Crazy. The trouble with plastic is once you start to hit 100c and its on its way. @120c in most cases its liquid.

vestibule:

The trouble with plastic is once you start to hit 100c and its on its way. @120c in most cases its liquid.

These are nylon connectors and the melting point is well above 150c, usually its 180 and above with people designing to spec not allowing melting points below 200. This user had it plugged in fully? i have an image to respond to that. https://cdn.discordapp.com/attachments/395665775077359626/1065924853217112094/image.png

There are so many plastics that could be used, but yeah nylon makes a lot of sense. No so much for heat but its plug ability. But obvs heat in this case is also a critical component Maybe they could have used one on the more denser plastics. Some thing like a anti static ABS but then that would be potentially brittle. Sure is a conundrum and I'm very glad its not my problem. 😕

For me, the question remains about the "real" need of this new power connector...

tried this CP-8920331 Premium Individually Sleeved 12+4pin PCIe Gen 5 12VHPWR 600W cable, Type 4, BLACK https://www.corsair.com/uk/en/Categories/Products/Accessories-|-Parts/PC-Components/Power-Supplies/Premium-Individually-Sleeved-Type-4-12VHPWR-Cable/p/CP-8920331 as an electronician I was already wary of the dubious decrease in wires but since they all went through that absurdely tiny connector anyway (smaller than a single pcie8 for those who never saw one and think it's larger, it's not it's slightly wider but less high and both the wires and connecting pins are smaller) here's my full experience : - the 1st thing I noticed when I bundled the wires toghether with ties was the warmth of the cables near the gpu connector, not burning hot but like 60°C "oh....." - "oh that's what the comb was for...to spread the wires and help cool them (at least for a portion of their length) - moved the comb around so has to spread them and allow airflow (from the only pc case design that should exist singe corsair crystal 680x and lianli 011 dynamic exist, a separate psu compartment and bottom fans airflow) and it worked the wires were noticeably cooler - "ok fine then it looks pretty ugly to be honest and you need to have a large wide spread of cables running through your case but okay... - no benefit in visual but neither in ease of install as you still need a good 30-35mm gap and you absolutely cannot bend them sideways that's a pretty big bummer "but wait there's more...." - benchmarked a while, open side panel so I can touch the cable fo warmth and check the connection plug was perfect and then... - then came the smell, anyone who already burned a gpu knows it the smell of power stages burning... EMERGENCY SHUTDOWN !!! - aftermath I expected the smell coming from the wires or corsair connectors but nope it was coming from the gpu power stage area where the infamous coils are - instant regret of trying this cable, put back the official nvidia 3-1 I got with the 450w 5090 and....everything was fine again and has been for 40hrs+ of gaming I only hope I didn't damage much my card but it was that's a certainty...I don't care to know why or how this happened I already have a work and Corsair and nvidia aren't paying me all I can say is don't even try or blame only yourself, you have been warned

Astyanax:

These are nylon connectors and the melting point is well above 150c, usually its 180 and above with people designing to spec not allowing melting points below 200. This user had it plugged in fully? i have an image to respond to that. https://cdn.discordapp.com/attachments/395665775077359626/1065924853217112094/image.png

Fully in or not is not the focus here.... Its a bullshit risk design. Why the hell make it so difficult for any user to use this piece of shite without running the risk of melting. Seriously? This is not a design geared towards customer ease-of-use or safety. Its geared towards NVidia's ever expanding need to fill its own pockets by lowering costs and quality of components to maximise profit.

kakiharaFRS:

tried this CP-8920331 Premium Individually Sleeved 12+4pin PCIe Gen 5 12VHPWR 600W cable, Type 4, BLACK https://www.corsair.com/uk/en/Categories/Products/Accessories-|-Parts/PC-Components/Power-Supplies/Premium-Individually-Sleeved-Type-4-12VHPWR-Cable/p/CP-8920331 as an electronician I was already wary of the dubious decrease in wires but since they all went through that absurdely tiny connector anyway (smaller than a single pcie8 for those who never saw one and think it's larger, it's not it's slightly wider but less high and both the wires and connecting pins are smaller) here's my full experience : - the 1st thing I noticed when I bundled the wires toghether with ties was the warmth of the cables near the gpu connector, not burning hot but like 60°C "oh....." - "oh that's what the comb was for...to spread the wires and help cool them (at least for a portion of their length) - moved the comb around so has to spread them and allow airflow (from the only pc case design that should exist singe corsair crystal 680x and lianli 011 dynamic exist, a separate psu compartment and bottom fans airflow) and it worked the wires were noticeably cooler - "ok fine then it looks pretty ugly to be honest and you need to have a large wide spread of cables running through your case but okay... - no benefit in visual but neither in ease of install as you still need a good 30-35mm gap and you absolutely cannot bend them sideways that's a pretty big bummer "but wait there's more...." - benchmarked a while, open side panel so I can touch the cable fo warmth and check the connection plug was perfect and then... - then came the smell, anyone who already burned a gpu knows it the smell of power stages burning... EMERGENCY SHUTDOWN !!! - aftermath I expected the smell coming from the wires or corsair connectors but nope it was coming from the gpu power stage area where the infamous coils are - instant regret of trying this cable, put back the official nvidia 3-1 I got with the 450w 5090 and....everything was fine again and has been for 40hrs+ of gaming I only hope I didn't damage much my card but it was that's a certainty...I don't care to know why or how this happened I already have a work and Corsair and nvidia aren't paying me all I can say is don't even try or blame only yourself, you have been warned

Well... that only makes me think, better not buy into that connector. Although it probably won't get better down the road, I doubt Nvidia will go back to the old connectors and multiple cable setups. Hope your hardware as fine! Did you test some more, is the rig running normal Shutdowns? Bad smells still coming from the PSU?

I was never thrilled about using this connector, and was pretty loud about it when it arrived. That said, I use one now and it has given me no grief, both with the Nvidia adapter and lately with a 2 8 pin to 12WHPWR from modDIY.com with a RTX 4080. It is what it is.

Astyanax:

These are nylon connectors and the melting point is well above 150c, usually its 180 and above with people designing to spec not allowing melting points below 200. This user had it plugged in fully? i have an image to respond to that. https://cdn.discordapp.com/attachments/395665775077359626/1065924853217112094/image.png

that image is proof that the connection sucks and yet you defend it as if you work for Nvidia?...that s*** is melted. user error my a**

Airbud:

that image is proof that the connection sucks and yet you defend it as if you work for Nvidia?...that s*** is melted. user error my a**

Yeah, sure, blame the electrican because your 2400w electric frypan wasn't plugged in all the way and it melted the terminal and burned your house down. Ignorance is User Error. Lying about it is fraud.

I suspect another issue beyond the flawed design is the quality. Most of our connectors are being made as cheaply as possible (always have) so this new, flawed design is just a recipe for potential failure.

Astyanax:

These are nylon connectors and the melting point is well above 150c, usually its 180 and above with people designing to spec not allowing melting points below 200. This user had it plugged in fully? i have an image to respond to that. https://cdn.discordapp.com/attachments/395665775077359626/1065924853217112094/image.png

Yep, it's the typical stress marks from earlier findings. It wasn't plugged in properly.

Cablemod 8pin also melt, just shitty QC i guess https://www.overclock.net/threads/cablemod-pci-e-cable-connector-melted-in-my-rtx3090.1795455/

Astyanax:

Yeah, sure, blame the electrican because your 2400w electric frypan wasn't plugged in all the way and it melted the terminal and burned your house down. Ignorance is User Error. Lying about it is fraud.

As I've mentioned in another thread: These are supposed to be built to be idiot proof. There's clip that snaps in, which should inform the user it is fully connected. If the clip snaps and the connector still burns, it's not user error. That's a design flaw no matter what way you look at it. You can't the average person to push the connector in as far as it'll go - most people who try to upgrade their RAM are afraid of snapping the motherboard. You can't expect shipped PCs to not have that connector slide out a millimeter, let alone have the consumer know to check on it.

If you look at the pictures, it's clear as day that this guy did a stupid, and is blaming Cablemod. He clearly looped the cable underneath the GPU, had the connector smashed up against the side-panel, bending it down, and causing a bad connection on the top row of pins.

schmidtbag:

As I've mentioned in another thread: These are supposed to be built to be idiot proof. There's clip that snaps in, which should inform the user it is fully connected. If the clip snaps and the connector still burns, it's not user error. That's a design flaw no matter what way you look at it. You can't the average person to push the connector in as far as it'll go - most people who try to upgrade their RAM are afraid of snapping the motherboard. You can't expect shipped PCs to not have that connector slide out a millimeter, let alone have the consumer know to check on it.

the clip has never "clicked" in any of these cases. you know what you do to make sure somethings connected properly and latched in properly, you tug on it.

Astyanax:

the clip has never "clicked" in any of these cases. you know what you do to make sure somethings connected properly and latched in properly, you tug on it.

How do you know the clip never clicked? For argument's sake, let's say the clip did click, and yet the connector burned anyway - that suggests the tolerances were too lose. While tugging on it is normally a good way to ensure whether it's firmly in place, the irony is that could have been enough to cause just enough of a gap to be a problem. No matter how you look at it though, there is a design flaw when there are so many burnt connectors.

schmidtbag:

How do you know the clip never clicked? For argument's sake, let's say the clip did click, and yet the connector burned anyway - that suggests the tolerances were too lose. While tugging on it is normally a good way to ensure whether it's firmly in place, the irony is that could have been enough to cause just enough of a gap to be a problem. No matter how you look at it though, there is a design flaw when there are so many burnt connectors.

Are there "so many" burnt connectors though? I'm pretty sure this has happened maybe a couple dozen times out of the hundreds of thousands of 3090 Ti/4090/4080's out there, and people are just eager to turn this into anti-Nvidia rage. You'll find plenty of burnt 8 pin connectors out there too if you go searching, but, just like these, those are all people being careless, and people took it for what it was since everyone uses 8-pin connectors instead of having Nvidia solely to point the finger at. I mean it's been weeks since we've even heard about this happening, and this particular case is clearly this guy being an idiot and trying to blame it on Cablemod.

This cable must have been designed after the GPU was, and power requirements were re-assessed.

Still feel like part of the problem is how big those cards are. They barely fit most midle tower cases. Often the cable has to be squeezed for the card to fit. I have a Corsair Airflow 4000D and looking at how big most 4080 are they would hardly fit in my case. Sqeezing the cable to fit the card surely doesn't help.