance from the edge (and 10A extension-cord sockets here got safety shrouds some years before the pins were changed).

Anyway: The IEC themselves rate the C13/C14, C15/C16, C15A/C16A, and C17/C18 (all variants on the same basic connector; C13/C14 being the base model, C15/C16 able to tolerate up to 120°C, C15A/C16A withstanding up to 155°C, and C17/C18 being a normal-temperature Class II version) for 10A, and the C19/C20, C21/C22, and C23/C24 (which again are Class I up to 70°C, Class I up to 155°C, and Class II up to 70°C respectively) for 16A; the UL/CSA ratings are 15A and 20A respectively. But why? The connectors are already small enough for their current ratings.

It also happens that most smaller 2-core extension cords in North America have 16AWG (nominally 1.31mm²) conductors, which are rated by the authorities there for 13A continuous (which looks accurate enough). This is still a bit undersized for the connector rating of 15A, and in the past, they also had 18AWG extension cords — make of that what you will.

At least those on 230V can use almost any PSU with any (legitimate) cord without the risk of hot wiring — but did you see the fuse rating on the British plug in my photo? At 5A, it will place the same limit on power draw as would be imposed by an 18AWG cord on 120V — and it's definitely the original fuse (the plug body bears the same 5A rating, and it's one of a batch of identical cords that I got — as a result of someone who was importing equipment and had no use for them). Well, you could increase it to 7A (maybe even to 10A) without danger, and while I can't deny that the Brits led the way in safe plugs and outlets, it seems that their regulators are missing a sanity check — remember that cord I mentioned with a 13A fuse and a C5 socket? For cords fixed to appliances, matching the fuse to the appliance's current draw is sensible enough, but for detachable cords, the fuse should be chosen to match the rating of the connector, with the appliance protected by its own fuse (which may be accessible or concealed as desired). Also out of place, for that matter, is the advice given on the tag (outside of the shot) — which not only describes plug replacement when it would be easier to just swap the cord (IEC 60320 connectors and all), but also states that:

I-Sheng (on the tag) wrote:With alternative plugs on (sic) approved 5 amp fuse must be fitted in the plug or adaptor or in the main fuse box.

But the other plug types (with the exception of some BS 546 models) don't even have fuses. :s Anyway, surely it would be more appropriate to simply provide the fuse rating (preferably more than half the connector rating) and a quick note that the user should swap the cord to operate the appliance/device/whatever elsewhere?

It may be inconvenient for the PSU makers to (hypothetically — I don't know if it has been done in a production model) provide a version with the C20 inlet (for 120V at full power) and another with the C14 inlet (for 230V, and that will refuse to run or at least restrict the available power on 120V), but better that than fire.

Anyway, at least those in the rest of Europe and Oz can freely use massive PC PSUs — provided of course that counterfeit cords are avoided like the plague, and that the user is happy with running a rig capable of doubling as a room heater :D. But who do you think is to blame for creating the dangerous situation in the first place? That's what the poll is about.
Information is far more fragile than the HDDs it's stored on.

Smart people don't buy "smart" devices without very carefully weighing up the risks and benefits beforehand.

My PC: Core i3 4130 on GA-H87M-D3H with GT640 OC 2GiB and 2 * 8GiB Kingston HyperX 1600MHz, Kingston SA400S37120G, WD3003FZEX-00Z4SA0 and HDS721010CLA630, Pioneer BDR-209DBKS and Optiarc AD-7200S, Seasonic G-360, Chenbro PC31031, Linux Mint Cinnamon 20.2 (with Windows 7 still accessible if needed).
Posts: 904
Joined: May 17th, 2013, 5:48 pm
Location: Albany, Western Australia


Who is online

Users browsing this forum: No registered users and 5 guests