News 12VHPWR adapters sporting heatsinks and thermal pads show how problematic the connector is

[Admin]

Extra cooling components on 16-pin adapters demonstrate that the standard is fragile compared to 8-pin and 6-pin connectors.

12VHPWR adapters sporting heatsinks and thermal pads show how problematic the connector is : Read more

 

[lmcnabney]

If the power is balanced 'perfectly' between the six +12V wires it means that 8.3 amps are running through each of those 16ga copper wires. There is a hard limit of 10 amps for 16ga for electrical standards in general. Using 8 pin PCIE cabling keeps the current down to 5.5 amps. The 16 pin connector was almost doomed to have a higher electrical failure rate. Stripping out overcurrent controls have only hastened it.

 

[artk2219]

If the power is balanced 'perfectly' between the six +12V wires it means that 8.3 amps are running through each of those 16ga copper wires. There is a hard limit of 10 amps for 16ga for electrical standards in general. Using 8 pin PCIE cabling keeps the current down to 5.5 amps. The 16 pin connector was almost doomed to have a higher electrical failure rate. Stripping out overcurrent controls have only hastened it.

Only a 20% margin of error with no current sensing or load balancing. You're right, this standard was destined to cause failures.

 

[Ogotai]

oh look, another band aid for a connector that needs to be scrapped

 

[TJ Hooker]

There is a hard limit of 10 amps for 16ga for electrical standards in general.

No there isn't, where did you see that? E.g. with 90 C rated insulation the NEC rates 16 AWG copper for 18 amps.

https://www.productinfo.schneider-e.../ConductorAmpacityBasedOnThe2017Nati-F20AB318

 

[qxp]

There is a discussion on what users want in a new connector here:

Proposed new Power Connector

With all the attention on melting graphics card connectors, this might be an opportunity to switch manufacturers to a connector that we like. Here is my proposal - please comment whether you like the idea or not, any changes and if we get enough people behind it we'll try to make them listen...

www.techpowerup.com

Short summary: 48V, one wire for positive, one for ground, two sensing wires to tell whether power wires are overloading (so you don't have to reset fuses) and two data wires to talk to PSU. 48V could also be used to provide motherboard with power ready for USB-C connectors.

 

[RodroX]

Good thing!

More items, no one asked for, to keep stealing money from gamers!!!

 

[vanadiel007]

I will skip these and wait on the liquid cooled version of this connector, with 240 mm radiator and RGB fans.

 

[redboy66]

We should call things as they are, the problem is not the connector, the problem is not the cables, the problem is not the PSU.
It is time for Nvidia to LOWER the consumption of its GPU

 

[TJ Hooker]

There is a discussion on what users want in a new connector here:

Proposed new Power Connector

With all the attention on melting graphics card connectors, this might be an opportunity to switch manufacturers to a connector that we like. Here is my proposal - please comment whether you like the idea or not, any changes and if we get enough people behind it we'll try to make them listen...

www.techpowerup.com

Short summary: 48V, one wire for positive, one for ground, two sensing wires to tell whether power wires are overloading (so you don't have to reset fuses) and two data wires to talk to PSU. 48V could also be used to provide motherboard with power ready for USB-C connectors.

This is nearly identical to the 48VHPWR connector that was added the PCIe CEM spec along with 12VHPWR (although not used in any consumer products, AFAIK). To the point I think that guy may have just regurgitated what was in the spec as if he came up with it himself.

But yeah, moving to a higher voltage would certainly make it a lot easier to power high end cards.

 

[InvalidError]

No there isn't, where did you see that? E.g. with 90 C rated insulation the NEC rates 16 AWG copper for 18 amps.

Generic AWG tables intended for cables going through in-wall insulation or conduits with up to three current-carrying pairs per 1" bundle and a normal operating temperature limit of 60C.

Ventilated cabinets like PCs do not need wiring derated for in-wall insulation.

 

[ekio]

Could we get the names of the people accountable for this complete disaster of a plug ?

 

[TryRebooting]

Could we get the names of the people accountable for this complete disaster of a plug ?

Nvidia and PCI SIG . There is no point to witch hunting for a individual, it's the collective responsibility of both organizations leadership to push the development and approval of the connector.

 

[Mr Majestyk]

If the power is balanced 'perfectly' between the six +12V wires it means that 8.3 amps are running through each of those 16ga copper wires. There is a hard limit of 10 amps for 16ga for electrical standards in general. Using 8 pin PCIE cabling keeps the current down to 5.5 amps. The 16 pin connector was almost doomed to have a higher electrical failure rate. Stripping out overcurrent controls have only hastened it.

The standard is tighter than that, it's between 8.5A-9A for these connectors. So even the best case is 9% margin of safety and worst case is ~2%

 

[endocine]

We should call things as they are, the problem is not the connector, the problem is not the cables, the problem is not the PSU.
It is time for Nvidia to LOWER the consumption of its GPU

Cost/Kwh is only going up, it does matter for a lot of people, and these things just keep using more and more of it every generation, its a backwards step

 

[TJ Hooker]

Generic AWG tables intended for cables going through in-wall insulation or conduits with up to three current-carrying pairs per 1" bundle and a normal operating temperature limit of 60C.

Ventilated cabinets like PCs do not need wiring derated for in-wall insulation.

Sure, I wasn't trying to imply my link was directly applicable to 12VHPWR or PC power connectors in general. It was just an easy example that shows there clearly isn't some sort of general, widely-accepted rule that says no more than 10 A over 16 AWG copper.

 

[melvis72]

Cablemod did recall the 1.0 adaptor, but it should have been noted they also issued a stop use of the replacement 1.1 and instructed users for a refund. I still have both sitting on my desk and I was hoping for an update, but it seems like they gave up due to issues on the design.

 

[qxp]

This is nearly identical to the 48VHPWR connector that was added the PCIe CEM spec along with 12VHPWR (although not used in any consumer products, AFAIK). To the point I think that guy may have just regurgitated what was in the spec as if he came up with it himself.

Actually no, this is the first time anyone mentioned it. I looked online and could not find the specs for PCIe CEM 48VHPWR - they appear to be behind a paywall. Do you have a link ? I am curious to compare.

But yeah, moving to a higher voltage would certainly make it a lot easier to power high end cards.

It is not just the higher voltage. There is a way to see how much voltage drop you have from power supply to the card so that the output can be disconnected when it is too big due to poor contact or excessive current.

 

[YSCCC]

Sure, I wasn't trying to imply my link was directly applicable to 12VHPWR or PC power connectors in general. It was just an easy example that shows their clearly isn't some sort of general, widely-accepted rule that says no more than 10 A over 16 AWG copper.

obviously from De8auer's video it takes 23A and still didn't melt, but the wire itself IMO is less of an issue, the pins and the resulting resistance (heat) in a plastic housing is the issue, since plastic don't conduct heat well going anywhere it will get hot isn't a good idea. In consumer electronics putting inside a plastic housing, the smaller your connector it gets the less you should allow it for heat generation

 

[DingusDog]

Power cables should be over engineered period. Hopefully a hefty class action lawsuit will ensure this sort of thing never happens again.

 

[InvalidError]

Sure, I wasn't trying to imply my link was directly applicable to 12VHPWR or PC power connectors in general. It was just an easy example that shows there clearly isn't some sort of general, widely-accepted rule that says no more than 10 A over 16 AWG copper.

I wasn't trying to contradict what you wrote. You asked where the other guy got his number from and the "ampacity" most people quote all of the time are from AWG tables intended for bundles in insulated walls. The worst-case scenario.

 

[FunWithComputers]

Any signs that buyers are having second thoughts about these GPUs on this basis? Any buyers looking to refund on the basis of this stuff?
Surely the wheels are coming off the hype train soon?
I quite liked der8auer's video - which seemed to be saying the problem was partly cutting the number of capacitors at the GPU end, and not monitoring the input from certain cables separately. Pretty surprising that the design of the cards at this basic level is open to question.

 

[Notton]

What I don't understand is why they went with 12 mini power connectors if they're going to fuse them to a single rail.

Take a look at what an XT90 can do with just 2 wires.
And the kicker is it's smaller, safer, and higher endurance than the 12VHPWR.
The cables aren't even that thick.

 

[BlueRaja]

It has been discovered that using previous-generation 12VHPWR cables with the RTX 5090 can result in melting issues regardless.

I keep seeing this misinformation all over the Internet, but it's incorrect. There is no "previous generation cable". 12VHPWR and 12V2x6 use the exact same cable. The only difference was in the connector pins, which are part of the GPU, not the cable

See for instance the Wikipedia page on 12VHPWR:

The connector was replaced by a minor revision called 12V-2x6 [..] The cables and their plugs remained unchanged.

 

[DingusDog]

12VHPWR has been weighed and measured and found wanting.