• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Check Your 4090's 12vhpwr Connectors To Make Sure They are in Pristine Condition

all these cables are made in china, shipped from china it seems... even cablemod they don't even tell you on them what model PSU they are for... just blank packaging


probably a 1 man company drop shipping from china
 
^^ good stuff

I agree, when your get your 4090 adaptor or cables (even if it's Cablemod) - inspect the connector for debris, perhaps blow some air if there is any loose debris to remove it. Then insert the cable straight, do not wiggle it and make sure it's fully in and then follow the normal rules on not bending the cable too sharply. That should prevent any issue other than if the cable has a manufacturing failure that you cannot prevent.

4090fe: let's make a great looking, excellent piece of engineering
Connector: Let's make an unsightly, fragile as glass mess, position it facing a wall and pump 600w thought it. Genius
 
Some more info here.

Yep and as I thought the longer this went on it was user error, In all honesty I'm glad GN did this research and video to call out all other so called experts because as a user like myself when you hear techtubers like Igor and Jay just guessing and blaming Nvidia a lot of people would listen when the fact is I've seen some pics of some of these on this Reddit and you can clearly see a gap between the plug base and the connector tip.

We know no user likes to be blamed for the issue or because it happened to them but it wouldn't surprise me in the slightest that the users that it has happened to they have tried to cover up their mistake or oversight by trying not to show that it wasn't connected properly.

It's not a dig at inexperienced users per say and the connector is stiff I know that from first hand experience but come on if you're really not that competent in connecting a cable together then you really shouldn't be going near this sort of equipment.

Look we all make mistakes I have in the past but experience has always taught make sure everything is plugged and seated to the best of your knowledge, There should not be any gap in any connector on the PC, You may get away with a PWM fan header not connected properly because of the low power but you're not going to get away with that at these sort of wattages.
 
Yep and as I thought the longer this went on it was user error, In all honesty I'm glad GN did this research and video to call out all other so called experts because as a user like myself when you hear techtubers like Igor and Jay just guessing and blaming Nvidia a lot of people would listen when the fact is I've seen some pics of some of these on this Reddit and you can clearly see a gap between the plug base and the connector tip.

We know no user likes to be blamed for the issue or because it happened to them but it wouldn't surprise me in the slightest that the users that it has happened to they have tried to cover up their mistake or oversight by trying not to show that it wasn't connected properly.

It's not a dig at inexperienced users per say and the connector is stiff I know that from first hand experience but come on if you're really not that competent in connecting a cable together then you really shouldn't be going near this sort of equipment.

Look we all make mistakes I have in the past but experience has always taught make sure everything is plugged and seated to the best of your knowledge, There should not be any gap in any connector on the PC, You may get away with a PWM fan header not connected properly because of the low power but you're not going to get away with that at these sort of wattages.


It wasn't only user error.
It was a number of things.
 
I like the solution that Steve and others have brought up. And this might be what PCI SIG is working on.

Since the connector already has 4 sense pins that's designed to be used for non power delivery purposes like identifying operational modes and only 2 pins are currently in use. They should just use the 3rd sense on the connectors and cut the pin's size down significantly to the point where it would only make contact when all the 12v pins are making good contact and then a simple yes/no logic would tell the GPU is the power cable is correctly inserted or not and if not, don't turn on.
 
Last edited:
So the percentages of melting cables was less than 1% and it is likely a 99% user error caused issue.

So much drama.
The adapters have two different types of connectors yet only one type of connector is showing melted in a reddit megathread.

A design that requires perfection in manufactuaring and implementation to avoid catastrophic failure is a bad design.

To be fair to the design, it seems to require two errors at the same time before it fails so terribly. (I could see this issue not getting caught in testing)

Failing to plug it in all the was doesn’t cause it to melt.

Bending the cable doesn’t cause it to melt.

Laterally loading a connector that *also* isn't fully inserted seems to be the most reliable way to get the connector to melt.

Some users have reported that some of these connectors simply don't want to plug in. Pictures of connectors with plastic *inside* the slot pins have been posted.

The 12VHPWR connector seems to lend itself to not being fully seated in a way that normal PCIe 8-pins don't.

Couple that with the fact that most of these cards can't fit into most cases without the cable hitting the side panel and causing some sort of bend in the cables, and you get a megathread exposing the issue.

The shortened sense pin change seems like an effective way to avoid the worst mode of failure.
 
The adapters have two different types of connectors yet only one type of connector is showing melted in a reddit megathread.

A design that requires perfection in manufactuaring and implementation to avoid catastrophic failure is a bad design.

To be fair to the design, it seems to require two errors at the same time before it fails so terribly. (I could see this issue not getting caught in testing)

Failing to plug it in all the was doesn’t cause it to melt.

Bending the cable doesn’t cause it to melt.

Laterally loading a connector that *also* isn't fully inserted seems to be the most reliable way to get the connector to melt.

Some users have reported that some of these connectors simply don't want to plug in. Pictures of connectors with plastic *inside* the slot pins have been posted.

The 12VHPWR connector seems to lend itself to not being fully seated in a way that normal PCIe 8-pins don't.

Couple that with the fact that most of these cards can't fit into most cases without the cable hitting the side panel and causing some sort of bend in the cables, and you get a megathread exposing the issue.

The shortened sense pin change seems like an effective way to avoid the worst mode of failure.
Don't disagree that things could be better, always space for improvement.

On the other hand, if it was simply down to such a poor design we would have a vastly higher incident rate and not the less than 1%. The single biggest cause of this issue is user error.

Majority of the mega threads on this topic, here and other places has had more people spouting ill informed / uninformed nonsense or opportunistic trolls causing grief and nothing to do with the frequency of the issue.
 
User error…
User error…

Refuse to accept user error as being the error

USER ERROR

Fully insert pins - NO I LIKE IT DANGLING

I want to blame someone for my stupidity - Nvidia it is all your fault.

Honestly, this connector should just be avoided until the format matures a bit tbh. But someone had to make the first jump. Might as well be the big bad Nvidia
 
Last edited:
User error…
User error…

Refuse to accept user error as being the error

USER ERROR

Fully insert pins - NO I LIKE IT DANGLING

I want to blame someone for my stupidity - Nvidia it is all your fault.

Honestly, this connector should just be avoided until the format matures a bit tbh. But someone had to make the first jump. Might as well be the big bad Nvidia
its not all the users fault if your connector is built to shoddy standards that dont lock in correctly and can be containated from the factory but you keep doing you and shill for nvidia all you like
 
Last edited:
In my experience, the reasoning around the connector not being fully inserted I could believe.
Over a good 20 yrs I have fixed many friends new pc builds and whilst troubleshooting there have been a lot of times a connector somewhere hasn't been pushed in far enough. It wouldn't always be the cause of the issue, but sometimes was. I think it was a mixture of inexperience, not knowing how important it is to fully insert a cable and being too careful thinking any more pressure may break things.
 
User error…
User error…

Refuse to accept user error as being the error

USER ERROR

Fully insert pins - NO I LIKE IT DANGLING

I want to blame someone for my stupidity - Nvidia it is all your fault.

Honestly, this connector should just be avoided until the format matures a bit tbh. But someone had to make the first jump. Might as well be the big bad Nvidia

Could have provided instructions with it saying IMPORTANT information to follow to make sure? Never heard of people having issues with PCIe cables? Pretty sure it isn't the first time someone hasnt fully inserted cable yeah now with this cable pulling more power they should have come with extra warning to make sure

Respect to the beta testers thank Nvidia while paying premium price for it lol
 
Last edited:
Back
Top Bottom