News Technician Repairs Hundreds of RTX 4090s With Melted Power Connectors Every Month

[TJ Hooker]

The new 12 pin connector is only using (12) 18 gauge wires instead of the 16 needed to hit the right gauge of wire for their 450w max

12VHPWR requires 16 gauge, not 18.

 

[TJ Hooker]

Didn't Steve also tested that sometimes the connector pulls more power from certain cables and not well balanced between all cables?

In the GN video they outright cut some of the 12VHPWR wires (I want to say at least 2), and still didn't see any meaningful rise in temp. So the current imbalance would have to be pretty severe.

 

[King_V]

The only changes made for 12V-2x6 are to shorten/retract the sense pins, such that they are last contact/first break. How does that help with anything other than improperly seated connectors?

That isn't the only change that was made.

New 12V-2x6 PSU Connector Tested: Keeps Cool, Even When Improperly Inserted

Thermals on the new connector are significantly lower compared to its 12VHPWR predecessor, even in worst-case scenarios.

www.tomshardware.com

Hardware Busters says the shortened connecting pins on the new connector only constitute part of the reason why the new connector works so well, in reference to the 12VHPWR connector. The primary upgrade that makes the connector work really well, even with an improper insertion, is its more conductive connecting pins that lower the voltage resistance of power coming through the connector. According to Hardware Busters, the lower resistance pins are the reason why thermals are so low, to the point where overheating is nearly impossible even with improper insertions.

 

[TJ Hooker]

That isn't the only change that was made.

New 12V-2x6 PSU Connector Tested: Keeps Cool, Even When Improperly Inserted

Thermals on the new connector are significantly lower compared to its 12VHPWR predecessor, even in worst-case scenarios.

www.tomshardware.com

Unfortunately that's just sloppy, incorrect reporting by TH (which I pointed out in the comments on that article). Whatever changes Linewell made to their cables aren't due to the change from 12VHPWR to 12V-2x6. This is evident by the fact that Linewell makes the power cables, which have received no changes with the new spec; it's only the board headers that changed, not the connectors attached to the cables. The new "pins" referred to are in fact the female terminals located in the cable plug (evident in the video around 5:45). Which, again, received no changes in the move to 12V-2x6.

From Aris, the person whose video TH is referencing in that article:

The “Cable Plus [sic]” side of the connector has not changed and is compatible with the new PCB header connector definition

ATX v3.1 & PCIe CEM 5.1 are official! - Hardware Busters

Hardware Busters - ATX v3.1 & PCIe CEM 5.1 are official! - PSUs

hwbusters.com

Intel's ATX PSU design guide also highlights the differences for 12V-2x6, which are purely related to sense pins on the board headers, see "Special Note on 12V-2x6 vs. 12VHPWR connectors" here:

https://edc.intel.com/content/www/us/en/design/ipla/software-development-platforms/client/platforms/alder-lake-desktop/atx-version-3-0-multi-rail-desktop-platform-power-supply-design-guide/2.1a/pci-express-pcie-add-in-card-connectors-recommended/

You can also look at the PCIe CEM 5.1 spec with change bars on to see there's no other physical changes*. But that requires a PCI SIG membership (unless they end up getting leaked on the web, not sure, haven't looked).

* There are some added notes/recommendations, see below.

 

[TJ Hooker]

To add on to my previous comment, there are some other changes to spec in the way of implementation recommendations/notes, rather than explicit electro-mechanical changes to the connectors. Stuff like (paraphrasing):

• There could be variation in contact resistance, such that there could be an imbalance in current between individual pins
• The connector could be subjected to side-loading, which could affect contact resistance
• There should be a minimum length of free wire coming out of the cable plug that should not be subjected to significant strain (i.e. the minimum 30mm of clearance for the 12pin connector I'm sure we've all heard of by now)

And that you should design/test with those points in mind.

As far as I can tell, all these notes are essentially already covered in the specs for the actual components that would be used to implement a 12pin PCIe power solution (e.g. amphenol Pwr CEM-5 connectors/headers), or in existing industry standard (e.g. EIA-364-23 Low Level Contact Resistance Test), or both. I'm not a hardware designer, but it seems like it's just taking best practices and standards that already exist in the industry, and highlighting a few of the most pertinent ones. I.e. I don't think any of it should be new info for a competent engineer in that field.

 

[MJS WARLORD]

I wonder how many people were going to buy one not only for its performance but to be able to say to others

Ha Ha .... look what i got ( knowing his pals cant afford one ) , i bet a lot of users changed their mind due to their failure rate .

I have now put myself in the ..... i saw the warnings and changed my mind category

 

[Phaaze88]

He's back with more!

View: https://www.youtube.com/watch?v=6H9XeH8G_mM


Who do you blame?