Jump to content

Boys, Keep your power dry and your adapter cables straight!

Mr.Vic20

By Mr.Vic20
in The Spawn Point

 Share

Recommended Posts

Spork3245

Spork3245

Posted
Posted
1 minute ago, Mr.Vic20 said:

I will likely do this for no other reason than its a better cable management solution. I hear you on the "in the back of my mind" element, fortunately work owns most of that space so I manage to forget about this until returning to this thread! :p


I’d go for the new right angle version for best cable management unless you’re vertically mounting your gpu. :santasun:

I’m going to attempt to “tie” my strimmer+ light-assets on top of the 12vhpwr cable so I didn’t care about the right angle one :lol: 

  • Haha 1
Link to comment
Share on other sites
Mr.Vic20

Mr.Vic20

Posted
  • Author
Posted
28 minutes ago, Spork3245 said:

Oh, the right angle thing is just an adapter, you still need a cable :money:

Ho now, what? so the 90 degree adapter doesn't run directly back to the PSU? I still need 3x8 pins? If so, can I just use the ones I'm currently running from the PSU, in other words, just swap the current adapter for the 90 degree one? 

Link to comment
Share on other sites
Spork3245

Spork3245

Posted
Posted
34 minutes ago, Mr.Vic20 said:

Ho now, what? so the 90 degree adapter doesn't run directly back to the PSU? I still need 3x8 pins? If so, can I just use the ones I'm currently running from the PSU, in other words, just swap the current adapter for the 90 degree one? 


No, the 90 degree adapter attaches to the GPU and takes a 12vhpwr cable, it’s just an adapter and not a cable. You could plug the 3x8pin adapter in it, or just a direct 12vhpwr cable. It’s like those right angle 24pin motherboard adapters and right angle 8pin adapters: just changes the angle of where you plug in. It’s purely for cable management.

  • Halal 1
Link to comment
Share on other sites
Mr.Vic20

Mr.Vic20

Posted
  • Author
Posted

Gentlemen, my PC woes are solved! The issue turnes out to be: Updating intel ME firmware! Apprently ASUS own website has the wrong one and it took crawling through their forums to find a thread by a guy who directed me to the right one. Now, all is well! :sun:

  • Hugs 1
Link to comment
Share on other sites
Spork3245

Spork3245

Posted
Posted
Just now, stepee said:

 

They better just ship them and not make me ship back the bad ones :P


I don’t think they want the bad ones back. I’m unsure if it’s a coupon code for a free cable (ASUS may or may not be doing this: I saw some codes in the wild from ASUS) or if Zotac is going to ship directly. Supposedly it’s happening in the next week.

  • Thanks 1
Link to comment
Share on other sites
cusideabelincoln

cusideabelincoln

Posted
Posted

Ugh, if this trend continues then it's not a good sign...

 

Maybe the first burnt connector with native ATX3.0 cable : nvidia

NEW.REDDIT.COM
0 votes and 38 comments so far on Reddit

 

I would suspect the reason we haven't seen more native 12vhpwr cables fail is because hardly anyone is using them. That one was from an MSI power supply.

 

Some of my own, personal, offshoot theories:

 

1. For some unknown reason the cards' VRMs are pulling more current through one pin over the others. Should not happen as they should all connect to the same plane when soldered into the PCB.

 

2. Maybe since the PCB is smaller and GPU more power hungry compared to previous cards, the PCB's temperature is hotter and that heat is making its way to the power pins. And this is exacerbated when the card is in an enclosed case and/or bad airflow. All of the other reviewers who have tried to force the cable to fail has done so on an open test bench.

 

3. Maybe it's a bad batch of plastic that just melts easily....

 

edit:

 

4. A special combination of any of the existing theories out there, which could explain why this isn't happening to more or most of the connectors.

  • Halal 1
Link to comment
Share on other sites
Spork3245

Spork3245

Posted
Posted
1 hour ago, cusideabelincoln said:

Ugh, if this trend continues then it's not a good sign...

 

Maybe the first burnt connector with native ATX3.0 cable : nvidia

NEW.REDDIT.COM
0 votes and 38 comments so far on Reddit

 

I would suspect the reason we haven't seen more native 12vhpwr cables fail is because hardly anyone is using them. That one was from an MSI power supply.

 

Some of my own, personal, offshoot theories:

 

1. For some unknown reason the cards' VRMs are pulling more current through one pin over the others. Should not happen as they should all connect to the same plane when soldered into the PCB.

 

2. Maybe since the PCB is smaller and GPU more power hungry compared to previous cards, the PCB's temperature is hotter and that heat is making its way to the power pins. And this is exacerbated when the card is in an enclosed case and/or bad airflow. All of the other reviewers who have tried to force the cable to fail has done so on an open test bench.

 

3. Maybe it's a bad batch of plastic that just melts easily....

 

edit:

 

4. A special combination of any of the existing theories out there, which could explain why this isn't happening to more or most of the connectors.


Worth noting that even CableMods is questioning the validity of this (and it’s not one of their cables). The original Facebook post of this shows no pictures of the full cable with the poster purposely gripping it oddly so you can only see the pins. I’m not trying to fling conspiracy theories, but I have a hard time believing the issue goes beyond the adapters.

Link to comment
Share on other sites
Spork3245

Spork3245

Posted
Posted
On 11/4/2022 at 8:05 AM, Spork3245 said:


Worth noting that even CableMods is questioning the validity of this (and it’s not one of their cables). The original Facebook post of this shows no pictures of the full cable with the poster purposely gripping it oddly so you can only see the pins. I’m not trying to fling conspiracy theories, but I have a hard time believing the issue goes beyond the adapters.


Looks like a few more are popping up with non-awful photos. So far the native cable issue seems limited to MSI MEG PSUs and it looks to be melting on the PSU-side more often than the GPU side, so it could be an issue from MSI’s manufacturing. FWIW I’m not overly familiar with MSI’s PSU quality pre-ATX 3.0 (I had no idea that they even made PSUs prior to people recommending them as one of the only native 3.0 units available :p ), so I can’t really comment one way or the other - with PSUs I pretty much only buy EVGA, Corsair, or Seasonic.


I think I said this awhile ago, but the push to move to ATX 3.0 before 3.0 PSUs were really available was always dumb AF to me. It’s like nVidia was rushing to get there because they didn’t want to be made fun of for putting 4 native 8pins on 4090s or something (they should’ve put 3x in the normal spot and a 4th hidden one on the side as an “OC” plug to have the power limit go above 100%). These issues because we’re at 1st gen of a seemingly rushed-out new tech only further cements my thoughts. nVidia would’ve been better off putting resources towards DP2.0 or 2.1 outs instead of this + a bundled adapter cable, as at least DP2/2.1 would make AI Frame Generation 1.0 more viable as we’d eventually have 4k monitors that exceed 120hz so screen tearing would be eliminated due to no vsync. I also think this is the first time that nVidia has been behind on new display technology; they were the first to switch to display port, first to add HDMI, first with HDMI 2.0, first with DP 1.4, and first with HDMI 2.1. Now they’re likely a generation away from adding DP2.1 support.

Link to comment
Share on other sites
Spork3245

Spork3245

Posted
Posted

An engineer at Galax did some tests with the adapter. When plugged in just before it “clicks” (locks) in, the plug would run over 100c under any load. When fully plugged in, it only reached 100c when they forced 1200watts through it (these cables are only supposed to do 600watts max iirc). Some users are now reporting that the plug and/or lock on the adapter may not fit well on some AIBs to where they had to “use a lot of force” before they felt and heard the lock click in. We’re still at the unknown phase, but this leads some credence to the original theory of the plug not being in all the way. This also potentially explains why no YouTubers or hardware sites have been able to successfully recreate the issue so far.

 

 

*I’m not sure if this explains the MSI 3.0 PSU issues as well, as they are melting on the PSU-side

Link to comment
Share on other sites
Spork3245

Spork3245

Posted
Posted
On 11/6/2022 at 6:02 PM, Spork3245 said:

An engineer at Galax did some tests with the adapter. When plugged in just before it “clicks” (locks) in, the plug would run over 100c under any load. When fully plugged in, it only reached 100c when they forced 1200watts through it (these cables are only supposed to do 600watts max iirc). Some users are now reporting that the plug and/or lock on the adapter may not fit well on some AIBs to where they had to “use a lot of force” before they felt and heard the lock click in. We’re still at the unknown phase, but this leads some credence to the original theory of the plug not being in all the way. This also potentially explains why no YouTubers or hardware sites have been able to successfully recreate the issue so far.

 

 

*I’m not sure if this explains the MSI 3.0 PSU issues as well, as they are melting on the PSU-side

 

 

  • Halal 1
Link to comment
Share on other sites
cusideabelincoln

cusideabelincoln

Posted
Posted

Some of the pictures look close to seated, so if it's true that 1-2mm of not being fully engaged is enough to cause this then yet again I have to reiterate what a poor design.

 

Of course all of the pictures show the very tip of the adapter melting first. The tip is closest to the PCB, and most surrounded by insulating plastic when fully inserted, so it gets hottest first and melts. Poor contact with the pins is the most likely cause for the heat buildup, but we still haven't figured out why it's happening. Is it purely manufacturing defect? Or a bad design that doesn't allow for enough tolerance/user variance?

Link to comment
Share on other sites
Spork3245

Spork3245

Posted
Posted

My cablemods cable arrived yesterday. It actually shipped in less than a week (from China, though) and took about 4 days after it shipped to arrive. I inspected my nVidia adapter that the 4090 came with (which I literally had folded and tied in half) after removing it, and there are ZERO signs of melting. Having just the cablemods cable does look better as there’s no bundle of additional wires spewing out of the card, but I’m having an issue figuring out how to properly “attach” my Strimmer+ cable to this until LianLi gets a 12vhpwr Strimmer out, though.

Link to comment
Share on other sites
cusideabelincoln

cusideabelincoln

Posted
Posted

 

 

TLDW:

 

Failures are caused by the pins not making proper contact with the connector, and this can happen with any cable or adapter. 

 

Not being fully inserted is the most common cause of failure, and they showed evidence of real-world failures not being fully inserted.

 

Debris (small plastic bits or metal shavings) can get inside the connector and and form a high-resistance path for current that ends up melting the plastic.

 

Not being fully inserted, or inserted at an angle, can also cause the tips of the connector to make contact with the pins. The tips aren't supposed to make contact under a normal connection, so this is similar to debris in that it forms a high-resistance path as the tips are making poor contact with the pins.

 

In other words, under a normal connection, there are specific contact points inside the metal terminals of the connector that make contact with the pins of the GPU. If these contact points are poor, due to user error, or if a new, poor contact  point is created by debris or user error, then the connector can melt.

 

So @stepee as long as your connector is clean, the terminals look straight and even, and you have it fully inserted into the card then you're safe.

  • Like 1
Link to comment
Share on other sites
Ghost_MH

Ghost_MH

Posted
Posted
On 11/16/2022 at 1:45 PM, cusideabelincoln said:

 

 

TLDW:

 

Failures are caused by the pins not making proper contact with the connector, and this can happen with any cable or adapter. 

 

Not being fully inserted is the most common cause of failure, and they showed evidence of real-world failures not being fully inserted.

 

Debris (small plastic bits or metal shavings) can get inside the connector and and form a high-resistance path for current that ends up melting the plastic.

 

Not being fully inserted, or inserted at an angle, can also cause the tips of the connector to make contact with the pins. The tips aren't supposed to make contact under a normal connection, so this is similar to debris in that it forms a high-resistance path as the tips are making poor contact with the pins.

 

In other words, under a normal connection, there are specific contact points inside the metal terminals of the connector that make contact with the pins of the GPU. If these contact points are poor, due to user error, or if a new, poor contact  point is created by debris or user error, then the connector can melt.

 

So @stepee as long as your connector is clean, the terminals look straight and even, and you have it fully inserted into the card then you're safe.

 

This all screams of poor design, and not on Nvidia but on the spec, itself. We've had these pinned cables like these for decades. The issues GN covers all seem like things that either shouldn't be a worry or should be easy to avoid.

Link to comment
Share on other sites
Spork3245

Spork3245

Posted
Posted
30 minutes ago, Mr.Vic20 said:

 

tldw (stolen from reddit):

•    ⁠Nvidia Official Statements:
    ⁠•    ⁠"We are actively investigating the reports. We are aware of about 50 cases globally. Our findings to date suggest that a common issue is that connectors are not fully plugged into the graphics card. To help ensure the connector is secure we recommend plugging the power dongle into the graphics card first to ensure it's firmly and evenly plugged in, before plugging the graphics card into the motherboard."
    ⁠•    ⁠"We are investigating additional ways to ensure that the connector is secure before powering on the graphics card. NVIDIA and our partners are committed to supporting our customers and ensuring an expedited RMA process, regardless of the cable or card used"
    ⁠•    ⁠"Nvidia has been able to test the cables that were RMA'ed by affected customers. In all of the cases a wear line is clearly visible that indicates the cable wasn't fully inserted into the 16-pin power connector"
    ⁠•    ⁠"Anybody who has an issue [relating to this] will be taken care of. We'll expedite an RMA"
    ⁠•    ⁠"Any issues with the burned cable or GPU, regardless of cable or GPU, it will be processed"
•    ⁠Nvidia Official Article: https://nvidia.custhelp.com/app/answers/detail/a_id/5413 
•    ⁠GN Updated research:
    ⁠•    ⁠0.04% to 0.05% failure rate range
    ⁠•    ⁠Sales source = 125,000 units. Updated GPU sales via 4 board partners
    ⁠•    ⁠Connector burn count source = 50 units. NVIDIA statement
    ⁠•    ⁠50/125000 = 0.04% failure rates
•    ⁠GN again showed that the cable could look like they are inserted but not fully seated and you can wiggle it out. This means you need to push the connector more to fully seat it.
•    ⁠GN referred to Tomshardware article talking about a paragraph within the warranty manual that stated it "may" void manufacturer warranty.
    ⁠•    ⁠GN Confirmed with Nvidia that this is not the case and Nvidia will honor the warranty
•    ⁠GN asked Nvidia about their foreign object debris findings but no confirmation on this side
•    ⁠How to improve the connector
    ⁠•    ⁠GN thinks latching mechanism can be improved
    ⁠•    ⁠Changing the sense pin length where if the connector is not fully seated, the card won't turn on
•    ⁠GN showed a way to properly install the cable
    ⁠•    ⁠Please watch the video

  • Halal 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
 Share
Go to topic listing

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...