4 ms·

600W from a tiny connector is questionable but the issue isn't caused by the 12VHPWR connector. Fault happens on the part that converts 4x older PCIe into the n

by tuvan 4y ago

600W from a tiny connector is questionable but the issue isn't caused by the 12VHPWR connector. Fault happens on the part that converts 4x older PCIe into the new headers. So this is an adapter issue and not something related to the PCI SIG design. According to Igor's labs investigation what happens is one or two of the welds in the adapter crack and adapter turns into 2x PCIe to 12VHPRW which is not what it's specced out for.

trashtester 4y ago
I think the 50A is a bigger issue than the 600W. Consider that the latest spec for usb-c allows up to 240W in a much smaller cable, but does so at 48V, meaning that the current is only 5A.
- willis936 4y ago
  Note that it's in a much smaller cable. Galax tested a 12VHPWR connector at 1530W and the outer jacket stayed under 70C. https://twitter.com/hms1193/status/1585257428291325958 https://twitter.com/hms1193/status/1585257428291325958
  - lazide 4y ago
    To be clear - the typical temperature rating you’ll find on mains power cabling in a house is 60c. The highest you’ll typically be able to find is 75c. 70 Celsius is NUTS for that usage, not ‘oh, that’s ok then’
    - wongarsu 4y ago
      It's easier to spec out heat resistive materials for a 5cm cable than for the entirety of a house's mains power cabling. Sure, 70°C is a lot, but I think the point is that it would work if it had to.
      - lazide 4y ago
        It’s a sign of a lot of resistive heat losses and requires specialized insulation and connectors at that level (same for mains). Typically insulation is starting to weaken or even melt at that point. It’s well outside normal expected operating temperatures for wire.
        willis936 4y ago
        Crazy high resistive heating when driven at multiple factors above the rated power isn't cause for alarm though. If you put 50A through your wall outlet you wouldn't be surprised when it got hot. The only application where pulling 1500W through a 12VHPWR connector is even close to possible is when using cryogenic cooling (when the conductive cooling will keep the power cable below ambient despite how many amps are crammed through them). This data point isn't exhaustive, but it does indicate that the actual safety margin is not as tight as is assumed from this news cycle.
        lazide 4y ago
        You and I have a very different concept of safe? I don’t think it’s saying what you think it’s saying. That brand new connectors and wire that are custom specced wouldn’t catch on fire (but come close!) at only 2x the power draw means it’s very likely any damage to a connector, wire, etc. especially with normal wire and connectors likely would cause damage and a fire, at much less than that power draw. Which is what we’ve been getting reports on. And that is just from minor physical damage it seems, ignoring corrosion or fraying wires over time which is usually the bigger problem.
        willis936 4y ago
        That is not an issue with the connector though. That is the adapter. That is my entire point. The connector is safe. Shoddy adapters are the problem. Putting multiple supplies in parallel is always very shady, doubly when it's high current through a small piece of metal. That's what's dangerous, not the 12VHPWR connector.
incrudible 4y ago
The issue is the 12VHPWR connector, not just the adapter. http://jongerow.com/12VHPWR/ http://jongerow.com/12VHPWR/ Yes, if "the right materials" are used and the cables are treated "properly" (not bent out of spec, not too many connection cycles) then it's "OK", but those shouldn't be requirements designed into something that cost-conscious consumers are going to be assembling.

by tuvan 4y ago

trashtester 4y ago

I think the 50A is a bigger issue than the 600W. Consider that the latest spec for usb-c allows up to 240W in a much smaller cable, but does so at 48V, meaning that the current is only 5A.

willis936 4y ago
Note that it's in a much smaller cable. Galax tested a 12VHPWR connector at 1530W and the outer jacket stayed under 70C. https://twitter.com/hms1193/status/1585257428291325958 https://twitter.com/hms1193/status/1585257428291325958
- lazide 4y ago
  To be clear - the typical temperature rating you’ll find on mains power cabling in a house is 60c. The highest you’ll typically be able to find is 75c. 70 Celsius is NUTS for that usage, not ‘oh, that’s ok then’
  - wongarsu 4y ago
    It's easier to spec out heat resistive materials for a 5cm cable than for the entirety of a house's mains power cabling. Sure, 70°C is a lot, but I think the point is that it would work if it had to.
    - lazide 4y ago
      It’s a sign of a lot of resistive heat losses and requires specialized insulation and connectors at that level (same for mains). Typically insulation is starting to weaken or even melt at that point. It’s well outside normal expected operating temperatures for wire.
      - willis936 4y ago
        Crazy high resistive heating when driven at multiple factors above the rated power isn't cause for alarm though. If you put 50A through your wall outlet you wouldn't be surprised when it got hot. The only application where pulling 1500W through a 12VHPWR connector is even close to possible is when using cryogenic cooling (when the conductive cooling will keep the power cable below ambient despite how many amps are crammed through them). This data point isn't exhaustive, but it does indicate that the actual safety margin is not as tight as is assumed from this news cycle.
        lazide 4y ago
        You and I have a very different concept of safe? I don’t think it’s saying what you think it’s saying. That brand new connectors and wire that are custom specced wouldn’t catch on fire (but come close!) at only 2x the power draw means it’s very likely any damage to a connector, wire, etc. especially with normal wire and connectors likely would cause damage and a fire, at much less than that power draw. Which is what we’ve been getting reports on. And that is just from minor physical damage it seems, ignoring corrosion or fraying wires over time which is usually the bigger problem.
        willis936 4y ago
        That is not an issue with the connector though. That is the adapter. That is my entire point. The connector is safe. Shoddy adapters are the problem. Putting multiple supplies in parallel is always very shady, doubly when it's high current through a small piece of metal. That's what's dangerous, not the 12VHPWR connector.

incrudible 4y ago

The issue is the 12VHPWR connector, not just the adapter. http://jongerow.com/12VHPWR/ http://jongerow.com/12VHPWR/ Yes, if "the right materials" are used and the cables are treated "properly" (not bent out of spec, not too many connection cycles) then it's "OK", but those shouldn't be requirements designed into something that cost-conscious consumers are going to be assembling.